SANDIA REPORT

SAND2023-13759
Printed November 2023

Xyce™ Parallel Electronic Simulator

Reference Guide, Version 7.8
Eric R. Keiter, Richard L. Schiek, Heidi K. Thornquist,
Ting Mei, Jason C. Verley, Karthik V. Aadithya,
Joshua D. Schickling, Gary L. Hennigan

Prepared by
Sandia National Laboratories
Albuquerque, New Mexico 87185
Livermore, California 94550
Issued by Sandia National Laboratories, operated for the United States Department of Energy by National Technology &
Engineering Solutions of Sandia, LLC.
NOTICE: This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the
United States Government, nor any agency thereof, nor any of their employees, nor any of their contractors, subcontractors, or their
employees, make any warranty, express or implied, or assume any legal liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process disclosed, or represent that its use would not infringe privately owned
rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or
otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States
Government, any agency thereof, or any of their contractors or subcontractors. The views and opinions expressed herein do not
necessarily state or reflect those of the United States Government, any agency thereof, or any of their contractors.
Printed in the United States of America. This report has been reproduced directly from the best available copy.
Available to DOE and DOE contractors from
U.S. Department of Energy
Office of Scientific and Technical Information
P.O. Box 62
Oak Ridge, TN 37831

Telephone: (865) 576-8401

Facsimile: (865) 576-5728
E-Mail: [email protected]
Online ordering: http://www.osti.gov/scitech
Available to the public from
U.S. Department of Commerce
National Technical Information Service
5301 Shawnee Road
Alexandria, VA 22312

Telephone: (800) 553-6847

Facsimile: (703) 605-6900
E-Mail: [email protected]
Online order: https://classic.ntis.gov/help/order-methods

NT OF E
ME N
RT
ER
A
DEP

GY

• •
IC A
U NIT

ER
ED

ST M
A TES OF A

2
ABSTRACT
This document is a reference guide to the Xyce Parallel Electronic Simulator, and is a companion document
to the Xyce Users’ Guide [1] . The focus of this document is (to the extent possible) exhaustively list device
parameters, solver options, parser options, and other usage details of Xyce. This document is not intended to
be a tutorial. Users who are new to circuit simulation are better served by the Xyce Users’ Guide [1] .

3
ACKNOWLEDGMENTS

We would like to acknowledge all the developers, DevOps engineers, and Scrum team members who have
contributed to the Xyce project over the years: Aaron Gibson, Alan Lundin, Antonio Gonzales, Ashley
Meek, Bart van Bloemen Waanders, Brad Bond, Brian Fett, Christina Warrender, David Baur, David
Collins, David Day, David Shirley, Debbie Serna, Deborah Fixel, Derek Barnes, Eric Rankin, Erik Zeek,
Herman "Buddy" Watts, Hue Lai, Jim Emery, Jonathan Woodbridge, Jonathen Kwok, Josh Smith, Keith
Santarelli, Laura Boucheron, Lawrence Musson, Lon Waters, Mary Meinelt, Michael Skoufis, Mingyu
"Genie" Hsieh, Nicholas Johnson, Peter Sholander, Philip Campbell, Rachel Campbell, Randall Lober,
Rebecca Arnold, Regina Schells, Richard Drake, Robert Hoekstra, Roger Pawlowski, Russell Hooper,
Samuel Browne, Scott Hutchinson, Simon Zou, Smitha Sam, Steven Verzi, Tamara Kolda, Thomas V.
Russo, Timur Takhtaganov, and Todd Coffey.
Also, thanks to for the original typesetting of this document in LATEX.

Trademarks

Xyce Electronic Simulator™ and Xyce™ are trademarks of National Technology & Engineering Solutions
of Sandia, LLC (NTESS).

Contact Information

Address Electrical Models & Simulation Dept.

Sandia National Laboratories
P.O. Box 5800, MS 1177
Albuquerque, NM 87185-1177

Outside Sandia
World Wide Web http://xyce.sandia.gov
Email [email protected]
Inside Sandia
World Wide Web https://info-ng.sandia.gov/xyce/
Email [email protected]

This document is copyright © 2002-2023 National Technology &

Engineering Solutions of Sandia, LLC.

4
CONTENTS

1. Introduction 19
1.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.2. How to Use this Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.3. Typographical conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2. Netlist Reference 21
2.1. Netlist Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.1.1. .AC (AC Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.1.2. .DATA (Data Table for sweeps) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.3. .DC (DC Sweep Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.3.1. Linear Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.3.2. Decade Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.3.3. Octave Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.1.3.4. List Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.1.3.5. Data Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.1.4. .DCVOLT (Initial Condition, Bias point) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1.5. .EMBEDDEDSAMPLING (Embedded Sampling) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.1.6. .END (End of Circuit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.1.7. .ENDS (End of Subcircuit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.1.8. .FFT (FFT Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.1.8.1. .OPTIONS FFT FFT_MODE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.1.8.2. Window Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.1.8.3. Additional FFT Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.1.8.4. Re-Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.1.8.5. Compatibility Between OUTPUT and FFT Options . . . . . . . . . . . . . . . . . . . 37
2.1.9. .FOUR (Fourier Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.1.10. .FUNC (Function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.1.11. .GLOBAL (Global Node) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.1.12. .GLOBAL_PARAM (Global parameter) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.1.13. .HB (Harmonic Balance Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.1.14. .IC (Initial Condition, Bias point) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.1.15. .INC or .INCLUDE or .INCL (Include file) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.1.16. .LIB (Library file) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.16.1. .LIB call statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.16.2. .LIB definition statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.17. .LIN (Linear Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.1.18. .MEASURE or .MEAS (Measure output) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.1.18.1. Measure Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.18.2. Measurement Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1.18.3. Expression Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.1.18.4. Re-Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

5
 2.1.18.5. RISE, FALL and CROSS Qualifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.1.18.6. Additional Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.1.18.7. Suppresssing Measure Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.1.18.8. ERROR Functions (ERR1 and ERR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.1.18.9. ERROR Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.1.18.10. Operator Support for AC Mode Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.1.18.11. Operator Support for NOISE Mode Measures . . . . . . . . . . . . . . . . . . . . . . . . 71
2.1.18.12. Behavior for Unsupported Modes and Types . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.1.18.13. Compatibility with .DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.1.18.14. HSPICE Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.1.18.15. Legacy Trig-Targ Mode (LTTM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.1.19. .MEASURE (Continuous results) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.1.19.1. Measure Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
2.1.19.2. RISE, FALL and CROSS Qualifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
2.1.19.3. AT and TD Qualifiers for TRIG-TARG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.1.19.4. HSPICE Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.1.20. .MEASURE FFT (Measure output for .FFT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
2.1.20.1. Measure Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.1.20.2. Re-Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.1.20.3. Additional Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2.1.21. .MODEL (Model Definition) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
2.1.21.1. LEVEL Parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
2.1.21.2. Model Binning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
2.1.21.3. Length Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
2.1.22. .NODESET (Approximate Initial Condition, Bias point) . . . . . . . . . . . . . . . . . . . . . . . . . 90
2.1.23. .NOISE (Noise Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
2.1.24. .OP (Bias Point Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.1.25. .OPTIONS Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
2.1.25.1. .OPTIONS DEVICE (Device Package Options) . . . . . . . . . . . . . . . . . . . . . . . . 95
2.1.25.2. .OPTIONS DIAGNOSTIC (Diagnostic Simulation Output) . . . . . . . . . . . . . . . 97
2.1.25.3. .OPTIONS TIMEINT (Time Integration Options) . . . . . . . . . . . . . . . . . . . . . . 97
2.1.25.4. .OPTIONS NONLIN (Nonlinear Solver Options) . . . . . . . . . . . . . . . . . . . . . . . 101
2.1.25.5. .OPTIONS LOCA (Continuation and Bifurcation Tracking Package Options) 103
2.1.25.6. .OPTIONS LINSOL (Linear Solver Options) . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.1.25.7. .OPTIONS LINSOL-HB (Linear Solver Options) . . . . . . . . . . . . . . . . . . . . . . 107
2.1.25.8. .OPTIONS LINSOL-AC (Linear Solver Options) . . . . . . . . . . . . . . . . . . . . . . 107
2.1.25.9. .OPTIONS OUTPUT (Output Options) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
2.1.25.10..OPTIONS RESTART (Checkpointing Options) . . . . . . . . . . . . . . . . . . . . . . . . 109
2.1.25.11..OPTIONS RESTART (Restarting Options) . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.1.25.12..OPTIONS RESTART: special notes for use with two-level-Newton . . . . . . . 110
2.1.25.13..OPTIONS SAMPLES (Sampling options) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.1.25.14..OPTIONS EMBEDDEDSAMPLES (Embedded Sampling options) . . . . . . . . . . 111
2.1.25.15..OPTIONS PCES (Fully intrusive PCE options) . . . . . . . . . . . . . . . . . . . . . . . 112
2.1.25.16..OPTIONS SENSITIVITY (Direct and Adjoint Sensitivity Options) . . . . . . 113
2.1.25.17..OPTIONS HBINT (Harmonic Balance Options) . . . . . . . . . . . . . . . . . . . . . . 114
2.1.25.18..OPTIONS DIST (Parallel Distribution Options) . . . . . . . . . . . . . . . . . . . . . . 114
2.1.25.19..OPTIONS FFT (FFT Options) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
2.1.25.20..OPTIONS MEASURE (Measure Options) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

6
 2.1.25.21..OPTIONS PARSER (Parser Options) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.1.26. .PARAM (Parameter) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.1.27. .PCE (Fully Intrusive Polynomial Chaos Expansion (PCE) Analysis) . . . . . . . . . . . . . . 120
2.1.28. .PREPROCESS REPLACEGROUND (Ground Synonym) . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
2.1.29. .PREPROCESS REMOVEUNUSED (Removal of Unused Components) . . . . . . . . . . . . . . . 124
2.1.30. .PREPROCESS ADDRESISTORS (Adding Resistors to Dangling Nodes) . . . . . . . . . . . . 125
2.1.31. .PRINT (Print output) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
2.1.31.1. Print AC Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
2.1.31.2. Print DC Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
2.1.31.3. Print Harmonic Balance Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
2.1.31.4. Print Noise Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
2.1.31.5. Print Transient Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
2.1.31.6. Print Homotopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
2.1.31.7. Print Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
2.1.31.8. Print Embedded Sampling Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
2.1.31.9. Print Intrusive PCE Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
2.1.31.10. Parameter Stepping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
2.1.31.11. Print Wildcards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
2.1.31.12. Device Parameters and Internal Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
2.1.32. .RESULT (Print results) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
2.1.32.1. Example Netlist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
2.1.33. .SAMPLING (Sampling UQ Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
2.1.34. .SAVE (Save operating point conditions) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
2.1.35. .SENS (Compute DC, AC or transient sensitivities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
2.1.36. .STEP (Step Parametric Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
2.1.36.1. Linear Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
2.1.36.2. Decade Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
2.1.36.3. Octave Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
2.1.36.4. List Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
2.1.36.5. Data Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
2.1.37. .SUBCKT (Subcircuit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
2.1.38. .TRAN (Transient Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
2.1.39. Miscellaneous Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
2.1.39.1. * (Comment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
2.1.39.2. ; (In-line Comment) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
2.1.39.3. + (Line Continuation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
2.2. Expressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
2.2.1. Expressions in .PARAM or .GLOBAL_PARAM Statements . . . . . . . . . . . . . . . . . . . . . . . . . 169
2.2.2. Expressions in .PRINT Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
2.2.3. Using Complex Values in Expressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
2.2.4. Expressions for Device Instance and Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 171
2.2.5. POLY expressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
2.2.5.1. Voltage-controlled sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
2.2.5.2. Current-controlled sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
2.2.5.3. B sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
2.3. Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
2.3.1. Voltage Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
2.3.1.1. Global nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

7
 2.3.1.2. Subcircuit Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
2.3.2. Legal Characters in Node and Device Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
2.3.3. Lead Currents and Power Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
2.3.4. Capacitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
2.3.5. Inductor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
2.3.6. Mutual Inductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
2.3.7. Resistor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
2.3.8. Diode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
2.3.9. Independent Current Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
2.3.10. Independent Voltage Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
2.3.11. Port Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
2.3.12. Voltage Controlled Voltage Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
2.3.13. Current Controlled Current Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
2.3.14. Voltage Controlled Current Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
2.3.15. Current Controlled Voltage Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
2.3.16. Nonlinear Dependent Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
2.3.17. Bipolar Junction Transistor (BJT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
2.3.17.1. The Level 1 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
2.3.17.2. VBIC Temperature Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
2.3.17.3. Level 1 BJT Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
2.3.17.4. Level 11 and 12 BJT Tables (VBIC 1.3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
2.3.17.5. Level 23 BJT Tables (FBH HBT_X) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
2.3.17.6. Level 230 BJT Tables (HICUM/L0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
2.3.17.7. Level 234 BJT Table (HICUM/L2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
2.3.17.8. Level 504 and 505 BJT Tables (MEXTRAM) . . . . . . . . . . . . . . . . . . . . . . . . 271
2.3.18. Junction Field-Effect Transistor (JFET) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
2.3.19. Metal-Semiconductor FET (MESFET) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
2.3.20. MOS Field Effect Transistor (MOSFET) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
2.3.20.1. Level 1 MOSFET Tables (SPICE Level 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
2.3.20.2. Level 2 MOSFET Tables (SPICE Level 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
2.3.20.3. Level 3 MOSFET Tables (SPICE Level 3) . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
2.3.20.4. Level 6 MOSFET Tables (SPICE Level 6) . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
2.3.20.5. Level 9 MOSFET Tables (BSIM3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
2.3.20.6. Level 10 MOSFET Tables (BSIM-SOI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
2.3.20.7. Level 14/54 MOSFET Tables (BSIM4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
2.3.20.8. Level 18 MOSFET Tables (VDMOS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
2.3.20.9. Levels 70 and 70450 MOSFET Tables (BSIM-SOI 4.6.1 and 4.5.0) . . . . . . 381
2.3.20.10. Level 77 MOSFET Tables (BSIM6 version 6.1.1) . . . . . . . . . . . . . . . . . . . . . 442
2.3.20.11. Level 102 MOSFET Tables (PSP version 102.5) . . . . . . . . . . . . . . . . . . . . . . 472
2.3.20.12. Level 103 and 1031 MOSFET Tables (PSP version 103.4) . . . . . . . . . . . . . . 490
2.3.20.13. Level 110 MOSFET Tables (BSIM CMG version 110.0.0) . . . . . . . . . . . . . . 563
2.3.20.14. Level 107 and 108 MOSFET Tables (BSIM CMG versions 107.0.0 and
108.0.0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
2.3.20.15. Levels 2000 and 2001 MOSFET Tables (MVS version 2.0.0) . . . . . . . . . . . . 664
2.3.20.16. Level 2002 MOSFET Tables (MVSG_CMC version 1.1.0) . . . . . . . . . . . . . 666
2.3.20.17. Level 260 MOSFET Tables (EKV version 2.6) . . . . . . . . . . . . . . . . . . . . . . . . 673
2.3.20.18. Level 301 MOSFET Tables (EKV version 3.0.1) . . . . . . . . . . . . . . . . . . . . . . 676
2.3.20.19. Level 10240 MOSFET Tables (L_UTSOI Version 102.4.0) . . . . . . . . . . . . . 684

8
 2.3.21.
Lossy Transmission Line (LTRA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
2.3.22.
Voltage- or Current-controlled Switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
2.3.23.
Generic Switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703
2.3.24.
Linear device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
2.3.25.
Lossless (Ideal) Transmission Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
2.3.26.
Lumped Transmission Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
2.3.27.
Ideal Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
2.3.27.1. Delay device instance parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
2.3.28. Behavioral Digital Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711
2.3.29. Y-Type Behavioral Digital Devices (Deprecated) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716
2.3.30. Accelerated mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
2.3.31. Power Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
2.3.32. Memristor Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
2.3.33. Subcircuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
2.4. TCAD Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
2.4.1. Physical Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
2.4.1.1. Material Models and Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
2.4.1.2. Effective Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
2.4.1.3. Electron Effective Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
2.4.1.4. Hole Effective Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
2.4.1.5. Intrinsic Carrier Concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
2.4.1.6. Bandgap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
2.4.2. Mobility Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
2.4.2.1. Analytic Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
2.4.2.2. Arora Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
2.4.2.3. Carrier-Carrier Scattering Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
2.4.2.4. Lombardi Surface Mobility Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
2.4.2.5. Edge Mobilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
2.4.2.6. Boundary Conditions for Electrode Contacts . . . . . . . . . . . . . . . . . . . . . . . . . 761
2.4.2.7. Neutral Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761
2.4.2.8. Schottky Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
2.4.2.9. Metal-Oxide-Semiconductor Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766
2.4.2.10. NMOS Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766

3. Command Line Arguments 769

3.1. Hspice extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
3.2. Redefined parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771

4. Runtime Environment 773

4.1. Running Xyce in Serial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
4.2. Running Xyce in Parallel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
4.3. Running Xyce on Sandia HPC and CEE Platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773

5. Setting Convergence Parameters for Xyce 775

5.1. Adjusting Transient Analysis Error Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
5.1.1. Setting RELTOL and ABSTOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
5.2. Adjusting Nonlinear Solver Parameters (in transient mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776

9
6. Quick Reference for Users of Other SPICE Circuit Simulators 777
6.1. Differences Between Xyce and PSpice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
6.1.1. Command Line Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
6.1.2. Device Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
6.1.3. .OPTIONS Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
6.1.4. .PROBE vs. .PRINT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
6.1.5. Converting PSpice ABM Models for Use in Xyce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
6.1.6. Usage of .STEP Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
6.1.6.1. Model Parameter Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
6.1.7. Behavioral Digital Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 780
6.1.8. Power Dissipation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 780
6.1.9. Dependent Sources with TABLE Syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 780
6.1.10. MODEL STATEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781
6.1.11. .NODESET and .IC Statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781
6.1.12. Piecewise Linear Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
6.1.13. .AC Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
6.1.14. Additional differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
6.1.15. Translating Between PSpice and Xyce Netlists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
6.2. Differences Between Xyce and Other SPICE Simulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
6.3. DC Operating Point Calculation Failures in Xyce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
6.3.1. Incompatible Voltage Constraints at Circuit Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
6.3.2. Multiple Voltage Constraints From Subcircuits or at Global Nodes . . . . . . . . . . . . . . . . 788
6.3.3. NODESET and IC Statements in Subcircuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
6.3.4. No DC Path to Ground for a Current Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
6.3.5. Inductor Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
6.3.6. Infinite Slope Transistions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
6.3.7. Simulation Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789

7. Quick Reference for Microsoft Windows Users 791

8. Rawfile Format 793

8.1. ASCII Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
8.2. Binary Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794
8.3. Special Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794

Bibliography 795

Appendices 798

A. Third Party Licenses 798

Index 815

Distribution 822

10
LIST OF FIGURES

Figure 2-1. Model Binning Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Figure 2-2. Scale Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Figure 2-3. BJT model schematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Figure 2-4. VBIC thermal network schematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Figure 2-5. Lumped Π Model for PowerGridBranch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
Figure 2-6. Equivalent Circuit for PowerGridbusShunt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
Figure 2-7. Equivalent Circuit for PowerGridTransformer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732
Figure 2-8. Neutral Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
Figure 2-9. Schottky Contact, N-type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763
Figure 2-10. Schottky Contact, P-type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764

11
This page intentionally left blank.

12
LIST OF TABLES

Table 1-1. Xyce typographical conventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Table 2-1. .FFT Window Function Definitions (N = Number of Points). . . . . . . . . . . . . . . . . . . . . . . . . 35

Table 2-2. RISE, FALL and CROSS Support in .MEASURE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Table 2-3. Options for Device Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Table 2-4. Options for Diagnostic Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Table 2-5. Options for Time Integration Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Table 2-6. Options for Nonlinear Solver Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Table 2-7. Options for Continuation and Bifurcation Tracking Package. . . . . . . . . . . . . . . . . . . . . . . . . 103
Table 2-8. Options for Linear Solver Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Table 2-9. Options for Linear Solver Package for HB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Table 2-10. Options for Sampling Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Table 2-11. Options for Embedded Sampling Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Table 2-12. Options for PCE Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Table 2-13. Options for Sensitivity Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Table 2-14. Options for HB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Table 2-15. Options for Parallel Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Table 2-16. Options for FFT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Table 2-17. Options for MEASURE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Table 2-18. Options for Parsing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Table 2-19. Print AC Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Table 2-20. Print DC Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Table 2-21. Print HB Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Table 2-22. Print NOISE Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Table 2-23. Print Transient Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Table 2-24. Print Homotopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Table 2-25. Print Sensitivities for .TRAN and .DC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Table 2-26. Print Sensitivities for .AC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Table 2-27. Print Transient Adjoint Sensitivities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Table 2-28. Print EMBEDDEDSAMPLING Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Table 2-29. Print PCE Analysis Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Table 2-32. Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Table 2-33. Arithmetic Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Table 2-34. SPICE Compatibility Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Table 2-35. Analog Device Quick Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Table 2-35. Analog Device Quick Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Table 2-36. Features Supported by Xyce Device Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Table 2-37. Capacitor Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Table 2-38. Capacitor Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Table 2-39. Inductor Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Table 2-40. Inductor Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192

13
Table 2-41. Nonlinear Mutual Inductor Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Table 2-42. Resistor Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Table 2-43. Resistor Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Table 2-44. Resistor Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Table 2-45. Resistor Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Table 2-46. Diode Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Table 2-47. Diode Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Table 2-48. JUNCAP200 Diode Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Table 2-49. JUNCAP200 Diode Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Table 2-50. Diode level 200 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Table 2-51. DIODE_CMC 2.0.0 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Table 2-52. DIODE_CMC 2.0.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Table 2-53. Diode level 2002 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Table 2-54. Pulse Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Table 2-55. Sine Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Table 2-56. Exponent Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Table 2-57. Pattern Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Table 2-58. Piecewise Linear Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Table 2-59. Frequency Modulated Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Table 2-60. Pulse Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Table 2-61. Sine Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Table 2-62. Exponent Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Table 2-63. Pattern Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Table 2-64. Piecewise Linear Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
Table 2-65. Frequency Modulated Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Table 2-66. Bipolar Junction Transistor Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Table 2-67. Bipolar Junction Transistor Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Table 2-68. VBIC 1.3 3T Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Table 2-69. VBIC 1.3 3T Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Table 2-70. VBIC 1.3 4T Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Table 2-71. VBIC 1.3 4T Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Table 2-72. FBH HBT_X v2.1 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Table 2-73. FBH HBT_X v2.1 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Table 2-74. HICUM L0 v1.32 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Table 2-75. HICUM L0 v1.32 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Table 2-76. BJT level 230 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Table 2-77. HICUM v2.4.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Table 2-78. BJT level 234 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Table 2-79. MEXTRAM 504.12.1 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Table 2-80. MEXTRAM 504.12.1 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Table 2-81. BJT level 504 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Table 2-82. MEXTRAM 504.12.1 with self heating Device Instance Parameters . . . . . . . . . . . . . . . . . 278
Table 2-83. MEXTRAM 504.12.1 with self heating Device Model Parameters . . . . . . . . . . . . . . . . . . . 278
Table 2-84. BJT level 505 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Table 2-85. JFET Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Table 2-86. JFET Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Table 2-87. JFET Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Table 2-88. JFET Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

14
Table 2-89. MESFET Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
Table 2-90. MESFET Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
Table 2-91. MOSFET level 1 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Table 2-92. MOSFET level 1 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Table 2-93. MOSFET level 2 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Table 2-94. MOSFET level 2 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Table 2-95. MOSFET level 3 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Table 2-96. MOSFET level 3 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Table 2-97. MOSFET level 6 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Table 2-98. MOSFET level 6 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Table 2-99. BSIM3 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Table 2-100. BSIM3 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Table 2-101. BSIM3 SOI Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
Table 2-102. BSIM3 SOI Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Table 2-103. BSIM4 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Table 2-104. BSIM4 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
Table 2-105. Power MOSFET Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Table 2-106. Power MOSFET Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Table 2-107. BSIM-SOI 4.6.1 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Table 2-109. MOSFET level 70 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Table 2-110. BSIM-SOI 4.5.0 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Table 2-112. MOSFET level 70450 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
Table 2-113. BSIM6 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Table 2-114. BSIM6 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Table 2-115. MOSFET level 77 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
Table 2-116. PSP102VA legacy MOSFET 102.5 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . 472
Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . 473
Table 2-118. MOSFET level 102 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
Table 2-119. PSP103VA MOSFET Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
Table 2-120. PSP103VA MOSFET Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Table 2-121. MOSFET level 103 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
Table 2-122. PSP103VA MOSFET with self-heating Device Instance Parameters . . . . . . . . . . . . . . . . . . 526
Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters . . . . . . . . . . . . . . . . . . . 527
Table 2-124. MOSFET level 1031 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Table 2-125. BSIM-CMG FINFET v110.0.0 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 563
Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 564
Table 2-127. MOSFET level 110 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
Table 2-128. BSIM-CMG FINFET v107.0.0 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 600
Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 600
Table 2-130. MOSFET level 107 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
Table 2-131. BSIM-CMG FINFET v108.0.0 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 630
Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 630
Table 2-133. MOSFET level 108 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Table 2-134. MVS ETSOI 2.0.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
Table 2-135. MVS HEMT 2.0.0 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
Table 2-136. MVSG-HV HEMT MODEL Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 666

15
Table 2-137. MVSG-HV HEMT MODEL Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
Table 2-138. MOSFET level 2002 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
Table 2-139. EKV MOSFET version 2.6 Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
Table 2-140. EKV MOSFET version 2.6 Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
Table 2-141. EKV3 MOSFET Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
Table 2-142. EKV3 MOSFET Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
Table 2-143. L_UTSOI MOSFET Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Table 2-144. L_UTSOI MOSFET Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Table 2-145. MOSFET level 10240 Output Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
Table 2-146. Lossy Transmission Line Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
Table 2-147. Controlled Switch Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
Table 2-148. LIN Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
Table 2-149. Ideal Transmission Line Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706
Table 2-150. Lumped Transmission Line Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
Table 2-151. Lumped Transmission Line Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
Table 2-152. Delay element Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
Table 2-153. Behavioral Digital Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
Table 2-154. Behavioral Digital Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
Table 2-155. DFF Truth-Table for DIGINITSTATE=3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
Table 2-156. DLTCH Truth-Table for DIGINITSTATE=3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715
Table 2-157. JKFF Truth-Table for DIGINITSTATE=3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715
Table 2-158. Behavioral Digital Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717
Table 2-159. Behavioral Digital Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717
Table 2-160. PowerGridBranch Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
Table 2-161. PowerGridBusShunt Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Table 2-162. PowerGridTransformer Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727
Table 2-163. PowerGridGenBus Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
Table 2-164. MemristorTEAM Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735
Table 2-165. MemristorTEAM Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736
Table 2-166. MemristorYakopcic Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736
Table 2-167. MemristorYakopcic Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736
Table 2-168. MemristorPEM Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
Table 2-169. MemristorPEM Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 738
Table 2-170. PDE Device Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
Table 2-171. 1D PDE (level 1) Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
Table 2-172. 2D PDE (level 2) Device Instance Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
Table 2-173. LAYER Composite Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Table 2-174. NODE Composite Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Table 2-175. NODE Composite Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
Table 2-176. DOPINGPROFILES Composite Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
Table 2-177. REGION Composite Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
Table 2-178. Description of the flatx, flaty doping parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 750
Table 2-179. Intrinsic Carrier Concentration Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
Table 2-180. Bandgap constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
Table 2-181. Analytic Mobility Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
Table 2-182. Arora Mobility Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Table 2-183. Carrier-Carrier Mobility Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
Table 2-184. Lombardi Surface Mobility Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760

16
Table 2-185. Material workfunction values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765
Table 2-186. Electron affinities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765

Table 3-1. List of Xyce command line arguments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769

Table 6-1. Incompatibilities with PSpice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783

Table 6-2. Incompatibilities with Other Circuit Simulators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785

Table 7-1. Issues for Microsoft Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791

Table 8-1. Xyce ASCII rawfile format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793

Table 8-2. Xyce binary rawfile format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794

17
This page intentionally left blank.

18
1. INTRODUCTION

Welcome to Xyce™
The Xyce™ Parallel Electronic Simulator has been written to support, in a rigorous manner, the simulation
needs of the Sandia National Laboratories electrical designers. It is targeted specifically to run on
large-scale parallel computing platforms but also runs well on a variety of architectures including single
processor workstations. It also aims to support a variety of devices and models specific to Sandia needs.

19
1.1. Overview

This document is intended to complement the Xyce Users’ Guide [1] . It contains comprehensive, detailed
information about a number of topics pertinent to the usage of Xyce. Included in this document is a netlist
reference for the input-file commands and elements supported within Xyce; a command line reference,
which describes the available command line arguments for Xyce; and quick-references for users of other
circuit codes, such as Orcad’s PSpice [2].

1.2. How to Use this Guide

This guide is designed so you can quickly find the information you need to use Xyce. It assumes that you are
familiar with basic Unix-type commands, how Unix manages applications and files to perform routine tasks
(e.g., starting applications, opening files and saving your work). Note that while Windows versions of Xyce
are available, they are command-line programs meant to be run under the Command Prompt, and are used
almost identically to their Unix counterparts.

1.3. Typographical conventions

Before continuing in this Reference Guide, it is important to understand the terms and typographical
conventions used. Procedures for performing an operation are generally indicated with the following
typographical conventions.

Notation Example Description

Commands entered from the
Typewriter text mpirun -np 4 keyboard on the command line
or text entered in a netlist.
Set nominal temperature using SPICE-type parameters used in
Bold Roman Font
the TNOM option. models, etc.
Feature that is designed
Gray Shaded Text DEBUGLEVEL primarily for use by Xyce
developers.
[text in brackets] Xyce [options] <netlist> Optional parameters.
Parameters to be inserted by the
<text in angle brackets> Xyce [options] <netlist>
user.
Parameter that may be multiply
<object with asterisk>* K1 <ind. 1> [<ind. n>*]
specified.
.PRINT TRAN Parameters that may only take
<TEXT1|TEXT2>
+ DELIMITER=<TAB|COMMA> specified values.
Table 1-1.: Xyce typographical conventions.

20
2. NETLIST REFERENCE

Chapter Overview
This chapter contains reference material directed towards working with circuit analyses in Xyce using the
netlist interface. Included are detailed command descriptions, start-up option definitions and a list of
devices supported by the Xyce netlist interface.

21
2.1. Netlist Commands

This section outlines the netlist commands that can be used with Xyce to setup and control circuit
analysis.

2.1.1. .AC (AC Analysis)

Calculates the frequency response of a circuit over a range of frequencies.

The .AC command can specify a linear sweep, decade logarithmic sweep, octave logarithmic sweep, or a
data table of multivariate values.

General Form .AC <sweep type> <points value>

+ <start frequency value> <end frequency value>

Examples .AC LIN 101 100Hz 200Hz

.AC OCT 10 1kHz 16kHz
.AC DEC 20 1MEG 100MEG
.AC DATA=<table name>

.param points=101, start=100Hz, end=200Hz

.AC LIN {points} {start} {end}

Arguments and
Options
sweep type
Must be LIN, OCT, DEC, or DATA as described below.
LIN Linear sweep
The sweep variable is swept linearly from the starting to the ending
value.
OCT Sweep by octaves
The sweep variable is swept logarithmically by octaves.
DEC Sweep by decades
The sweep variable is swept logarithmically by decades.
DATA Sweep values from a table
Sweep variables are applied based on the rows of a data table. This
format allows magnitude and phase to be swept in addition to frequency.
If using this format, no other arguments are needed on the .AC line.
points value
Specifies the number of points in the sweep, using an integer greater than or
equal to 1.

22
 start frequency value
end frequency value
The end frequency value must not be less than the start frequency value, and
both must be greater than zero. The whole sweep must include at least one
point.

Comments AC analysis is a linear analysis. The simulator calculates the frequency response by
linearizing the circuit around the DCOP bias point.
If specifying the sweep points using the DATA type, one can also sweep the magnitude
and phase of an AC source, as well as the values of linear model parameters. However,
unlike the use of DATA for .STEP and .DC, it is not possible to sweep nonlinear device
parameters. This is because changing other nonlinear device parameters would alter
the correct DCOP solution, and the AC sweep happens after the DCOP calculation in
the analysis flow. To sweep a nonlinear device parameter on an AC problem, add a
.STEP command to the netlist to provide an outer parametric sweep around the
analysis.
A .PRINT AC must be used to get the results of the AC sweep analysis. See Section
2.1.31.
Some devices that may be expected to work in AC analysis do not at this time. This
includes, but is not limited to, the lossy transmission line (LTRA) and lossless
transmission line (TRA). The LTRA and TRA models will need to be replaced with
lumped transmission line models (YTRANSLINE).
Power calculations (P(<device>) and W(<device>) are not supported for any
devices for AC analysis. Current variables (e.g., I(<device>)) are only supported for
devices that have “branch currents” that are part of the solution vector. This includes
the V, E, H and L devices. It also includes the voltage-form of the B device.

23
2.1.2. .DATA (Data Table for sweeps)

User-defined data table, which can be used to specify sweep points for .AC, .DC, .NOISE or .STEP

General Form .DATA [<name>]

+ <parameter name> [parameter name]*
+ <parameter value> [parameter value]*
.ENDDATA

Examples .data test

+ r1 r2
+ 8.0000e+00 4.0000e+00
+ 9.0000e+00 4.0000e+00
.enddata

Arguments and
Options name
Name of the data table.
parameter name
Name of sweep parameter. This can be a device instance parameter, a device
model parameter or a user-defined parameter specified using
.GLOBAL_PARAM or a globally scoped .PARAM statement.
parameter value
Value of sweep parameter for the given sweep point. This must be a double
precision number. Each row of the table corresponds to a different sweep
step, so multiple parameters can be changed simultaneously.

Comments Each column of a data table corresponds to a different parameter, and each row
corresponds to a different sweep point.
If using .DATA with .DC or .STEP, then any instance parameter, model parameter, or
user-defined parameter in the global scope.
However, if using .DATA with .AC or .NOISE, then one can sweep the magnitude and
phase of an AC source, and linear model parameters (such as resistance and
capacitance) in addition to the traditional AC sweep variable, frequency. Parameters
associated with nonlinear models (like transistors) are not allowed. This is because AC
analysis is a linear analysis, performed after the DCOP calculation. Changing
nonlinear device model parameters would result in a different DCOP solution, so
changing them during the AC (or NOISE) analysis phase is not valid.
Another caveat, for both .AC and .NOISE, is that all of the frequency values in the data
table must be positive. If .DATA is used with .NOISE then the integrals for the total
input noise and total output noise will only be calculated, and sent to stdout, if the
frequencies in the data table are monotonically increasing.

24
2.1.3. .DC (DC Sweep Analysis)

Calculates the operating point for the circuit for a range of values. Primarily, this capability is applied to
independent voltage sources, but it can also be applied to most device parameters. Note that this may be
repeated for multiple sources in the same .DC line.

The .DC command can specify a linear sweep, decade logarithmic sweep, octave logarithmic sweep, a list
of values, or a data table of multivariate values.

2.1.3.1. Linear Sweeps

General Form .DC [LIN] <sweep variable name> <start> <stop> <step>
+ [<sweep variable name> <start> <stop> <step>]*

Examples .DC LIN V1 5 25 5

.DC VIN -10 15 1
.DC R1 0 3.5 0.05 C1 0 3.5 0.5

.param start=5, stop=25, points=5

.DC {start} {stop} {points}

Comments A .PRINT DC must be used to get the results of the DC sweep analysis. See Section
2.1.31.

A .OP comand will result in a linear DC analysis if there is no .DC specified.

If the stop value is smaller than the start value, the step value should be negative. If a
positive step value is given in this case, only a single point (at the start value) will be
performed, and a warning will be emitted.

2.1.3.2. Decade Sweeps

General Form .DC DEC <sweep variable name> <start> <stop> <points>
+ [DEC <sweep variable name> <start> <stop> <points>]*

Examples .DC DEC VIN 1 100 2

.DC DEC R1 100 10000 3 DEC VGS 0.001 1.0 2

.param start=1, stop=100, points=2

.DC DEC VIN {start} {stop} {points}

25
Comments The stop value should be larger than the start value. If a stop value smaller than the
start value is given, only a single point at the start value will be performed, and a
warning will be emitted. The points value must be an integer.

2.1.3.3. Octave Sweeps

General Form .DC OCT <sweep variable name> <start> <stop> <points>
+ [OCT <sweep variable name><start> <stop> <points>]...

Examples .DC OCT VIN 0.125 64 2

.DC OCT R1 0.015625 512 3 OCT C1 512 4096 1

.param start=0.125, stop=64, points=2

.DC OCT VIN {start} {stop} {points}

Comments The stop value should be larger than the start value. If a stop value smaller than the
start value is given, only a single point at the start value will be performed, and a
warning will be emitted. The points value must be an integer.

2.1.3.4. List Sweeps

General Form .DC <sweep variable name> LIST <val> <val> <val>*
+ [ <sweep variable name> LIST <val> <val>* ]*

Examples .DC VIN LIST 1.0 2.0 5.0 6.0 10.0

.DC VDS LIST 0 3.5 0.05 VGS LIST 0 3.5 0.5
.DC TEMP LIST 10.0 15.0 18.0 27.0 33.0

.param val1=0, val2=3.5, val3=0.5

.DC VDS LIST {val1} {val2} {val3}

2.1.3.5. Data Sweeps

General Form .DC DATA=<data table name>

Examples .DC data=resistorValues

.data resistorValues
+ r1 r2
+ 8.0000e+00 4.0000e+00
+ 9.0000e+00 4.0000e+00
.enddata

26
2.1.4. .DCVOLT (Initial Condition, Bias point)

The .DCVOLT sets initial conditions for an operating point calculation. It is identical in function to the .IC
command. See section 2.1.14 for detailed guidance.

27
2.1.5. .EMBEDDEDSAMPLING (Embedded Sampling)

Calculates a full analysis (for .DC or .TRAN only) over a distribution of parameter values. Embedded
sampling operates similarly to .STEP, except that the parameter values are generated from random
distributions rather than sweeps, and that the loop over parameters happens at the inner-most part of the
calculation, so all samples are propagated simultaneously. If used in conjunction with projection-based
PCE methods, then the sample points are not based on random samples. Instead they are based on the
quadrature points.

General Form .EMBEDDEDSAMPLING

+ param=<parameter name>,[parameter name]*
+ type=<parameter type>,[parameter type]*
+ means=<mean>,[mean]*
+ std_deviations=<standard deviation>,[standard deviation]*
+ lower_bounds=<lower bound>,[lower bound]*
+ upper_bounds=<upper bound>,[upper bound]*
+ alpha=<alpha>,[alpha]*
+ beta=<beta>,[beta]*

Examples .EMBEDDEDSAMPLING
+ param=R1
+ type=normal
+ means=3K
+ std_deviations=1K

.EMBEDDEDSAMPLING
+ param=R1,R2
+ type=uniform,uniform
+ lower_bounds=1K,2K
+ upper_bounds=5K,6K

.EMBEDDEDSAMPLING
+ useExpr=true

.options EMBEDDEDSAMPLES numsamples=10000

.options EMBEDDEDSAMPLES numsamples=25000

+ OUTPUTS={R1:R},{V(1)}
+ SAMPLE_TYPE=MC

.options EMBEDDEDSAMPLES numsamples=1000

+ MEASURES=maxSine
+ SAMPLE_TYPE=LHS

.options embeddedsamples numsamples=30

+ covmatrix=1e6,1.0e-3,1.0e-3,4e-14
+ OUTPUTS={V(1)},{R1:R},{C1:C}

28
Arguments and
Options param
Names of the parameters to be sampled. This may be any of the parameters
that are valid for .STEP, including device instance, device model, or global
parameters. If more than one parameter, then specify as a comma-separated
list.
type
Distribution type for each parameter. This may be uniform, normal or
gamma. If more than one parameter, then specify as a comma-separated list.
means
If using normal distributions, the mean for each parameter must be specified.
If more than one parameter, then specify as a comma-separated list.
std_deviations
If using normal distributions, the standard deviation for each parameter must
be specified. If more than one parameter, then specify as a comma-separated
list.
lower_bounds
If using uniform distributions, the lower bound must be specified. This is
optional for normal distributions. If used with normal distributions, may
alter the mean and standard deviation. If more than one parameter, then
specify as a comma-separated list.
upper_bounds
If using uniform distributions, the upper bound must be specified. This is
optional for normal distributions. If used with normal distributions, may
alter the mean and standard deviation. If more than one parameter, then
specify as a comma-separated list.
alpha
If using gamma distributions, the alpha value for each parameter must be
specified. If more than one parameter, then specify as a comma-separated list.
beta
If using gamma distributions, the beta value for each parameter must be
specified. If more than one parameter, then specify as a comma-separated list.
useExpr
If this argument is set to true, then the sampling algorithm will set up
random inputs from expression operators such as AGAUSS and AUNIF. In this
case it will also ignore the list of parameters on the .EMBEDDEDSAMPLING
command line. For a complete description of expression-based random
operators, see the expression documentation in section 2.2.

Comments In addition to the .EMBEDDEDSAMPLING command, this analysis requires a

.options EMBEDDEDSAMPLES command as well. The .EMBEDDEDSAMPLING
command specifies parameters and their attributes, either using the useExpr option,

29
or with comma-separated lists. The .options EMBEDDEDSAMPLES command
specifies analysis options, including the number of samples, the type of sampling
(LHS or MC) and the outputs and/or measures for which to compute statistics. This
line also allows one to specify a non-intrusive Polynomial Chaos Expansion (PCE)
method (either regression or projection PCE). To see the details of the
.options EMBEDDEDSAMPLES command , see table 2-11.
On the .EMBEDDEDSAMPLING command line, if not using useExpr, parameters and
their attributes must be specified using comma-separated lists. The comma-separated
lists must all be the same length.
The .PRINT ES command provides output based on the contents of those print-lines,
and also the NUMSAMPLES and OUTPUT arguments on the .OPTIONS
EMBEDDEDSAMPLES line. If the OUTPUT_SAMPLE_STATS argument on a .PRINT ES
line is set to “true” then the statistics for the MEAN, MEANPLUS, MEANMINUS, STDDEV
and VARIANCE will be output for each variable in the OUTPUT argument. If the
OUTPUT_ALL_SAMPLES argument on a .PRINT ES line is set to “true” then the values
of all NUMSAMPLES samples, for each variable requested in the OUTPUTS argument,
will be output.

30
2.1.6. .END (End of Circuit)

Marks the end of netlist file.

31
2.1.7. .ENDS (End of Subcircuit)

Marks the end of a subcircuit definition.

32
2.1.8. .FFT (FFT Analysis)

Performs Fast Fourier Transform analysis of transient analysis output.

General Form .FFT <ov> [NP=<value>] [WINDOW=<value>] [ALFA=<value>]

+ [FORMAT=<value>] [START=<value>] [STOP=<value>]
+ [FREQ=value] [FMIN=value] [FMAX=value]

Examples .FFT v(1)

.FFT v(1,2) NP=512 WINDOW=HANN
.FFT {v(3)-v(2)} START=1 STOP=2

Arguments and
Options
ov The desired solution output to be analyzed. Only one output variable can be
specified on a .FFT line. However, multiple .FFT lines with the same output
variable but, for example, different windowing functions may be used in a
netlist. The available outputs are:
• V(<circuit node>) the voltage at <circuit node>
• V(<circuit node>,<circuit node>) to output the voltage
difference between the first <circuit node> and second <circuit
node>
• I(<device>) the current through a two terminal device
• I<lead abbreviation>(<device>) the current into a particular lead
of a three or more terminal device (see the Comments, below, for details)
• N(<device parameter>) a specific device parameter (see the
individual devices in Section 2.3 for syntax)

NP The number of points in the FFT. This value must be a power of 2. If the
user-entered value is not a power of two then it will be rounded to the nearest
power of 2. The minimum allowed value of NP is 4. The default value is 1024.
WINDOW
The windowing function that will be applied to the sampled waveform
values. The allowed values are as follows, where table 2-1 gives their exact
definitions:
• RECT (or RECTANGULAR) = rectangular window (default)
• BART (or BARTLETT) = Bartlett (triangular) window
• BARTLETTHANN = Bartlett-Hann window
• BLACK = Blackman window
• BLACKMAN = “Conventional Blackman” window
• HAMM (or HAMMING) = Hamming window
• HANN (or HANNING) = Hanning window

33
 • HARRIS (or BLACKMANHARRIS) = Blackman-Harris window
• NUTTALL = Nuttall window
• COSINE2 = Power-of-cosine window, with exponent 2.
• COSINE4 = Power-of-cosine window, with exponent 4.
• HALFCYCLESINE = Half-cycle sine window
• HALFCYCLESINE3 = Half-cycle sine window, with exponent 3.
• HALFCYCLESINE6 = Half-cycle sine window, with exponent 6.
ALFA
This parameter is supported for HSPICE compatibility. It currently has no
effect though, since the GAUSS and KAISER windows are not supported.
FORMAT
The allowed values are NORM and UNORM. If NORM is selected then the
magnitude values will be normalized to the largest magnitude. If UNORM is
selected then the actual magnitude values will be output instead. The default
value for this parameter depends on the .OPTIONS FFT FFT_MODE setting.
If FFT_MODE=0, which is the default options setting, then the default value is
NORM. If FFT_MODE=1 then the default value is UNORM.
START
The start time for the FFT analysis. The default value is the start time for the
transient analysis. FROM is an allowed synonym for START.
STOP
The end time for the FFT analysis. The default value is the end time for the
transient analysis. TO is an allowed synonym for STOP.
FREQ
The “first harmonic” of the frequencies provided in the output file. The
default value for FREQ is 1/(STOP - START). If FREQ is given then it is
rounded to the nearest integer multiple of the default value. The Xyce Users’
Guide [1] provides an example.
FMIN
This parameter can use to adjust the harmonics included in the “additional
metrics” defined in Section 2.1.8.3. It has a default value of 1.
FMAX
This parameter can use to adjust the harmonics included in the “additional
metrics” defined in Section 2.1.8.3. It has a default value of NP/2.

Comments Multiple .FFT lines may be used in a netlist. All results from FFT analyses will be
returned to the user in a file with the same name as the netlist file suffixed with .fftX
where X is the step number.

<lead abbreviation> is a single character designator for individual leads on a

device with three or more leads. For bipolar transistors these are: c (collector), b
(base), e (emitter), and s (substrate). For mosfets, lead abbreviations are: d (drain), g

34
 (gate), s (source), and b (bulk). SOI transistors have: d, g, s, e (bulk), and b (body). For
PDE devices, the nodes are numbered according to the order they appear, so lead
currents are referenced like I1(<device>), I2(<device>), etc.
For this analysis, the phase data is always output in degrees.
In Xyce, WINDOW=TRIANGULAR is not an allowed synonym for WINDOW=BART. This is
to avoid confusion with other analysis packages such as SciPy and Matlab.

2.1.8.1. .OPTIONS FFT FFT_MODE

The setting for .OPTIONS FFT FFT_MODE is used to control whether the Xyce FFT processing and output
are more compatible with HSPICE (0) or Spectre (1). This setting affects the format of the window
functions, the conversion from two-sided to one-sided results, and whether the default output for the
magnitude values is normalized, or not. The default setting for FFT_MODE is 0.
If FFT_MODE=0 then symmetric window functions are used. If FFT_MODE=1 then periodic window
functions are used. The next subsection provides more details on that difference.
If FFT_MODE=0 then the two-sided to one-sided conversion doubles the magnitudes of the 1,2,...,NP
harmonics. If FFT_MODE=1 then that conversion only doubles the magnitudes of the 1,2,...,(NP-1)
harmonics.
The default value for the FORMAT parameter depends on the .OPTIONS FFT FFT_MODE setting. If
FFT_MODE=0 then the default value for that parameter is NORM. If FFT_MODE=1 then the default value is
UNORM.

2.1.8.2. Window Functions

Table 2-1 gives the definitions of the window functions implemented in Xyce. For HSPICE compatibility,
the BLACK window type is actually the “-67 dB Three-Term Blackman-Harris window” [3] rather than the
“Conventional Blackman Window” used by Spectre, Matlab and SciPy. The Convential Blackman Window
can be selected with the BlACKMAN window type instead. The definition of the BART window type [4] was
chosen to match Spectre, Matlab and SciPy. The Xyce definition may differ from HSPICE.
As mentioned in in the previous subsection, the choice of symmetric vs. periodic window functions can be
selected via the use of .OPTIONS FFT FFT_MODE=<0|1>. If symmetric windows are used then 𝐿 = 𝑁 − 1
in the formulas in table 2-1, where 𝑁 is the number of points in the FFT. If periodic windows are used then
𝐿 = 𝑁.

Table 2-1. .FFT Window Function Definitions (N = Number of Points).

Value Description Definition

Rectangular 𝑤(𝑖) = 1, for 0 ≤ 𝑖 < 𝑁

RECT
 2𝑖
BART Barlett [4] 𝑤(𝑖) = 𝐿 , if 𝑖 < 0.5 · (𝑁 − 1)
2 − 2𝑖𝐿 , otherwise
𝑤(𝑖) =
BARTLETTHANN Bartlett-Hann [3]
0.62−0.48 · | 𝐿𝑖 −0.5| +0.38 · 𝑐𝑜𝑠(2𝜋 · ( 𝐿𝑖 −0.5)), for 0 ≤ 𝑖 < 𝑁
HANN Hanning [4] 𝑤(𝑖) = 𝑠𝑖𝑛2 ( 𝜋𝑖
𝐿 ), for 0 ≤ 𝑖 < 𝑁

35
 Value Description Definition
HAMM Hamming [4] 𝑤(𝑖) = 0.54 − 0.46 · 𝑐𝑜𝑠( 2𝐿𝜋𝑖 ), for 0 ≤ 𝑖 < 𝑁
“Conventional
BLACKMAN 𝑤(𝑖) = 0.42 − 0.5 · 𝑐𝑜𝑠( 2𝐿𝜋𝑖 ) + 0.08 · 𝑐𝑜𝑠( 4𝐿𝜋𝑖 ), for 0 ≤ 𝑖 < 𝑁
Blackman window” [3]
-67 dB Three-Term
𝑤(𝑖) = 0.42323 − 0.49755 · 𝑐𝑜𝑠( 2𝐿𝜋𝑖 ) + 0.07922 ·
BLACK Blackman-Harris
𝑐𝑜𝑠( 4𝐿𝜋𝑖 ), for 0 ≤ 𝑖 < 𝑁
window [3]
-92 dB Four-Term
𝑤(𝑖) = 0.35875 − 0.48829 · 𝑐𝑜𝑠( 2𝐿𝜋𝑖 ) + 0.14128 · 𝑐𝑜𝑠( 4𝐿𝜋𝑖 ) −
HARRIS Blackman-Harris
0.01168 · 𝑐𝑜𝑠( 6𝐿𝜋𝑖 ), for 0 ≤ 𝑖 < 𝑁
window [3]
Four-Term Nuttall,
𝑤(𝑖) = 0.3635819 − 0.4891775 · 𝑐𝑜𝑠( 2𝐿𝜋𝑖 ) + 0.1365995 ·
NUTTALL Minimum Sidelobe
𝑐𝑜𝑠( 4𝐿𝜋𝑖 ) − 0.0106411 · 𝑐𝑜𝑠( 6𝐿𝜋𝑖 ), for 0 ≤ 𝑖 < 𝑁
(Blackman-Nuttall) [3]
Power-of-cosine
COSINE2 window, with exponent 𝑤(𝑖) = 0.5 − 0.5 · 𝑐𝑜𝑠( 2𝐿𝜋𝑖 ), for 0 ≤ 𝑖 < 𝑁
2
Power-of-cosine
COSINE2 window, with exponent 𝑤(𝑖) = 0.375 − 0.5 · 𝑐𝑜𝑠( 2𝐿𝜋𝑖 ) + 0.125 · 𝑐𝑜𝑠( 4𝐿𝜋𝑖 ), for 0 ≤ 𝑖 < 𝑁
4
HALFCYCLESINE Half-cycle sine window 𝑤(𝑖) = 𝑠𝑖𝑛( 𝜋𝑖
𝐿 )
Half-cycle sine window,
HALFCYCLESINE3 𝑤(𝑖) = 𝑠𝑖𝑛3 ( 𝜋𝑖
𝐿 )
with exponent 3
Half-cycle sine window,
HALFCYCLESINE6 𝑤(𝑖) = 𝑠𝑖𝑛6 ( 𝜋𝑖
𝐿 )
with exponent 6

2.1.8.3. Additional FFT Metrics

The following additional metrics will be sent to stdout if .OPTIONS FFT FFTOUT=1 is used in the netlist.
These definitions are also used by the corresponding measure types on .MEASURE FFT lines.
Define the integer index of the “first harmonic” ( 𝑓0 ) as follows (where it has a default value of 1 and the
𝑅𝑂𝑈𝑁 𝐷 () function rounds to the nearest integer):
 𝐹 𝑅𝐸𝑄
𝑅𝑂𝑈𝑁 𝐷 ( 𝑆𝑇𝑂𝑃−𝑆𝑇 𝐴𝑅𝑇 ), if FREQ given
𝑓0 =
1, otherwise
Finally, define 𝑚𝑎𝑔[𝑖] as the magnitude of the FFT coefficient at index 𝑖, and 𝑁 as the number of points in
the FFT.
The Signal to Noise-plus-Distortion Ratio (SNDR) is calculated as follows, where the summation in the
denominator includes all of the frequencies except for the first harmonic frequency 𝑓0 :
𝑚𝑎𝑔[ 𝑓0 ]
𝑆𝑁 𝐷 𝑅 = 20 · 𝑙𝑜𝑔10( 𝑠𝑞𝑟𝑡 ( Í 𝑚𝑎𝑔[𝑖 ]∗𝑚𝑎𝑔[𝑖 ] ) ), for 1 ≤ 𝑖 ≤ 0.5 · 𝑁 , and 𝑖 ≠ 𝑓0

The Effective Number of Bits (ENOB) is calculated as follows, where the units is “bits”:
𝑆 𝑁 𝐷𝑅−1.76
𝐸 𝑁𝑂𝐵 = 6.02

For the Signal to Noise Ratio (SNR) metric define the upper frequency limit for the SNR metric as follows,
with the caveat that if 𝑓2 is less than 𝑓0 then its value is set to 𝑓0 :
 𝐹 𝑀 𝐴𝑋
𝑅𝑂𝑈𝑁 𝐷 ( 𝑆𝑇𝑂𝑃−𝑆𝑇 𝐴𝑅𝑇 ), if FMAX given
𝑓2 =
0.5 · 𝑁, otherwise

36
The SNR is then calculated as follows, where the summation in the denominator only uses indexes that are
either not an integer-multiple of 𝑓0 or that are greater than the upper frequency limit of 𝑓2 . (Note: for the
default case of 𝑓0 = 1 and FMAX not given there are no “noise frequencies”.)
𝑚𝑎𝑔[ 𝑓0 ]
𝑆𝑁 𝑅 = 20 · 𝑙𝑜𝑔10( 𝑠𝑞𝑟𝑡 ( Í 𝑚𝑎𝑔[𝑖 ]∗𝑚𝑎𝑔[𝑖 ] ) ), for 𝑖 > 𝑓2 or 𝑖% 𝑓0 ≠ 0

For the Total Harmonic Distortion (THD) metric define the upper frequency limit ( 𝑓2 ) in the summation in
the numerator as:
 𝐹 𝑀 𝐴𝑋
𝑅𝑂𝑈𝑁 𝐷 ( 𝑆𝑇𝑂𝑃−𝑆𝑇 𝐴𝑅𝑇 ), if FMAX given
𝑓2 =
0.5 · 𝑁, otherwise
The THD is then calculated using only the indexes that are multiples of 𝑓0 . So, if 𝑓0 =2 then the summation
in the numerator would only include indexes 4,6,8,etc.
Í
𝑇 𝐻𝐷 = 20 · 𝑙𝑜𝑔10( 𝑠𝑞𝑟𝑡 ( 𝑚𝑎𝑔[𝑖 ]∗𝑚𝑎𝑔[𝑖 ] )
𝑚𝑎𝑔[ 𝑓0 ] ), for 2 · 𝑓0 ≤ 𝑖 ≤ 𝑓2 , and 𝑖% 𝑓0 = 0
For Spurious Free Distortion Ratio (SFDR) metric define the upper and lower frequency limits ( 𝑓2 and 𝑓1 )
considered in the denominator of the calculation as:
 𝐹 𝑀 𝐴𝑋
𝑅𝑂𝑈𝑁 𝐷 ( 𝑆𝑇𝑂𝑃−𝑆𝑇 𝐴𝑅𝑇 ), if FMAX given
𝑓2 =
0.5 · 𝑁, otherwise
and:
𝐹𝑀𝐼𝑁
 𝑅𝑂𝑈𝑁 𝐷 ( 𝑆𝑇𝑂𝑃−𝑆𝑇 𝐴𝑅𝑇 ), if FMIN given



𝑓1 = 𝑓0 , if FMIN not given, and 𝑓2 ≥ 𝑓0
 1,

if FMIN not given, and 𝑓2 < 𝑓0

The SFDR is then calculated as:
𝑚𝑎𝑔[ 𝑓0 ]
𝑆𝐹 𝐷 𝑅 = 20 · 𝑙𝑜𝑔10( 𝑀 𝐴𝑋 (𝑚𝑎𝑔[𝑖 ] ) ), for 𝑓1 ≤ 𝑖 ≤ 𝑓2 , and 𝑖 ≠ 𝑓0

2.1.8.4. Re-Measure

Xyce can re-calculate (or re-measure) the values for .FFT statements using existing Xyce output files.
Section 2.1.18.4 discusses this topic in more detail for both .MEASURE and .FFT statements. One additional
caveat is that FFT_ACCURATE is set to 0 during the re-measure operation. This should have no impact on the
accuracy of the re-measured results if the output file was previously generated with FFT_ACCURATE set to
1.

2.1.8.5. Compatibility Between OUTPUT and FFT Options

Some .OPTIONS OUTPUT settings are incompatible with using the default setting of .OPTIONS FFT
FFT_ACCURATE=1. The use of .OPTIONS OUTPUT OUTPUTTIMEPOINTS is compatible. However, the use
of .OPTIONS OUTPUT INITIAL_INTERVAL is not. In that latter case, .OPTIONS FFT FFT_ACCURATE
will be automatically set to 0.

37
2.1.9. .FOUR (Fourier Analysis)

Performs Fourier analysis of transient analysis output.

General Form .FOUR <freq> <ov> [ov]*

Examples .FOUR 100K v(5)

.FOUR 1MEG v(5,3) v(3)
.FOUR 20MEG SENS
.FOUR 40MEG {v(3)-v(2)}

Arguments and
Options freq
The fundamental frequency used for Fourier analysis. Fourier analysis is
performed over the last period (1/freq) of the transient simulation. The DC
component and the first nine harmonics are calculated.

ov The desired solution output, or outputs, to be analyzed. Fourier analysis can

be performed on several outputs for each fundamental frequency, freq. At
least one output must be specified in the .FOUR line. The available outputs
are:
• V(<circuit node>) the voltage at <circuit node>
• V(<circuit node>,<circuit node>) to output the voltage
difference between the first <circuit node> and second <circuit
node>
• I(<device>) the current through a two terminal device
• I<lead abbreviation>(<device>) the current into a particular lead
of a three or more terminal device (see the Comments, below, for details)
• N(<device parameter>) a specific device parameter (see the
individual devices in Section 2.3 for syntax)
• SENS transient direct sensitivities (see Section 2.1.35 for more details
about setting up the .SENS command)

Comments Multiple .FOUR lines may be used in a netlist.

All results from Fourier analysis will be returned to the user in a file with the same
name as the netlist file suffixed with .four#, where the suffixed number (#) starts at 0
and increases for multiple iterations (.STEP iterations) of a given simulation.
<lead abbreviation> is a single character designator for individual leads on a
device with three or more leads. For bipolar transistors these are: c (collector), b
(base), e (emitter), and s (substrate). For mosfets, lead abbreviations are: d (drain), g
(gate), s (source), and b (bulk). SOI transistors have: d, g, s, e (bulk), and b (body). For
PDE devices, the nodes are numbered according to the order they appear, so lead
currents are referenced like I1(<device>), I2(<device>), etc.

38
For this analysis, the phase data is always output in degrees.

39
2.1.10. .FUNC (Function)

User defined functions that can be used in expressions appearing later in the same scope as the .FUNC
statement.

General Form .FUNC <name>([arg]*) {<body>}

Examples .FUNC E(x) {exp(x)}

.FUNC DECAY(CNST) {E(-CNST*TIME)}
.FUNC TRIWAV(x) {ACOS(COS(x))/3.14159}
.FUNC MIN3(A,B,C) {MIN(A,MIN(B,C))}

Arguments and
Options name
Function name. Functions cannot be redefined and the function name must
not be the same as any of the predefined functions (e.g., SIN and SQRT).

arg The arguments to the function. .FUNC arguments cannot be node names. The
number of arguments in the use of a function must agree with the number in
the definition. Parameters, TIME, FREQ, and other functions are allowed in
the body of function definitions. Two constants EXP and PI cannot be used a
argument names. These constants are equal to 𝑒 and 𝜋, respectively, and
cannot be redefined.
body
May refer to other (previously defined) functions; the second example,
DECAY, uses the first example, E.

Comments The <body> of a defined function is handled in the same way as any math expression;
it must be enclosed in curly braces .

The scoping rules for functions are:

• If a .FUNC, statement is included in the main circuit netlist, then it is accessible

from the main circuit and all subcircuits.

• .FUNC statements defined within a subcircuit are scoped to that subciruit

definition. So, their functions are only accessible within that subcircuit
definition, as well as within “nested subcircuits” also defined within that
subcircuit definition.

Additional illustative examples of scoping are given in the “Working with Subcircuits
and Models” section of the Xyce Users’ Guide [1] .

Rules for function names are as follows:

• They should start with a letter or the underscore (_) character, for maximal
compatibilty with other Spice-like simulators. The hash (#) at (@) and backtick
(‘) symbols also work, but they are not reserved characters.

40
• These arithmetic operators % ^ & ~ * - + < > / | should not be used anywhere in
function names, as they cause problems with expression parsing.
• Parentheses (“(” or “)”), braces (“{” or “}”), commas, semi-colons, double
quotes and single quotes are also not allowed.

41
2.1.11. .GLOBAL (Global Node)

The .GLOBAL command provides another way to designate certain nodes as global nodes, besides starting
their node name with the two characters “$G” as discussed in section 2.3.1. A typical usage of such global
nodes is to define a VDD or VSS signal that all subcircuits need to be able to access, but without having to
provide VSS and VDD input nodes to every subcircuit.

General Form .GLOBAL <node1> [ node2 node3 ... ]

Examples .GLOBAL g1
.subckt rsub a b
Rab a b 2
* since node G1 is global, it may be used here without
* being listed on the .subckt line
Rbg G1 b 3
.ends

Comments The name of the global node can be any legal node name, per section 2.3.2.

42
2.1.12. .GLOBAL_PARAM (Global parameter)

User-defined global parameter that can be used in expressions throughout the netlist.
General Form .GLOBAL_PARAM [<name>=<value>]*

Examples .GLOBAL_PARAM T={27+100*time}

name
Global parameter name. Global parameters may be redefined.
If the same name is used on multiple parameters, Xyce by
default will use the last parameter of that name. By
default, no warning will be emitted. To change this
behavior, one can use the -redefined_param command line
option, described in section 3-1.
value
The value may be a number or an expression.

Comments A .PARAM defined in the top level netlist is equivalent to a .GLOBAL_PARAM, and they
can be combined as needed. Thus, you may use parameters defined by .PARAM in
expressions used to define global parameters, and you may also use global parameters
in .PARAM definitions. However, a .GLOBAL_PARAM can only depend on .PARAM
parameter from the top level circuit scope.
Like .PARAM parameters, .GLOBAL_PARAM may depend on time dependent quantities
in the circuit. They may also be frequency dependent. They cannot, however, be
dependent on solution variaables such as voltage nodes.
To load an external data file with time voltage pairs of data on each line into a global
parameter, use this syntax:
.GLOBAL_PARAM extdata = {tablefile("filename")}
or
.GLOBAL_PARAM extdata = {table("filename")}
where filename would be the name of the file to load. Other interpolators that can
read in a data table from a file include fasttable,spline, akima, cubic, wodicka
and bli. See 2.2 for further information.
There are several reserved words that may not be used as names for parameters. These
reserved words are:
• Time
• Freq
• Hertz
• Vt

43
 • Temp
• Temper
• GMIN
Global parameters are accessible, and have the same value, throughout all levels of the
netlist hierarchy. It is not legal to redefine global parameters in different levels of the
netlist hierarchy. Also, global parameters can only be defined in the top level circuit
scope. Parameters defined inside of subcircuits must be of the .PARAM type.

44
2.1.13. .HB (Harmonic Balance Analysis)

Calculates steady states of nonlinear circuits in the frequency domain.

General Form .HB <fundamental frequencies>

Examples .HB 1e4

.hb 1e4 2e2

Arguments and
Options fundamental frequencies
Sets the fundamental frequencies for the analysis.

Comments Harmonic balance analysis calculates the magnitude and phase of voltages and
currents in a nonlinear circuit. Use a .OPTIONS HBINT statement to set additional
harmonic balance analysis options.
The .PRINT HB statement must be used to get the results of the harmonic balance
analysis. See section 2.1.31.
Some devices that may be expected to work in HB analysis do not at this time. This
includes some use cases of B sources (but not all). A time-dependent B source will not
work with HB. However, a B source that is purely dependent (such as a nonlinear
resistor) will work. This same guidance applies to the E,F,G, and H dependent sources.

45
2.1.14. .IC (Initial Condition, Bias point)

The .IC/.DCVOLT command sets initial conditions for operating point calculations. These operating point
conditions will be enforced the entire way through the nonlinear solve. Initial conditions can be given for
some or all of the circuit nodes.
As the conditions are enforced for the entire solve, only the nodes not specified with .IC statements will
change over the course of the operating point calculation.
Note that it is possible to specify conditions that are not solvable. Consult the Xyce Users’ Guide [1] for
more guidance.

General Form .IC V(<node>)=<value>

.IC <node> <value>
.DCVOLT V(<node>)=<value>
.DCVOLT <node> <value>

Examples .IC V(2)=3.1

.IC 2 3.1
.DCVOLT V(2)=3.1
.DCVOLT 2 3.1

Comments The .IC capability can only set voltage values, not current values.
The .IC capability can not be used within subcircuits to set voltage values on global
nodes.

46
2.1.15. .INC or .INCLUDE or .INCL (Include file)

Include specified file in netlist.

The file name can be surrounded by single or double quotes, ’filename’ or "filename", but this is not
necessary. The directory for the include file is assumed to be the execution directory unless a full or relative
path is given as a part of the file name.

General Form .INC <include file name>

.INCLUDE <include file name>
.INCL <include file name>

Examples .INC models.lib

.INC ’models.lib’
.INC "models.lib"
.INCLUDE models.lib
.INCLUDE ’models.lib’
.INCLUDE "path_to_library/models.lib"

Comments If <include file name> uses an absolute path then that path is used. Otherwise, the
search-path order for <include file name> is:
• Relative to the directory that contains <include file name>.
• Relative to the directory that contains the file with the top-level netlist.
• Relative to the Xyce execution directory.

47
2.1.16. .LIB (Library file)

The .LIB command is similar to .INCLUDE, in that it brings in an external file. However, it is designed to
only bring in specific parts of a library file, as designated by an entry name. Note that the Xyce version of
.LIB has been designed to be compatible with HSPICE [5], not PSpice [6].

There are two forms of the .LIB statement, the call and the definition. The call statement reads in a
specified subset of a library file, and the definition statement defines the subsets.

2.1.16.1. .LIB call statement

General Form .LIB <file name> <entry name>

Examples .LIB models.lib nom

.LIB ’models.lib’ low
.LIB "models.lib" low
.LIB "path/models.lib" high

Arguments and
Options file name
Name of file containing netlist data. Single or double quotes (" or ’) may be
used around the file name.
entry name
Entry name, which determines the section of the file to be included. These
sections are defined in the included file using the definition form of the .LIB
statement.

The library file name can be surrounded by quotes (single or double), as in "path/filename" but this is not
necessary. The directory for the library file is assumed to be the execution directory unless a full or relative
path is given as a part of the file name. The section name denotes the section or sections of the library file to
include.

If <file name> uses an absolute path then that path is used. Otherwise, the search-path order for <file name>
is:

• Relative to the directory that contains <file name>.

• Relative to the directory that contains the file with the top-level netlist.

• Relative to the Xyce execution directory.

2.1.16.2. .LIB definition statement

The format given above is when the .LIB command is used to reference a library file; however, it is also
used as part of the syntax in a library file.

48
General Form .LIB <entry name>
<netlist lines>*
.endl <entry name>

Examples
* Library file res.lib
.lib low
.param rval=2
r3 2 0 9
.endl low

.lib nom
.param rval=3
r3 2 0 8
.endl nom

Arguments and
Options entry name
The name to be used to identify this library component. When used on a
.LIB call line, these segments of the library file will be included in the
calling file.

Note that for each entry name, there is a matched .lib and .endl. Any valid netlist commands can be
placed inside the .lib and .endl statements. The following is an example calling netlist, which refers to
the library in the examples above:
* Netlist file res.cir
V1 1 0 1
R 1 2 {rval}
.lib res.lib nom
.tran 1 ps 1ns
.end

In this example, only the netlist commands that are inside of the “nom” library will be parsed, while the
commands inside of the “low” library will be discarded. As a result, the value for resistor r3 is 8, and the
value for rval is 3.

49
2.1.17. .LIN (Linear Analysis)

Extracts linear transfer parameters (S-, Y- and Z-parameters) for a general multiport network. Those
parameters can be output in either Touchstone format [7].

General Form .LIN [SPARCALC=<1|0>] [FORMAT=<TOUCHSTONE2|TOUCHSTONE>]

+ [LINTYPE=<S|Y|Z>] [DATAFORMAT=<RI|MA|DB>]
+ [FILE=<output filename>] [WIDTH=<print field width>]
+ [PRECISION=<floating point output precision>]

Examples .LIN
.LIN FORMAT=TOUCHSTONE DATAFORMAT=MA FILE=foo

Arguments and
Options SPARCALC=<1|0>
If this is set to 1 then the LIN analysis is done at the frequency values
specified on the .AC line. The default value is 1.
FORMAT=<TOUCHSTONE2|TOUCHSTONE>
Output file format
TOUCHSTONE Output file is in Touchstone 1 format
TOUCHSTONE2 Output file is in Touchstone 2 format. The default is
TOUCHSTONE2.
LINTYPE=<S|Y|Z>
The type of parameter data (S, Y or Z) in the output file. The default is S.
DATAFORMAT=<RI|MA|DB>
Format for the S-, Y- or Z-parameter data
RI Real-imaginary format
The data is output as the real and imaginary parts for each extracted S-,
Y- or Z-parameter. This is the default.
MA Magnitude-angle format
The data is output as the magnitude and the phase angle of each
extracted S-, Y- or Z-parameter. For compatibility with Touchstone
formats, the angle values are in degrees.
DB Magnitude(dB)-angle format
The data is output as the magnitude (in dB) and the phase angle of each
extracted S-, Y- or Z-parameter. For compatibility with Touchstone
formats, the angle values are in degrees.
FILE=<output filename>
Specifies the name of the file to which the output will be written. For
HSPICE compatibility FILENAME= is an allowed synonym for FILE= on
.LIN lines.
WIDTH=<print field width>
Controls the output width used in formatting the output.

50
 PRECISION=<floating point precision>
Number of floating point digits past the decimal for output data.

Comments The .LIN command line functions like a .PRINT line for the extracted S-, Y- or
Z-parameter data. So, a netlist can have multiple .LIN lines with different values for
the LINTYPE, DATAFORMAT and FILE arguments on each line. If there are multiple
.LIN lines in the netlist, then a linear analysis will be performed if SPARCALC=1 on
any of those .LIN lines.
The default filename for both Touchstone formats is <netlistName>.sNp where N is
the number of “ports” (P devices) specified in the netlist.
The Xyce Touchstone output is based on the Touchstone standard [7]. So, it differs
slightly from the corresponding HSPICE output. In particular, the full matrix of S-, Y-
or Z-parameters is always output.
The HSPICE SPARDIGIT and FREQDIGIT arguments are not supported. Instead, the
PRECISION argument is used for all of the output values.
The output of individual S-parameters via the .PRINT AC line is supported.
If the -r <raw-file-name> and -a command line options are used with .LIN with
SPARCALC=1 then Xyce will exit with a parsing error.
The -o command line option can be used with .LIN. In that case, the output defaults
to Touchstone 2 format and any FILE=<filename> argument on the .LIN line is
ignored.

51
2.1.18. .MEASURE or .MEAS (Measure output)

The .MEASURE statement allows calculation or reporting of simulation metrics to an external file, as well as
to the standard output and/or a log file. One can measure when simulated signals reach designated values,
or when they are equal to other simulation values. The .MEASURE statement is supported for .TRAN, .DC,
.AC and .NOISE analyses. It can be used with .STEP in all four cases. For HSPICE compatibility, .MEAS is
an allowed synonym for .MEASURE.
The syntaxes for the .MEASURE statements are shown below. The AVG, DERIV, EQN, ERR, ERR1, ERR2,
FIND-AT, FIND-WHEN, INTEG, MIN, MAX, PP, RMS, WHEN and TRIG-TARG measures are supported for all
four “measure modes” (TRAN, AC, DC and NOISE). Note that each measure type (e.g., MAX) may be listed
twice. This is because only a subset of the allowed “qualifiers” (e.g., FROM and TO) may be supported for the
AC, DC and NOISE measure modes.
The ERROR measure is Xyce-specific, and is supported for TRAN, AC, DC and NOISE measure modes. The
DUTY, FREQ, FOUR, OFF_TIME and ON_TIME measures are also Xyce-specific, and are only supported for
TRAN measure mode.

General Form .MEASURE TRAN <result name> AVG <variable>

+ [MIN_THRESH=<value>] [MAX_THRESH=<value>]
+ [FROM=<time>] [TO=<time>] [TD=<time>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> DERIV <variable> AT=<value>

+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> DERIV <variable>

+ WHEN <variable>=<variable2 >|<value>
+ [MINVAL=<value>] [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> DUTY <variable>

+ [ON=<value>] [OFF=<value>] [MINVAL=<value>]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> EQN <variable>

+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> <ERR|ERR1|ERR2>

+ <variable1 > <variable2 > [FROM=<value>] [TO=<value>]
+ [MINVAL=<value>] [IGNOR|YMIN=<value>] [YMAX=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> ERROR <variable> FILE=<value>

+ INDEPVARCOL=<value> DEPVARCOL=<value> [COMP_FUNCTION=<value>]

52
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> FIND <variable> AT=<value>

+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> FIND <variable>

+ WHEN <variable>=<variable2 >|<value>
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> FOUR <variable> AT=freq

+ [NUMFREQ=<value>] [GRIDSIZE=<value>]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> FREQ <variable>

+ [ON=<value>] [OFF=<value>] [MINVAL=<value>]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> INTEG <variable>

+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> MAX <variable>

+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST] [RFC_LEVEL=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>]
+ [PRINT=<value>] [OUTPUT=<value>]

.MEASURE TRAN <result name> MIN <variable>

+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST] [RFC_LEVEL=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>]
+ [PRINT=<value>] [OUTPUT=<value>]

.MEASURE TRAN <result name> OFF_TIME <variable>

+ [OFF=<value>] [MINVAL=<value>]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> ON_TIME <variable>

+ [ON=<value>] [MINVAL=<value>]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

53
.MEASURE TRAN <result name> PP <variable>
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST] [RFC_LEVEL=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> RMS <variable>

+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> WHEN <variable>=<variable2 >|<value>

+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE|TRAN> <result name>

+ TRIG <variable1 >=<variable2 >|<value>
+ [TD=<val>] [RISE=r] [FALL=f] [CROSS=c]
+ TARG <variable3 >=<variable4 >|<value>
+ [TD=<val>] [RISE=r] [FALL=f] [CROSS=c]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE|TRAN> <result name>

+ TRIG AT=<value> TARG AT=<value>
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name>

+ TRIG <variable1 > FRAC_MAX=<value>
+ [RISE=r] [FALL=f] [CROSS=c]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ TARG <variable2 > FRAC_MAX=<value>
+ [RISE=r] [FALL=f] [CROSS=c]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> AVG <variable>

+ [FROM=<value>] [TO=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> DERIV <variable> AT=<value>

+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> DERIV <variable>

54
+ WHEN <variable>=<variable2 >|<value>
+ [MINVAL=<value>] [FROM=<value>] [TO=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> EQN <variable>

+ [FROM=<value>] [TO=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> <ERR|ERR1|ERR2>

+ <variable1 > <variable2 > [FROM=<value>] [TO=<value>]
+ [MINVAL=<value>] [IGNOR|YMIN=<value>] [YMAX=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> ERROR <variable>

+ FILE=<value> [DEPVARCOL=<value>] [COMP_FUNCTION=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> FIND <variable> AT=<value>

+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> FIND <variable>

+ WHEN <variable>=<variable2 >|<value>
+ [FROM=<value>] [TO=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> INTEG <variable>

+ [FROM=<value>] [TO=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> MAX <variable>

+ [FROM=<value>] [TO=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>]
+ [PRINT=<value>] [OUTPUT=<value>]

.MEASURE <AC|DC|NOISE> <result name> MIN <variable>

+ [FROM=<value>] [TO=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>]
+ [PRINT=<value>] [OUTPUT=<value>]

.MEASURE <AC|DC|NOISE> <result name> PP <variable>

+ [FROM=<value>] [TO=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name> RMS <variable>

55
 + [FROM=<value>] [TO=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC|DC|NOISE> <result name>

+ WHEN <variable>=<variable2 >|<value>
+ [FROM=<value>] [TO=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

Examples .MEASURE TRAN hit1_75 WHEN V(1)=0.75 MINVAL=0.02

.MEASURE TRAN hit2_75 WHEN V(1)=0.75 MINVAL=0.08 RISE=2
.MEASURE TRAN avgAll AVG V(1)
.MEASURE TRAN dutyAll DUTY V(1) ON=0.75 OFF=0.25
.MEASURE DC maxV1 MAX V(1)
.MEAS DC minV2 MIN V(2)
.MEASURE AC maxV1R MAX VR(1)
.MEASURE NOISE maxonoise MAX ONOISE

Arguments and
Options result name
Measured results are reported to the output and log file. Additionally, for
TRAN measures, the results are stored in files called circuitFileName.mt#,
where the suffixed number (#) starts at 0 and increases for multiple iterations
(.STEP iterations) of a given simulation. Each line of this file will contain
the measurement name, <result name>, followed by its value for that run.
The <result name> must be a legal Xyce character string. For DC
measures, the results are stored in the files circuitFileName.ms#, while
AC and NOISE measures use the files circuitFileName.ma#.
If multiple measures are defined with the same <result name> then Xyce
uses the last such definition, and issues warning messages about (and
discards) any previous measure definitions with the same <result name>.
measure type
AVG, DERIV, DUTY, EQN, ERR, ERR1, ERR2, ERROR, FIND, FREQ,
FOUR, INTEG, MAX, MIN, OFF_TIME, ON_TIME, PP, RMS, WHEN,
TRIG, TARG
The third argument specifies the type of measurement or calculation to be
done. The only exception is the TARG clause which comes later in the
argument list, after the TRIG clause has been specified.
By default, the measurement is performed over the entire simulation. The
calculations can be limited to a specific measurement window by using the
qualifiers FROM, TO, TD, RISE, FALL, CROSS and MINVAL, which are
explained below.
The supported measure types and their definitions are:

56
AVG Computes the arithmetic mean of <variable> for the simulation, or
within the extent of the measurement window. The measurement
window can be limited with the qualifiers FROM, TO and TD for TRAN
measures, and with FROM and TO for AC, DC and NOISE measures.
DERIV Computes the derivative of <variable> at a user-specified time (by
using the AT qualifier) or when a user-specified condition occurs (by
using the WHEN qualifier). If the WHEN qualifier is used then the
measurement window can be limited with the qualifiers FROM, TO, RISE,
FALL and CROSS for all measure modes. In addition, the TD qualifier is
supported for TRAN measures. The MINVAL qualifier is used as a
comparison tolerance for both AT and WHEN. For HSPICE compatibility,
DERIVATIVE is an allowed synonym for DERIV.
DUTY Fraction of time that <variable> is greater than ON and does not fall
below OFF either for the entire simulation, or the measurement window.
The qualifier MINVAL is used as a tolerance on the ON and OFF values,
so that the thresholds become (ON − MINVAL) and (OFF − MINVAL). The
measurement window can be limited with the qualifiers FROM, TO, and
TD for TRAN measures.
EQN Calculates the value of <variable> during the simulation. The
measurement window can be limited with the qualifiers FROM, TO and
TD for TRAN measures, and with FROM and TO for AC, DC and NOISE
measures. As noted in the “Additional Examples” subsection, the
variable can use the results of other measure statements.
ERRx Calculates the error between two simulation variables, where the ERR1
and ERR2 functions (and the use of the MINVAL, YMIN and YMAX
qualifiers in those functions) are defined further in the “Error Functions
(ERR1 and ERR2)” subsection. The ERR measure type is a synonym for
the ERR1 measure type. The measurement window can be limited with
the qualifiers FROM and TO.
ERROR Calculates the norm between the measured waveform and a
“comparison waveform” specified in a file. The supported norms are L1,
L2 and INFNORM. The default norm is the L2 norm.
FIND-AT Returns the value of <variable> at the time when the AT clause
is satisfied. The AT clause is described in more detail later in this list.
FIND-WHEN Returns the value of <variable> at the time when the WHEN
clause is satisfied. The WHEN clause is described in more detail later in
this list.
FOUR Calculates the fourier transform of the transient waveform for
<variable>, given the fundamental frequency AT. All frequencies
output by the measure will be multiples of that fundamental frequency,
and will always start at that fundamental frequency. The values of the
DC component and the first NUMFREQ-1 harmonics are determined
using an interpolation of GRIDSIZE points. The default values for
NUMFREQ and GRIDSIZE are 10 and 200, respectively. The measurement
window can be limited with the qualifiers FROM, TO and TD for TRAN

57
 measures. For this measure, the phase data is always output in degrees.
FREQ An estimate of the frequency of <variable>, found by cycle
counting during the simulation. Cycles are defined through the values of
ON and OFF with MINVAL being used as a tolerance so that the thresholds
becomes (ON − MINVAL) and (OFF + MINVAL). The measurement
window can be limited with the qualifiers FROM, TO and TD for TRAN
measures.
INTEG Calculates the integral of outVal through second order numerical
integration. The integration window can be limited with the qualifiers
FROM, TO and TD for TRAN measures, and with FROM and TO for AC, DC
and NOISE measures. For HSPICE compatibility, INTEGRAL is an
allowed synonym for INTEG.
MAX Returns the maximum value of <variable> during the simulation. The
measurement window can be limited with the qualifiers FROM, TO, TD,
RISE, FALL and CROSS for TRAN measures, and with FROM and TO for
AC, DC and NOISE measures.
MIN Returns the minimum value of <variable> during the simulation. The
measurement window can be limited with the qualifiers FROM, TO, TD,
RISE, FALL and CROSS for TRAN measures, and with FROM and TO for
AC, DC and NOISE measures.
OFF_TIME Returns the time that <variable> is below OFF during the
simulation or measurement window, normalized by the number of
cycles of the waveform during the simulation or measurement window.
OFF uses MINVAL as a tolerance, and the threshold becomes (OFF +
MINVAL). The measurement window can be limited with the qualifiers
FROM, TO and TD for TRAN measures.
ON_TIME Returns the time that <variable> is above ON during the
simulation or measurement window, normalized by the number of
cycles of the waveform during the simulation or measurement window.
ON uses MINVAL as a tolerance, and the threshold becomes (ON −
MINVAL). The measurement window can be limited with the qualifiers
FROM, TO and TD for TRAN measures.
PP Returns the difference between the maximum value and the minimum
value <variable> during the simulation. The measurement window
can be limited with the qualifiers FROM, TO, TD, RISE, FALL and CROSS
for TRAN measures, and with FROM and TO for AC, DC and NOISE
measures.
RMS Computes the root-mean-squared value of <variable> during the
simulation, which is defined as ”the square root of the area under the
<variable> curve, divided by the period of interest”. The measurement
window can be limited with the qualifiers FROM, TO and TD for TRAN
measures, and with FROM and TO for AC, DC and NOISE measures.
TRIG
TARG Measures the time between a trigger event and a target event. The
trigger is specified with TRIG <variable1 >=<variable2 > or TRIG

58
 <variable1 >=<value> or TRIG AT=<value>. The target is then
specified as TARG <variable3 >=<variable4 > or TARG
<variable3 >=<value>. or TARG AT=<value>. It is also possible to
use this measure to find a rise time for variable when the rise time is
defined as the time to go from some small fraction of the maxima to
some other fraction of the maxima. For example, the syntax for finding a
rise time from 10% to 90% of the maxima is:
TRIG V(node) FRAC_MAX=0.1 TARG V(node) FRAC_MAX=0.9
WHEN Returns the time (or frequency or DC sweep value) when
<variable> reaches <variable2 > or the constant value, value. The
measurement window can be limited with the qualifiers FROM, TO, RISE,
FALL and CROSS for all measure modes. In addition, the TD qualifier is
supported for TRAN measures. The qualifier MINVAL acts as a tolerance
for the comparison. For example when <variable2 > is specified, the
comparison used is when <variable> = <variable2 > ± MINVAL or
when a constant, value is given: <variable> = value ± MINVAL. If
the conditions specified for finding a given value were not found during
the simulation then the measure will return the default value of 0. The
user may change this default value with the DEFAULT_VAL qualifier
described below. Note: The use of FIND and WHEN in one measure
statement is also supported.

variable
variable𝑛
value
These quantities represents the test for the stated measurement. <variable>
is a simulation quantity, such as a voltage or current. One can compare it to
another simulation variable or a fixed quantity. Additionally, the
<variable> may be a Xyce expression delimited by { } brackets. As noted
above, an example is V(1)=0.75

AT=value
A time at which the measurement calculation will occur. This is used by the
DERIV and FIND measures and the TRIG clause. Note that ill-considered use
of the FROM, TO, TD and AT qualifiers in the same TRIG-TARG measure
statement can cause an empty measurement window, and thus a failed
measure. Finally, the FROM and TO qualifiers take precedence over the AT
qualifier for DERIV and FIND measures.

FROM=value
A time (or frequency or DC sweep value) after which the measurement
calculation will start. For DC measures, this qualifier uses the first variable on
the .DC line.

TO=value
A time (or frequency or DC sweep value) at which the measurement
calculation will stop. For DC measures, this qualifier uses the first variable on
the .DC line.

59
TD=value
A time delay before which this measurement should be taken or checked.
Note that ill-considered use of both FROM and TO qualifiers and a TD qualifier
in the same measure statement can cause an empty measurement window,
and thus a failed measure.
MIN_THRESH=value
A minimum threshold value above which the measurement calculation will
be done and below which it will not be done. This is only used by the AVG
measure.
MAX_THRESH=value
A maximum threshold value above which the measurement calculation will
not be done and below which it will be done. This is only used by the AVG
measure.
RISE=r|LAST
The number of rises after which the measurement should be checked. If
LAST is specified, then the last rise found in the simulation will be used. It is
recommended that only one of the qualifiers RISE, FALL or CROSS be used in
a given measure statement. The exception is TRIG-TARG measures. In that
case, different RISE, FALL and CROSS criteria can be specified for TRIG and
TARG.
FALL=f|LAST
The number of falls after which the measurement should be checked. If LAST
is specified, then the last fall found in the simulation will be used.

CROSS=c|LAST
The number of zero crossings after which the measurement should be
checked. If LAST is specified, then the last zero crossing found in the
simulation will be used.
RFC_LEVEL=value
The level used to calculate rises, falls and crosses when the “level-crossing”
mode is used by measure types that do not support the WHEN qualifier. So,
RFC_LEVEL is used by the MAX, MIN and PP measures. Its usage is discussed
further in the subsection on “Rise, Fall and Cross Qualifiers”.
MINVAL=value
For the DERIV, DUTY, FIND, FREQ, OFF_TIME, ON_TIME and WHEN measures,
this is allowed difference between outVal and the variable to which it is
being compared. This has a default value of 1.0e-12. One may need to
specify a larger value to avoid missing the test condition in a transient run.
The descriptions of those seven measures detail how MINVAL is used by each
measure. For the ERR1 and ERR2 measures, if the absolute value of
<variable1 > is less than MINVAL, then MINVAL replaces the value of the
denominator of the ERR1 or ERR2 expression. For all measure types, that
support the FROM, TO and/or TD qualifiers, MINVAL also functions as a relative
tolerance for the comparison of the simulation time (or sweep value) to the
bounds of the measurement window. This allows for numerical-roundoff
errors if the FROM, TO and/or TD qualifiers are expressions.

60
YMIN=value
If the absolute value of <variable1 > in ERR1 or ERR2 measure is less than
the YMIN value then the ERR1 or ERR2 calculation does not consider that
point. The default is 1.0e-15. IGNOR and IGNORE are synonyms for YMIN.
YMAX=value
If the absolute value of <variable1 > in ERR1 or ERR2 measure is greater
than YMAX value then the ERR1 or ERR2 calculation does not consider that
point. The default is 1.0e15.
FRAC_MAX=value
A fractional value of the maximum value of <variable>. This is useful for
ensemble runs where the maximum value of a waveform is not known in
advance. FRAC_MAX is used by the TRIG and TARG measures for TRAN
measure mode, only.
ON=value
The value at which a signal is considered to be “on” for FREQ, DUTY and
ON_TIME measure calculations. This has a default value of 0.
OFF=value
The value at which a signal is considered to be “off” for FREQ, DUTY and
ON_TIME measure calculations. This has a default value of 0.
DEFAULT_VAL=value
If the conditions specified for finding a given measure’s value are not found
during the simulation then the measure will return a default value of 0. As
examples, a measure will fail if the condition specified by a WHEN or AT
qualifier is not found. It will also fail if the user specifies a set of FROM, TO
and TD values for a given measure that yields an empty measurement
interval. The default value for a given measure is settable by the user by
adding the qualifier DEFAULT_VAL=<retval> on that measure line. The
.OPTIONS MEASURE DEFAULT_VAL=<value> setting can be used to set the
default value of all of the measures in the netlist. The measure value in the
standard output or log file will always be FAILED. The measure value in the
circuitFileName.mt# (or circuitFileName.ms# or
circuitFileName.ma#) files will also be FAILED by default (.OPTIONS
MEASURE MEASFAIL=1). If .OPTIONS MEASURE MEASFAIL=0 is used in
the netlist then the measure value in the output file will be the default value.
See Section 2.1.25 for more details on the .OPTIONS settings. As a final note,
the FOUR measure is a special case since it produces multiline output. Failed
FOUR measures will be reported as FAILED in the circuitFileName.mt# (
or circuitFileName.ms# or circuitFileName.ma#) files, irrespective
of the various MEASFAIL and DEFAULT_VAL settings.
PRECISION=value
The default precision for .MEASURE output is 6 digits after the decimal point.
This argument provides a user configurable precision for a given .MEASURE
statement that applies to both the .mt# ( or .ms# or .ma#) files and standard
output. If .OPTIONS MEASURE MEASDGT=<val> is given in the netlist then
that value overrides the PRECISION parameters given on individual

61
 .MEASURE lines.
PRINT=value
This parameter controls where the .MEASURE output appears. The default is
ALL, which produces measure output in both the .mt# (or .ms# or .ma#) file
and to the standard output. A value of STDOUT only produces measure output
to standard output, while a value of NONE suppresses the measure output to
both the .mt# (or .ms# or .ma#) file and standard output. The subsection on
“Suppresing Measure Output” gives examples and also discuss the
interactions of this parameter with .OPTIONS MEASURE
MEASPRINT=<val>.
OUTPUT=value
This parameter is only supported for the MAX and MIN measures. The default
is VALUE. For TRAN measures, a value of VALUE will print the maximum (or
minimum) value to the .mt# file. A value of TIME will print the time of the
maximum (or minimum) value to the .mt# file. For DC measures, a value of
SV will output the value of the first variable on the .DC line to the .ms# file.
For AC and NOISE measures, a value of FREQ will print the frequency at
which the maximum (or minimum) value occurs to the .ma# file. This
parameter does not affect the descriptive output that is printed to the standard
output. The “Additional Examples” subsection gives an example for the MAX
measure.
VAL=value
This parameter is only implemented for the TRIG and TARG measures. It is
not the preferred Xyce syntax. It is only supported for HSPICE compatibility
(see that subsection, below, for details).
GOAL=value
This parameter is not implemented in Xyce, but is included for compatibility
with HSPICE netlists.
WEIGHT=value
This parameter is not implemented in Xyce, but is included for compatibility
with HSPICE netlists.
FILE=value
The filename for the “comparison file” used for the ERROR measure. This
qualifier is required for the ERROR measure.
INDEPVARCOL=value
The column index, in the “comparison file”, of the independent variable (e.g,
the simulation time or frequency) used in an ERROR measure. This qualifier
is required for the TRAN, AC and NOISE measure modes. For those modes, the
INDEPVARCOL and DEPVARCOL qualifiers must have different values. The
INDEPVARCOL qualifier is not used for DC mode ERROR measures, and will be
“silently ignored” in that case. Finally, note that the column indices in Xyce
output files start with 0.
DEPVARCOL=value
The column index, in the “comparison file”, of the dependent variable used
in an ERROR measure. This qualifier is required for the ERROR measure for all

62
 four measure modes (TRAN, AC, DC and NOISE). For the TRAN, AC and NOISE
measure modes, the DEPVARCOL and INDEPVARCOL qualifiers must have
different values. Finally, note that the column indices in Xyce output files
start with 0.
COMP_FUNCTION=value
This is the norm used by the ERROR measure to compare the simulation
values for the measured variable with the corresponding values in the
“comparison file” specified with the FILE qualifier. The allowed values are
L1NORM, L2NORM and INFNORM. Any other values will default to L2NORM.
This qualifier is optional for the ERROR measure, and has a default value of
L2NORM. The descriptive output for each ERROR measure, that is printed to
standard output, will explicitly state which norm was used for each ERROR
measure.

2.1.18.1. Measure Output

As previously mentioned, measured results are reported to the output and log file. Additionally, for TRAN
measures, the results are stored in files called circuitFileName.mt#, where the suffixed number (#)
starts at 0 and increases for multiple iterations (.STEP iterations) of a given simulation. For DC measures,
the results are stored in the files circuitFileName.ms#, while AC and NOISE measures use the files
circuitFileName.ma#.
A user-defined measure can also be output at each time-step via inclusion in a .PRINT command. For
example, this netlist excerpt outputs the integral of V(1) at each time step. The measure value TINTV1 is
then also output at the end of the simulation to both the standard output and the .mt# (or .ms# or .ma#)
files.

.MEASURE TRAN TINTV1 INTEG V(1)

.PRINT TRAN FORMAT=NOINDEX V(1) TINTV1

The output for successful and failed measures to the standard output (and log files) provides more
information than just the measure’s calculated values. As an example, for a successful and failed MAX
measure the standard output would be:

MAXVAL = 0.999758 at time = 0.000249037

Measure Start Time= 0 Measure End Time= 0.001

Netlist warning: MAXFAIL failed. TO value < FROM value

MAXFAIL = FAILED at time = 0
Measure Start Time= 1 Measure End Time= 0.001

In general, information on the measurement window, the time(s) that the measure’s value(s) were calculated
and a possible cause for a failed measure are output to standard output for all measures except for FOUR.
This information is similar, but not identical, to HSPICE’s verbose output. For a failed FOUR measure, the
standard output will have “FAILED”, but there may be less information provided as to why the FOUR
measure failed.
In this example, the circuitFileName.mt# file would have the following output:

63
MAXVAL = 0.999758
MAXFAIL = -1

2.1.18.2. Measurement Windows

There is an implicit precedence when multiple qualifiers are specified to limit the measurement window for
a given .MEASURE statement for TRAN measures. In general, Xyce first considers the time-window criteria of
the FROM, TO and TD qualifiers. If the simulation time is within that user-specified time-window then the
RISE, FALL, CROSS are qualifiers are counted and/or the TRIG, TARG and WHEN qualifiers are evaluated.
The following netlist excerpt shows simple examples where the .MEASURE statement may return the default
value because the measure “failed”. For riseSine, this may occur because V(1) never has an output value
of 1.0 because of the time steps chosen by Xyce. So, careful selection of the threshold values in WHEN,
TRIG and TARG clauses may be needed in some cases. For fallPulseFracMax, the simulation interval is
too short and the TARG value of 0.3 for V(2) is not reached within the specified one-second simulation time.
For maxSine, the FROM, TO and TD values yield an empty time interval, which is typically an error in netlist
entry.

VS 1 0 SIN(0 1.0 0.5 0 0)

VP 2 0 PULSE( 0 10 0.2 0.2 0.2 0.5 2)
R1 1 0 100K
R2 2 0 100K
.TRAN 0 1
.PRINT TRAN FORMAT=NOINDEX V(1) V(2)
.MEASURE TRAN riseSine TRIG V(1)=0 TARG V(1)=1.0
.MEASURE TRAN fallPulseFracMax TRIG V(2) FRAC_MAX=0.97
+ TARG V(2) FRAC_MAX=0.03
.MEASURE TRAN maxSine MAX V(1) FROM=0.2 TO=0.25 TD=0.5

The intent in Xyce is for the measurement window to be the intersection of the FROM-TO and TD windows, if
both are specified. As noted above, the use of both FROM-TO and TD windows can lead to an empty
measurement window. So, that usage is not recommended.

2.1.18.3. Expression Support

These measure “qualifiers” (TO, FROM, TD, RISE, FALL, CROSS, AT, OFF, ON, DEFAULT_VAL and VAL)
support expressions. The caveat is that the expression must evaluate to a constant at the time that each
measure object is made. So, that expression can not depend on solution variables or lead currents. This
limitation matches HSPICE. It also can not depend on a global parameter. Finally, it can not depend on
another measure’s value, which is an allowed syntax in HSPICE.
Simple examples of allowed syntaxes for qualifiers are as follows, where all three measures will get the
same answer:

.PARAM t1=0.2
.PARAM t2=0.3
.MEASURE TRAN M1 PP V(1) FROM=‘0.1+0.2’

64
.MEASURE TRAN M2 PP V(1) FROM={0.1+t1}
.MEASURE TRAN M3 PP V(1) TO={t2}

Expressions should also work in FIND-WHEN, WHEN and TRIG-TARG measures. The preferred Xyce
syntax with curly braces and the three legal HSPICE syntaxes for expressions should all work. However,
note that the two HSPICE expression syntaxes shown below are only legal in Xyce .MEASURE statements.

.PARAM a1=0.1
.PARAM a2=0.7
.MEASURE TRAN M4 FIND V(2) WHEN V(1)={a1}
.MEASURE TRAN M4PAR FIND V(2) WHEN V(1)=PAR(‘a1’) ; HSPICE exp. syntax
.MEASURE TRAN M4PAREN FIND V(2) WHEN V(1)=(‘a1’) ; HSPICE exp. syntax
.MEASURE TRAN M5 WHEN V(1)={a1}
.MEASURE TRAN M6 TRIG {v(1)-0.1} VAL={a1} TARG {v(1)-0.5} VAL={a2}

2.1.18.4. Re-Measure

Xyce can re-calculate (or re-measure) the values for .MEASURE and/or .FFT statements using existing Xyce
output files. This is useful for tuning .MEASURE and/or .FFT statements to better capture response metrics
for a circuit when the underlying simulation runtime is long. To use this functionality, add the command
line argument -remeasure <file>, where <file> is a Xyce-generated .prn, .csv or .csd output file.

There are several important limitations with -remeasure:

• The data required by the .MEASURE and/or .FFT statements must have been output in the simulation
output file. When using -remeasure, Xyce does not recalculate the full solution, but uses the data
supplied in the output file instead. Thus, everything a .MEASURE and/or .FFT statement needs to
calculate its results must be in the output file. So, the nodal voltages (e.g., node A), lead currents (e.g,
for device R1) and branch currrents requested by the .MEASURE statements must have been used, at
least once, on the .PRINT statement in the form of V(A), N(a) or I(R1). They can not only appear
on the .PRINT line within an expression or a voltage-difference operator.

• Only voltage node values, lead currents and branch currents can be used in .MEASURE statements
while using -remeasure. Power values will not be interpreted correctly during a re-measure
operation. A work-around for that limitation is illustrated below.

• -remeasure only works with .tran or .dc analyses. However, it can be used with .STEP in both
cases. It is not currently supported for .ac analyses.

• For .tran analyses, -remeasure works with .prn, .csv and .csd formatted output data. However,
it might only work with .csv and .csd files generated by Xyce.

• For .dc analyses, -remeasure works with .prn and .csd formatted output data. However, it might
only work with .csd files generated by Xyce.

• -remeasure will fail if the netlist has a .op statement that precedes the .tran or .dc statement.
This can be fixed by either moving the .op statement or by temporarily commenting the .op
statement out during -remeasure.

65
As an example in using -remeasure, consider a netlist called myCircuit.cir which had previously been
run in Xyce and produced the output file myCircuit.cir.prn. One could run -remeasure with the
following command:

Xyce -remeasure myCircuit.cir.prn myCircuit.cir

A work-around for re-measuring power values (e.g., for device R1) is to use this combination of .PRINT
and .MEASURE lines in the netlist. As noted above, expressions will work with re-measure if all of the
quantities used in the expression also appear outside of an expression on the .PRINT line.

R1 a b 1
.PRINT TRAN V(a) V(b) I(R1)
.MEASURE TRAN PR1B MAX {(V(a)-V(b))*I(R1)}

2.1.18.5. RISE, FALL and CROSS Qualifiers

The RISE, FALL and CROSS qualifiers are supported for more measures types, and in more ways, in Xyce
than in HSPICE for TRAN meaures. This sections explains those differences and supplies some examples.
One key difference is that Xyce supports two different “modes” for these qualifiers for TRAN meaures.
The first mode is “level-crossing”, where the RISE, FALL and CROSS counts are incremented each time the
measured signal (e.g, V(a)) crosses the user-specified level (termed crossVal here). This mode should
work identically to HSPICE for the DERIV-WHEN, FIND-WHEN, WHEN and TRIG-TARG measures.
If we define currentVal and lastVal as the current and previous values of V(a), and riseCount,
fallCount and crossCount as the number of rises, falls and crosses that have occurred, then the
pseudo-code for the “level-crossing” mode is:

if ( (currentVal-crossVal >= 0.0) AND (lastVal-crossVal < 0.0) )

{
riseCount++;
crossCount++;
}
else if( (currentVal-crossVal) <= 0.0) AND (lastVal-crossVal > 0.0) )
{
fallCount++;
crossCount++;
}

For DERIV-WHEN, FIND-WHEN, WHEN, measures, the cross value is set by the value (or second variable) in
the WHEN clause. For TRIG-TARG measures, the cross values are set separately by the the values (or second
variables) in the TRIG and TARG clauses.
The second mode is termed “absolute”. In this mode, Xyce attempts to auto-detect whether the measured
waveform has started a new rise or fall. However, the crossCount is still evaluated against a fixed
crossVal of 0. This mode may be useful for pulse waveforms with sharp rises and falls, where the
waveform’s maximum (or minimum) level is not exactly known in advance. It may not work well with noisy
waveforms.

66
If we define two Boolean variables isRising and isFalling then the pseudo-code for the “absolute”
mode is:

if( (currentVal > lastVal) AND !isRising )

{
isRising= true;
isFalling = false;
riseCount++;
}
else if( (currentVal < lastVal) AND !isFalling )
{
isRising = false;
isFalling = true;
fallCount++;
}
if ( ( (currentVal >= 0.0) AND (lastVal < 0.0) ) OR
( (currentVal <= 0.0) AND (lastVal > 0.0) ) )
{
crossCount++;
}

The following table shows which of these two modes are supported for which Xyce measure types.

Table 2-2. RISE, FALL and CROSS Support in .MEASURE.

Measure Level-Crossing Absolute
The crossVal is set by the value of the
No
DERIV-WHEN WHEN clause
FIND-WHEN and The crossVal is set by the value of the
No
WHEN WHEN clause
A fixed crossVal can be set with
MAX Default, if RFC_LEVEL is not set
RFC_LEVEL
A fixed crossVal can be set with
MIN Default, if RFC_LEVEL is not set
RFC_LEVEL
A fixed crossVal can be set with
PP Default, if RFC_LEVEL is not set
RFC_LEVEL
The levels are set separately by the values in
TRIG and TARG Only if FRAC_MAX is used
the TRIG and TARG clauses

As simple examples of these two modes for the MAX measure, consider the following netlist:

*examples of RFC modes

VPWL1 1 0 PWL(0 0 0.2 0.5 0.4 0 0.6 0.75 0.8 0 1.0 0.75 1.2 0.0)
R1 1 0 100
.TRAN 0 1.2s
.MEASURE TRAN MAX1 MAX V(1) RISE=1
.MEASURE TRAN MAX2 MAX V(1) RISE=1 RFC_LEVEL=0.6
.MEASURE TRAN MAX3 MAX V(1) FALL=1 RFC_LEVEL=0.5

67
.PRINT TRAN V(1) MAX1 MAX2 MAX3
.END

The descriptive output to standard output would then be:

MAX1 = 5.000000e-01 at time = 2.000000e-01

Measure Start Time= 0.000000e+00 Measure End Time= 1.200000e+00
Rise 1: Start Time= 1.000000e-10 End Time= 4.000000e-01

MAX2 = 7.500000e-01 at time = 6.000000e-01

Measure Start Time= 0.000000e+00 Measure End Time= 1.200000e+00
Rise 1: Start Time= 5.600000e-01 End Time= 9.500000e-01

MAX3 = 7.500000e-01 at time = 1.000000e+00

Measure Start Time= 0.000000e+00 Measure End Time= 1.200000e+00
Fall 1: Start Time= 6.700000e-01 End Time= 1.060000e+00

The MAX1 measure uses the “absolute” mode, so the first rise begins with the very first time-step. The
maximum value in that first rise interval for measure MAX1 then occurs at time=0.2s. The MAX2 measure
uses the “level-crossing” mode with a user-specified RFC_LEVEL of 0.6V. So, the first rise interval for the
MAX2 measure begins at time=0.56s, and the maximum value in that first rise interval occurs at time=0.6s.
The MAX3 measure illustrates an important point. A “fall” is not recorded for the MAX3 measure at t=0.2
seconds, but a “rise” (and “cross”) would be recorded, since the value of V(1) is exactly equal to the
user-specified RFC_LEVEL. So, the first fall interval for measure MAX3 begins at time=0.67s, when V(1) first
passes through the user-specified RFC_LEVEL of 0.5V.

2.1.18.6. Additional Examples

Pulse width measurements in Xyce can be done as follows, based on this netlist excerpt. This may be useful
for ensemble runs, where the maximum value of a one-shot pulse is not known in advance. The first syntax
uses three measure statements to measure the 50% pulse width, and works with noisy waveforms. The
second syntax uses only one measure statement, but may not always work with noisy waveforms.

* pulse-width measurement example 1

.measure tran rise50FracMax trig v(1) frac_max=0.5 targ v(1) frac_max=1
.measure tran fall50FracMax trig v(1) frac_max=1 targ v(1) frac_max=0.5
.measure tran 50width EQN{rise50FracMax + fall50FracMax}

* pulse-width measurement example 2

.measure tran 50widthFracMax trig v(1) frac_max=0.50
+ targ v(1) frac_max=0.50 FALL=1

In some cases, the user may wish to print out both the measure value and measure time (or the value of the
first variable on the .DC line) of a MAX or MIN measure to the .mt0 file. For a TRAN measure, this can be
done for these two measures with the OUTPUT keyword as follows:

68
* printing maximum value and time of maximum value to .mt0 file
.TRAN 0 1
V1 1 0 PWL 0 0 0.5 1 1 0
R1 1 0 1
.MEASURE TRAN MAXVAL MAX V(1)
.MEASURE TRAN TIMEOFMAXVAL MAX V(1) OUTPUT=TIME

The output to the .mt0 file would be:

MAXVAL = 1.000000e+00
TIMEOFMAXVAL = 5.000000e-01

The descriptive output to standard output would be the same for both measures though. The measure value
and measure time are not re-ordered in the descriptive output when OUTPUT=VALUE is used for the MAX or
MIN measures.

MAXVAL = 1.000000e+00 at time = 5.000000e-01

Measure Start Time= 0.000000e+00 Measure End Time= 1.000000e+00

TIMEOFMAXVAL = 1.000000e+00 at time = 5.000000e-01

Measure Start Time= 0.000000e+00 Measure End Time= 1.000000e+00

For a DC measure, one would use OUTPUT=SV instead of OUTPUT=TIME. In that case, the “sweep value”
(SV) is the value of the first variable on the .DC line. For an AC or NOISE measure, one would use
OUTPUT=FREQ.

2.1.18.7. Suppresssing Measure Output

If the Xyce output is post-processed with other programs, such as Dakota, it may be desirable to only print a
subset of the measure values to the .mt# (or .ms# or .ma#) files, but to print all of the measure output to
standard output. As an example, these .MEASURE statements:

.TRAN 0 2ms
.measure tran minSineOne min V(1) print=none
.measure tran minSinTwo min V(2) print=stdout
.measure tran minSinThree min V(3) print=all
.measure tran sinSinFive min V(4)

would produce the following measure output in the .mt0 file:

MINSINTHREE = -3.851422e-01
MINSINFOUR = -1.998548e+00

and the following measure output in standard output:

69
MINSINTWO = -1.188589e+00 at time = 7.400000e-04
Measure Start Time= 0.000000e+00 Measure End Time= 2.000000e-03

MINSINTHREE = -3.851422e-01 at time = 2.400000e-04

Measure Start Time= 0.000000e+00 Measure End Time= 2.000000e-03

MINSINFOUR = -1.998548e+00 at time = 7.500000e-04

Measure Start Time= 0.000000e+00 Measure End Time= 2.000000e-03

.OPTIONS MEASURE MEASPRINT=<val> also provides the option to accomplish these same effects, but for
all of the measure statements in the netlist. The interactions between these two features are as follows. If
MEASPRINT=ALL is used, which is the default setting, then the PRINT qualifier on a given .MEASURE line
will override that setting. However, MEASPRINT=NONE and MEASPRINT=STDOUT will take precedence over
the PRINT qualifiers on individual .MEASURE lines. Finally, the MEASPRINT option will be ignored during
remeasure, but the PRINT qualifiers on individual measure lines will be used.
.OPTIONS MEASURE MEASOUT=<val> provides another way to suppress the output of the .mt# (or .ms#
or .ma#) files. See Section 2.1.25 for more details. If given, this option takes precedence over the
MEASPRINT option setting. However, it is also ignored during remeasure.

2.1.18.8. ERROR Functions (ERR1 and ERR2)

This subsection defines the calculation functions for the ERR1 and ERR2 measure types. For the ERR1
measure, the measure value is calculated as follows, where 𝑀𝑖 and 𝐶𝑖 are the first and second variables on
the measure line and N is the number of time, frequency or DC sweep values included in the measure
calculation: v
u
t 𝑁  2
1 ∑︁ 𝑀𝑖 − 𝐶𝑖
𝐸 𝑅𝑅1 = (2.1)
𝑁 𝑖=1 𝑚𝑎𝑥(𝑀 𝐼 𝑁𝑉 𝐴𝐿, | 𝑀𝑖 |)

For the ERR2 measure, the value is:

𝑁
1 ∑︁ 𝑀𝑖 − 𝐶𝑖
𝐸 𝑅𝑅2 = (2.2)
𝑁 𝑖=1 𝑚𝑎𝑥(𝑀 𝐼 𝑁𝑉 𝐴𝐿, | 𝑀𝑖 |)

For both measures, if the absolute value of 𝑀𝑖 is less than the YMIN value or greater than the YMAX value
then the ERR1 or ERR2 calculation does not consider that point. The default for YMIN is 1.0-e15. The default
for YMAX is 1.0e15.

2.1.18.9. ERROR Measure

The Xyce ERROR measure is not the functional equivalent of the ERR1 or ERR2 measures. It is intended to
solve a different problem, namely the comparison of data in multiple simulation runs to an assumed “gold
standard” read in from a file. It also uses different comparision functions then the ERR1 and ERR2
measures.. This subsection lists some important caveats with the use of the ERROR measure.
• The comparison file, specified with the FILE qualifier, can be .prn, .csv and .csd formatted output
data. However, the ERROR measure might only work with .csv and .csd files generated by Xyce.

70
 • The data in the comparison file is assumed to be “non-step data”, from one simulation iteration. The
simulated data can use .STEP though and the ERROR measure values will be re-evaluated for each
step.

• For TRAN (or AC or NOISE) measures, the values of the measured waveform are interpolated to the
simulation times (or frequencies) in the comparison waveform. So, the norm calculation is inherently
windowed to the time (or frequency) interval of the comparison waveform. For the best interpolation
results for AC or NOISE measures, it is recommended that the frequency extent of the comparison
waveform be greater than or equal to the frequency extent of the measured waveforms.

• For DC measures, interpolation is not used. So, the values of the simulated and comparion waveforms
are compared at the values specified by the DEPVARCOL qualifier. Any value for the INDEPVARCOL
qualifier specified on a DC measure line will be “silently ignored”.

• The time and frequency window constraints (TO, FROM and TD qualifiers) are not supported for the
ERROR measure. So, as noted above, the effective window for the norm calculation is set by the extent
of the comparison waveform.

• The values in the column in the comparison file specified with the INDEPVARCOL qualifier must be
monotonically increasing for a TRAN, AC or NOISE measure. Otherwise, Xyce will not run the
simulation.

• The ERROR measure currently supports the L1, L2 and INFNORM, with the default being the L2
norm. If anything other than L1, L2 or INFNORM is specified, Xyce will default to the L2 norm. The
descriptive output for each ERROR measure, that is printed to standard output, will explicitly state
which norm was used for each ERROR measure. (Note: The norm value is selected with the
COMP_FUNCTION qualifier, and the allowed values are L1NORM, L2NORM and INFNORM.)

As a final note, the ERROR measure can enable the use of Xyce simulation output in optimization problems,
like device calibration. However, for internal Sandia users, there may be better approaches that leverage the
combined capabilities of Sandia’s Dakota and Xyce software packages.

2.1.18.10. Operator Support for AC Mode Measures

All of the operators supported on .PRINT AC lines are supported for AC measure mode. The linear
parameter operators (e.g., SR(1,1)) are only supported when a .LIN analysis is done, but their values can
be used in .MEASURE AC statements in that case.

One caveat is that AC mode measures that use V(a) will actually measure VR(a). The same caveat applies
to the use of S(1,1). An AC mode measure would measure SR(1,1) instead.

2.1.18.11. Operator Support for NOISE Mode Measures

All of the operators supported on .PRINT NOISE lines are supported for NOISE measure mode. One
caveat is that NOISE mode measures that use V(a) will actually measure VR(a).

71
2.1.18.12. Behavior for Unsupported Modes and Types

The .MEASURE statement is supported for .TRAN, .AC, .DC and .NOISE analyses. It can be used with .STEP
in all four cases. So, Xyce does not support HB measure mode. If that mode is included in the netlist then
Xyce parsing will fail and emit error messages. Similarly, Xyce parsing will fail if the requested measure
type is not supported for a given measure mode (e.g., OFF_TIME for a AC, DC or NOISE measure).

2.1.18.13. Compatibility with .DATA

The .DATA command can be used to specify table-based .AC, .DC or .NOISE sweeps for those three
analysis types. For AC and NOISE measures, the “swept variable” then uses the frequency values in the table
specified on the .AC or .NOISE line.

For DC measures, the swept variable uses the row index in the table specified on the .DC line. An example is
as follows:

* example of .DATA with DC measures

V1 1 0 1
R1 1 2 1
R2 2 0 1

.data test
+ r1 r2
+ 1.0e+00 4.0e+00
+ 4.0e+00 6.0e+00
+ 6.0e+00 4.0e+00
.enddata

.DC data=test
.print DC V(1) V(2)

.OPTIONS MEASURE MEASDGT=1

.MEASURE DC MAXV2TO MAX V(2) TO=2
.MEASURE DC MAXV2FROM MAX V(2) FROM=2

.END

The measure results reported in stdout will be as follows where the respective maximum values occur for
the R1 and R2 values given in the first and second rows of the test table:

MAXV2TO = 8.0e-01 at Table Row value = 1.0e+00

Measure Start Table Row Value= 1.0e+00 Measure End Table Row Value= 2.0e+00

MAXV2FROM = 6.0e-01 at Table Row value = 2.0e+00

Measure Start Table Row Value= 2.0e+00 Measure End Table Row Value= 3.0e+00

72
All valid measure types will return an answer when a data-based sweep is used on the .AC, .DC or .NOISE
line. However, the results for AVG, DERIV, FIND, INTEG, RMS and WHEN measures may be “non-physical” if
the frequency values in the data table are not monotonically increasing. In addition, for DC measures the
effective step size between table rows is equal to one.

2.1.18.14. HSPICE Compatibility

There are known incompatibilities between the Xyce and HSPICE implementation of .MEASURE. They
include the following:

• Since .AC and .NOISE are separate analysis types in Xyce, there are separate AC and NOISE measure
modes.

• Several of the Xyce measure types (DUTY, EQN, FREQ, FOUR, ON_TIME, and OFF_TIME) and qualifers
(e.g., FRAC_MAX) are not found in HSPICE. Several HSPICE measure types are not supported in Xyce.

• The default, in both HSPICE and Xyce. for DERIV-WHEN, FIND-WHEN, WHEN and TRIG-TARG
measures is to use CROSS=0 if a RISE, FALL or CROSS value is not explicitly given in the WHEN, TRIG
or TARG clause, However, the Xyce and HSPICE results may differ in those cases if the measured
signal is either a constant value or meets the measure criteria at the first simulation step (e.g., t=0).

• The HSPICE qualifers of REVERSE and PREVIOUS are not supported in Xyce.

• The HSPICE .POWER statement, which prints out a table with the AVG, RMS, MIN and MAX
measures for each specified signal, is not supported in Xyce.

• Xyce generally supports more qualifiers (FROM, TO, TD, RISE, FALL and CROSS) for the measurement
windows for a given measure-type. So, some legal Xyce syntaxes may not be legal in HSPICE.

• The Xyce EQN measure can calculate an expression based on other measure values. So, one of its
usages is similar to the HSPICE PARAM measure. However, their syntaxes are different.

• A mismatch between the measure mode and the analysis mode (e.g., a DC measure in a netlist that
uses a .TRAN analysis statement) will cause a Xyce netlist parsing error. That same mismatch might
be silently ignored by HSPICE.

• How Xyce and HSPICE handle “steps” may be different. In Xyce, the “steps” in the measured data
(e.g., the generation of new .mt# or .ms# or .ma# files) are triggered by the variable(s) on the .STEP
line, but not by the variable(s) on the .DC line.

• Expressions on .MEASURE lines are supported in fewer contexts then in HSPICE. See the “Expression
Support” subsection for more details.

• The settings for the MEASFAIL and MEASOUT options are only used if those options are explicitly
given in the netlist. Otherwise, the Xyce defaults will be used.

The following HSPICE syntax (VAL=0.9) is supported in Xyce for TRIG and TARG measures. However, the
preferred Xyce syntax would use targ v(1)=0.9 instead.

.measure tran riseSine trig v(1) AT=0.0001 targ v(1) VAL=0.9 RISE=1

73
The remainder of this subsection discusses alternate syntaxes for Xyce measure lines that are supported for
improved HSPICE compatibility. The definitions of the measure syntaxes given at the beginning of this
.MEASURE section give the preferred Xyce syntaxes. However, PARAM (and the equivalent EQN) measure
lines are allowed with, or without, the equal sign after the PARAM keyword. So, these two Xyce measure
statements are equivalent:

.measure tran noEqualSgn PARAM {v(1)+1.0}

.measure tran equalSgn PARAM={v(1)+1.0}

There are multiple expression syntaxes that are allowed in various contexts on HSPICE measure lines. So,
all of these example syntaxes are allowed in expression contexts on Xyce measure lines. (Note: Only the
first single-quote-delimited expression format is supported in all Xyce expression contexts, in addition to
the Xyce curly-braces format.)

.measure tran curlyBraces MAX {V(1)+1}

.measure tran singleQuote MAX ‘V(1)+1’
.measure tran parenSingleQuote MAX (‘V(1)+1’)
.measure tran parSyntax MAX PAR(‘V(1)+1’)

Undelimited expressions are allowed in some contexts in HSPICE. However, the syntax for the
notLegalInXyce measure shown below is not allowed in Xyce, since it uses an undelimited expression.

.measure tran PLUS PP PAR(‘V(1)+V(2)’)

.measure tran notLegalInXyce PARAM PLUS+2.0 ; not legal

2.1.18.15. Legacy Trig-Targ Mode (LTTM)

For the Xyce 7.5 release, the code for the TRIG-TARG measure was extensively re-written to provide better
HSPICE (and ngspice) compatibility. However, mostly for backwards compatibility for internal Sandia
users, the previous behavior can be recovered by using .OPTION MEASURE USE_LTTM=1 in the netlist. It is
anticipated that this feature will be removed in a future release. However, support for the FRAC_MAX
qualifier, which is not in either HSPICE or ngspice, will likely be continued.

The allowed syntaxes for this mode are shown below. Note that this mode lacks several features. In
particular:

• The AT qualifer is only allowed in the TRIG clause.

• The FROM, TO and TD qualifiers apply to both the TRIG and TARG clauses.

• Expression support, especially in the TARG clause, was less available.

74
General Form .MEASURE TRAN <result name> TRIG <variable>=<variable2 >|<value>
+ [RISE=r1|LAST] [FALL=f1|LAST] [CROSS=c1|LAST]
+ TARG <variable3 >=<variable4 >|<value>
+ [RISE=r2|LAST] [FALL=f2|LAST] [CROSS=c2|LAST]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN <result name> TRIG AT=<value>

+ TARG <variable2 >=<variable3 >|<value>
+ [RISE=r2|LAST] [FALL=f2|LAST] [CROSS=c2|LAST]
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

75
2.1.19. .MEASURE (Continuous results)

“Continuous” measure results are supported for DERIV-AT, DERIV-WHEN, FIND-AT, FIND-WHEN, WHEN and
TRIG-TARG measures for .TRAN, .DC, .AC and .NOISE analyses. They are identical to the
“non-continuous” versions, except that they can return more than one measured value in some cases.

General Form .MEASURE <AC_CONT|DC_CONT|NOISE_CONT|TRAN_CONT> <result name>

+ DERIV <variable> AT=<value>
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN_CONT <result name>

+ DERIV <variable> WHEN <variable>=<variable2 >|<value>
+ [MINVAL=<value>] [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC_CONT|DC_CONT|NOISE_CONT> <result name>

+ DERIV <variable> WHEN <variable>=<variable2 >|<value>
+ [MINVAL=<value>] [FROM=<value>] [TO=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC_CONT|DC_CONT|NOISE_CONT|TRAN_CONT> <result name>

+ FIND <variable> AT=<value>
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN_CONT <result name>

+ FIND <variable> WHEN <variable>=<variable2 >|<value>
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC_CONT|DC_CONT|NOISE_CONT> <result name>

+ FIND <variable> WHEN <variable>=<variable2 >|<value>
+ [FROM=<value>] [TO=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE TRAN_CONT <result name>

+ WHEN <variable>=<variable2 >|<value>
+ [FROM=<value>] [TO=<value>] [TD=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

76
 .MEASURE <AC_CONT|DC_CONT|NOISE_CONT> <result name>
+ WHEN <variable>=<variable2 >|<value>
+ [FROM=<value>] [TO=<value>]
+ [RISE=r|LAST] [FALL=f|LAST] [CROSS=c|LAST]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC_CONT|DC_CONT|NOISE_CONT|TRAN_CONT> <result name>

+ TRIG <variable1 >=<variable2 >|<value>
+ [TD=<val>] [RISE=r] [FALL=f] [CROSS=c]
+ TARG <variable3 >=<variable4 >|<value>
+ [TD=<val>] [RISE=r] [FALL=f] [CROSS=c]
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE <AC_CONT|DC_CONT|NOISE_CONT|TRAN_CONT> <result name>

+ TRIG AT=<value> TARG AT=<value>
+ [MINVAL=<value>] [DEFAULT_VAL=<value>]
+ [PRECISION=<value>] [PRINT=<value>]

Examples .MEASURE TRAN_CONT DERIV1At5 DERIV V(1) AT=5

.MEASURE DC_CONT deriv2 DERIV WHEN V(2)=0.75
.MEASURE AC_CONT find1at5 FIND V(1) AT=5
.MEASURE NOISE_CONT find1 FIND V(1) WHEN V(2)=1 RISE=2
.MEASURE TRAN_CONT whenv1 WHEN V(1)=5
.MEASURE TRAN_CONT TrigTargAT TRIG AT=2ms TARG AT=8ms
.MEASURE TRAN_CONT TrigTargAT1 TRIG V(1)=0.2 CROSS=1
+ TARG AT=8ms
.MEASURE TRAN_CONT TrigTargAT2 TRIG AT=2ms
+ TARG V(1)=0.2 CROSS=1
.MEASURE TRAN_CONT TrigTarg TRIG V(1)=0.2 CROSS=1
+ TARG V(1)=0.3 CROSS=1 TD=8ms

Arguments and
Options result name
Measured results are reported to the log file and (possibly) multiple output
files. Section 2.1.19.1 below gives more information on the output files
produced by continuous mode measures.
The <result name> must be a legal Xyce character string. If multiple
measures are defined with the same <result name> then Xyce uses the last
such definition, and issues warning messages about (and discards) any
previous measure definitions with the same <result name>.
measure type
DERIV, FIND, WHEN, TRIG, TARG

77
The third argument specifies the type of measurement or calculation to be
done.
By default, the measurement is performed over the entire simulation. The
calculations can be limited to a specific measurement window by using the
qualifiers FROM, TO, TD, RISE, FALL, CROSS and MINVAL, which are
explained below and in section 2.1.18..
The supported “continuous” measure types and their definitions are:
DERIV Computes the derivative of <variable> at a user-specified time (by
using the AT qualifier) or when a user-specified condition occurs (by
using the WHEN qualifier). If the WHEN qualifier is used then the
measurement window can be limited with the qualifiers FROM, TO, RISE,
FALL and CROSS for all measure modes. In addition, the TD qualifier is
supported for TRAN_CONT measures. The MINVAL qualifier is used as a
comparison tolerance for both AT and WHEN. For HSPICE compatibility,
DERIVATIVE is an allowed synonym for DERIV.
FIND-AT Returns the value of <variable> at the time when the AT clause
is satisfied. The AT clause is described in more detail later in this list.
FIND-WHEN Returns the value of <variable> at the time when the WHEN
clause is satisfied. The WHEN clause is described in more detail later in
this list.
WHEN Returns the time (or frequency or DC sweep value) when
<variable> reaches <variable2 > or the constant value, value. The
measurement window can be limited with the qualifiers FROM, TO, RISE,
FALL and CROSS for all measure modes. In addition, the TD qualifier is
supported for TRAN_CONT measures. The qualifier MINVAL acts as a
tolerance for the comparison. For example when <variable2 > is
specified, the comparison used is when <variable> = <variable2 > ±
MINVAL or when a constant, value is given: <variable> = value ±
MINVAL. If the conditions specified for finding a given value were not
found during the simulation then the measure will return the default
value of -1. The user may change this default value with the
DEFAULT_VAL qualifier. Note: The use of FIND and WHEN in one
measure statement is also supported.
TRIG
TARG Measures the time between a trigger event and a target event. The
trigger is specified with TRIG <variable1 >=<variable2 > or TRIG
<variable1 >=<value> or TRIG AT=<value>. The target is specified
as TARG <variable3 >=<variable4 > or TARG
<variable3 >=<value> or TARG AT=<value>. The measurement
window can be limited with the qualifiers TD, RISE, FALL and CROSS for
all measure modes. The qualifier MINVAL acts as a tolerance for the
comparison. For example when <variable2 > is specified, the
comparison used is when <variable1 > = <variable2 > ± MINVAL or
when a constant, value is given: <variable1 > = value ± MINVAL. If
the conditions specified for finding a given value were not found during

78
 the simulation then the measure will return the default value of -1. The
user may change this default value with the DEFAULT_VAL qualifier.
variable
variable𝑛
value
These quantities represents the test for the stated measurement. <variable>
is a simulation quantity, such as a voltage or current. One can compare it to
another simulation variable or a fixed quantity. Additionally, the
<variable> may be a Xyce expression delimited by { } brackets. As noted
above, an example is V(2)=0.75
Additional information on the TO, FROM, TD, RISE, FALL, CROSS, MINVAL,
DEFAUAL_VAL, PRECISION and PRINT qualifiers is given in section 2.1.18.

2.1.19.1. Measure Output

As discussed in section 2.1.18.1, measured results for AC, DC, NOISE and TRAN mode measures are reported
to the log file. Additionally, for TRAN measures, the results are stored in files called
circuitFileName.mt#, where the suffixed number (#) starts at 0 and increases for multiple iterations
(.STEP iterations) of a given simulation. For DC measures, the results are stored in the files
circuitFileName.ms#, while AC and NOISE measures use the files circuitFileName.ma#.
For AC_CONT, DC_CONT, NOISE_CONT and TRAN_CONT mode measures, the output for successful and failed
measures is sent to the standard output (and log files), as described in section 2.1.18.1. There are two
options for the output files though. The default is for each continuous mode measure to generate its own
output file where, for example for a non-step transient analysis, the file name would be
circuitFileName_resultname.mt0 where the result (measure) name is always output in lower-case.
This default matches HSPICE. The second option uses .OPTIONS MEASURE USE_CONT_FILES=0. In that
case, the results for all of the continuous mode measures are sent to the circuitFileName.mt# file.
An example is as follows.

VPWL1 1 0 pwl(0 0 2.5m 1 5m 0 7.5m 1 10m 0)

R1 1 0 1

.TRAN 0 10ms
.PRINT TRAN V(1)

.MEASURE TRAN MAXV1 MAX V(1)

.MEASURE TRAN_CONT FindV1 WHEN V(1)=0.5
.MEASURE TRAN_CONT FindV1AT FIND V(1) AT=0.6ms

.END

The result for measure MAXV1 is sent to <netlistName>.mt0. The results for measures FindV1 and
FindV1AT are then sent to individual files, named <netlistName>_findv1.mt0 and
<netlistName>_findv1at.mt0. Note that the measure names have been lower-cased in the output file
names. The contents of those files are then as follows.

79
FINDV1 = 1.250000e-03
FINDV1 = 3.750000e-03
FINDV1 = 6.250000e-03
FINDV1 = 8.750000e-03

and:

FINDV1AT = 2.400000e-01

Note that FIND-AT measures will still only return one measure value, even for TRAN_CONT measure mode.
However, in this simple example, the specified FIND-WHEN measure returns all four times where V(1)
equals 0.5. The next subsection will describe how the RISE, FALL and CROSS qualifiers can used to return
only a subset of those four crossings.

2.1.19.2. RISE, FALL and CROSS Qualifiers

Xyce supports non-negative values for the RISE, FALL and CROSS qualifiers for all continuous measure
types. It supports negative values for the RISE, FALL and CROSS qualifiers for the DERIV-WHEN, FIND-WHEN
and WHEN measure types. However, their interpretation is slightly different for TRAN and TRAN_CONT
measure modes, as illustrated by the following netlist for the TRAN_CONT measure mode and WHEN measure.
The rules are then the same for the other continuous measures modes and the RISE and FALL qualifiers.

VPWL1 1 0 pwl(0 0 2.5m 1 5m 0 7.5m 1 10m 0)

R1 1 0 1

.TRAN 0 10ms
.PRINT TRAN V(1)

.MEASURE TRAN FindV1_CROSS3 WHEN V(1)=0.5 CROSS=3

.MEASURE TRAN_CONT FindV1_CONT_CROSS3 WHEN V(1)=0.5 CROSS=3

.MEASURE TRAN FindV1_CROSS_NEG3 WHEN V(1)=0.5 CROSS=-3

.MEASURE TRAN_CONT FindV1_CONT_CROSS_NEG3 WHEN V(1)=0.5 CROSS=-3

.END

The <netlistName>.mt0 file will contain the results for both TRAN mode measures. The result for the
FindV1_CROSS3 is the time of the third crossing. The result for the FindV1_CROSS_NEG3 is the time of the
second crossing, which is also the “third to last” (or negative third) crossing in this case.

FINDV1_CROSS3 = 6.250000e-03
FINDV1_CROSS_NEG3 = 3.750000e-03

The <netlistName>_findv1_cont_cross3.mt0 output file will have two values. For non-negative
values of CROSS, a TRAN_CONT measure will return all crossings, starting with the specified value. This is
the third and fourth crossings in this case.

80
FINDV1_CONT_CROSS3 = 6.250000e-03
FINDV1_CONT_CROSS3 = 8.750000e-03

For negative values of CROSS, a TRAN_CONT measure will only return one value. That is the third-to-last
crossing in this case. So, the <netlistName>_findv1_cont_cross3_neg3.mt0 file only has one value
in it. As a final note, a CROSS value of either 5 or -5 would produce failed measures in this example.

FINDV1_CONT_CROSS_NEG3 = 3.750000e-03

2.1.19.3. AT and TD Qualifiers for TRIG-TARG

The following rules apply to the AT and TD qualifiers for TRIG-TARG measures:

• Separate AT values can be given for the TRIG and TARG clauses.
• Separate TD values can be given for the TRIG and TARG clauses.
• The AT value takes precedence over the TD qualifier if both are given in a TRIG or TARG clause.
• If the TD value is only given for the TRIG clause then that value will be used for both the TRIG and
TARG clauses.
• An AT value that is outside of the simulation window, or a TD value that is greater than the end
simulation time or the largest AC, DC or NOISE sweep value, will produce a failed measure.
• A TD value that is less than 0, or the smallest AC, DC or NOISE sweep value, is essentially ignored.

2.1.19.4. HSPICE Compatibility

There are known incompatibilities between the Xyce and HSPICE implementation of continuous measures.
They include the following:

• Xyce will not return a trig or targ value that is outside of the simulation bounds. In some case,
HSPICE will return a trig or targ value that is earlier than the start of the simulation window.
• Xyce does not support negative values for the RISE, FALL or CROSS qualifiers for the continuous
version of the TRIG-TARG measure.

81
2.1.20. .MEASURE FFT (Measure output for .FFT)

The .MEASURE FFT statement allows calculation or reporting of simulation metrics, from data associated
with .FFT analyses, to an external file as well as to the standard output and/or a log file, So, it is only
supported for .TRAN, analyses. It can be used with .STEP. For HSPICE compatibility, .MEAS is an allowed
synonym for .MEASURE.

The syntaxes for the .MEASURE FFT statements are shown below.

General Form .MEASURE FFT <result name> ENOB <variable> [BINSIZ=<value>]

+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE FFT <result name> EQN <variable>

+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE FFT <result name> FIND <variable> AT=<value>

+ [PRECISION=<value>] [PRINT=<value>]

.MEASURE FFT <result name> SFDR <variable>

+ [MINFREQ=<value>] [MAXFREQ=<value>] [BINSIZ=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE FFT <result name> SNDR <variable> [BINSIZ=<value>]

+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE FFT <result name> SNR <variable> [MAXFREQ=<value>]

+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

.MEASURE FFT <result name> THD <variable>

+ [NBHARM=<value>] [MAXFREQ=<value>]
+ [DEFAULT_VAL=<value>] [PRECISION=<value>] [PRINT=<value>]

Examples .FFT V(1) NP=16

.MEASURE FFT ENOBVAL ENOB V(1)
.MEASURE FFT EQNVAL EQN VR1AT2
.MEASURE FFT VR1AT2 VR(1) AT=2
.MEASURE FFT SFDRVAL SFDR V(1)
.MEASURE FFT SNDRVAL SNDR V(1)
.MEASURE FFT SNRVAL SNR V(1)
.MEASURE FFT THDVAL THD V(1)

Arguments and
Options result name
Measured results are reported to the output and log file. Additionally, the
results are stored in files called circuitFileName.mt#, where the suffixed
number (#) starts at 0 and increases for multiple iterations (.STEP iterations)

82
 of a given simulation. Each line of this file will contain the measurement
name, <result name>, followed by its value for that run. The <result
name> must be a legal Xyce character string.
If multiple measures are defined with the same <result name> then Xyce
uses the last such definition, and issues warning messages about (and
discards) any previous measure definitions with the same <result name>.
measure type
ENOB, EQN, FIND, SFDR, SNDR, SNR, THD
The third argument specifies the type of measurement or calculation to be
done. By default, the measurement is performed over the time window
defined by the START and STOP parameters on the associated .FFT line. So,
the FROM, TO and TD qualifiers have no effect on FFT-based measures.
The supported measure types are:
ENOB Calculates the “Effective Number of Bits”, where that metric is
defined in Section 2.1.8.3 which covers the .FFT command.
EQN Calculates the value of <variable> during the simulation. That
variable can use the results of other measure statements. PARAM is an
allowed synonym for EQN as a measure type. For FFT measure mode, an
EQN measure will be reported as “failed” until the associated FFT has
been calculated.
FIND Returns the requested FFT cofficient at the requested frequency.
Examples of the mapping of Xyce operators (e.g., VM and IM) to FFT
cofficients is given in the “Additional Examples” subsection below. FIND
measures can be also used in conjunction with EQN measures to generate
fairly arbitrary FFT-based measures. The FIND measure for FFT measure
mode does not support expressions, or the P and W operators. It also
does not support multi-terminal lead current operators, such as IC().
SFDR Calculates the “Spurious Free Dynamic Range”, where that metric is
defined in Section 2.1.8.3.
SNDR Calculates the “Signal to Noise-plus-Distortion Ratio”, where that
metric is defined in Section 2.1.8.3.
SNR Calculates the “Signal to Noise Ratio”, where that metric is defined in
Section 2.1.8.3.
THD Calculates the “Total Harmonic Distortion”, where that metric is
defined further below and also in in Section 2.1.8.3.
variable
The <variable> is a simulation quantity, such as a voltage or current.
Additionally, the <variable> may be a Xyce expression delimited by { }
brackets. The only constraint is that the variable on the .MEASURE FFT
line must be an exact match for the ov on at least one .FFT line in the netlist.
If there are multiple .FFT lines in the netlist with the same ov then the
corresponding .MEASURE FFT statements will use the first such one.

83
AT=value
A frequency at which the measurement calculation will occur. This is used by
the FIND measure only. The entered AT value will be rounded to the nearest
harmonic frequency, as defined by the FREQ, START and STOP parameters on
the associated .FFT line. An AT value that rounds to a harmonic frequency of
less than zero, or to more than NP/2, will produce a failed measure in Xyce,
where NP is the number of points specified on the associated .FFT line.. The
behavior of these ”failed” cases may differ from commercial simulators.

BINSIZ=value
This parameter is implemented in Xyce for the ENOB, SFDR and SNDR
measure types. It can be used to account for any “broadening” of the spectral
energy in the first harmonic of the signal, as discussed below. BINSIZ has a
default value of 0.
DEFAULT_VAL=value
If the conditions specified for finding a given value are not found during the
simulation then the measure will return the default value of -1 in the
circuitFileName.mt# file. The measure value in the standard output or
log file will be FAILED. The default return value for the
circuitFileName.mt# file is settable by the user for each measure by
adding the qualifier DEFAULT_VAL=<retval> on that measure line. If either
.OPTIONS MEASURE MEASFAIL=<val> or .OPTIONS MEASURE
DEFAULT_VAL=<val> are given in the netlist then those values override the
DEFAULT_VAL parameters given on individual .MEASURE FFT lines. See
Section 2.1.25 for more details.
MAXFREQ=value
The maximum frequency over which to perform a SFDR, SNR or THD
measure. The entered MAXFREQ value will be rounded to the nearest
harmonic frequency, as defined by the FREQ, START and STOP parameters on
the associated .FFT line. The default value is NP/2, where NP is the number
of points specified on the associated .FFT line.

MINFREQ=value
The minimum frequency over which to perform a SFDR or THD measure. The
entered MINFREQ value will be rounded to the nearest harmonic frequency,
as defined by the FREQ, START and STOP parameters on the associated .FFT
line. The default value is 1.
NBHARM=value
The maximum (integer) number of harmonics over which to perform a THD
measure. The default value is NP/2. The NBHARM qualifier has precedence
over the MAXFREQ qualifier if both are given on a .MEASURE line.

PRECISION=value
The default precision for .MEASURE output is 6 digits after the decimal point.
This argument provides a user configurable precision for a given .MEASURE
statement that applies to both the .mt# file and standard output. If .OPTIONS
MEASURE MEASDGT=<val> is given in the netlist then that value overrides
the PRECISION parameters given on individual .MEASURE lines.

84
 PRINT=value
This parameter controls where the .MEASURE output appears. The default is
ALL, which produces measure output in both the .mt# and the standard
output. A value of STDOUT only produces measure output to standard output,
while a value of NONE suppresses the measure output to both the .mt# file
and standard output.

2.1.20.1. Measure Definitions

The ENOB, SNDR, SNR and SFDR measure types use the same definitions as the metrics produced by .FFT
lines. Section 2.1.8.3 provides more details on those definitions. (Note: The MAXFREQ and MINFREQ
qualifiers from the .MEASURE FFT lines are mapped into the FMAX and FMIN parameters used in those
equations.) There are two exceptions.
The first exception is the THD measure. If the optional NBHARM qualifier is not used then the definition given
in Section 2.1.8.3 is used. If the NBHARM qualifier is used then it takes precedence over the MAXFREQ
qualifier. The THD measure definition is then as follow. Let 𝑓0 be the integer index of the “first harmonic”
( 𝑓0 ), as defined in Section 2.1.8.3, from the associated .FFT line. Then the effective value of the upper
frequency limit ( 𝑓2 ) in the THD calculation is NBHARM· 𝑓0 , with the caveat that all of the harmomics will be
used if NBHARM < 0 or NBHARM > NP/2.
The second exception is the BINSIZ qualifier for the ENOB, SFDR and SNDR measures. For a non-zero value
of BINSIZ, the “signal power” is considered to reside in the harmonic indexes between ( 𝑓0 ± BINSIZ),
where the DC value is still excluded from the measure calculations. (Note: This definition for BINSIZ may
differ from HSPICE.)

2.1.20.2. Re-Measure

Xyce can re-calculate (or re-measure) the values for .MEASURE FFT statements using existing Xyce output
files. Section 2.1.18.4 discusses this topic in more detail for both .MEASURE and .FFT statements.

2.1.20.3. Additional Examples

This section provides a simple example how to use the FIND measure, along with the V(), VR(), VI(),
VM(), VP() and VDB() operators, to obtain the real and imaginary parts of the FFT coefficients, along with
the magnitude and phase of those coefficients, at a specified frequency. Those coefficient values are
unnormalized.

* Example of obtaining FFT coefficients

.TRAN 0 1
.PRINT TRAN V(1)
.OPTIONS FFT FFT_ACCURATE=1 FFTOUT=1
V1 1 0 1
R1 1 0 1
.FFT V(1) NP=8 WINDOW=HANN

* Unnormalized one-sided FFT cofficients for V(1) at F=1.0

85
* Magnitude
.MEASURE FFT V1AT1 FIND V(1) AT=1.0
* Real part
.MEASURE FFT VR1AT1 FIND VR(1) AT=1.0
* Imaginary part
.MEASURE FFT VI1AT1 FIND VI(1) AT=1.0
* Magnitude
.MEASURE FFT VM1AT1 FIND VM(1) AT=1.0
* Phase
.MEASURE FFT VP1AT1 FIND VP(1) AT=1.0
* Magnitude in dB
.MEASURE FFT VDB1AT1 FIND VDB(1) AT=1.0

.END

The .MEASURE output is then:

FINDV1AT1 = 5.200051e-01
FINDVR1AT1 = -4.804221e-01
FINDVI1AT1 = -1.989973e-01
FINDVM1AT1 = 5.200051e-01
FINDVP1AT1 = -1.575000e+02
FINDVDB1AT1 = -5.679847e+00

86
2.1.21. .MODEL (Model Definition)

The .MODEL command provides a set of device parameters to be referenced by device instances in the
circuit.

General Form .MODEL <model name> <model type> (<name>=<value>)*

Examples .MODEL RMOD R (RSH=1)

.MODEL MOD1 NPN BF=50 VAF=50 IS=1.E-12 RB=100 CJC=.5PF TF=.6NS
.MODEL NFET NMOS(LEVEL=1 KP=0.5M VTO=2V)

Arguments and
Options model name
The model name used to reference the model.
model type
The model type used to define the model. This determines if the model is
(for example) a resistor, or a MOSFET, or a diode, etc. For transistors, there
will usually be more than one type possible, such as NPN and PNP for BJTs,
and NMOS and PMOS for MOSFETs.
name
value
The name of a parameter and its value. Most models will have a list of
parameters available for specification. Those which are not set will receive
default values. Most will be floating point numbers, but some can be integers
and some can be strings, depending on the definition of the model.

Comments The scoping rules for models are:

• If a .MODEL, statement is included in the main circuit netlist, then it is accessible

from the main circuit and all subcircuits.

• .MODEL statements defined within a subcircuit are scoped to that subciruit

definition. So, their models are only accessible within that subcircuit definition,
as well as within “nested subcircuits” also defined within that subcircuit
definition.

Additional illustative examples of scoping are given in the “Working with Subcircuits
and Models” section of the Xyce Users’ Guide [1] .

A model name can be the same as a device name in Xyce. However, that usage will
generate a warning message during netlist parsing. The reason is that it can lead to
ambiguous .PRINT lines when a model parameter and instance parameter, for a given
device, have the same name but a different meaning. For example, R1 could be used as
both a resistor device-name, and as a resistor model-name. However, .PRINT TRAN
R1:R would then be ambiguous. In addition, the use of duplicate model and device
names is not recommended if those names will be used within a Xyce expression since
that can result in an ambiguous expression.

87
2.1.21.1. LEVEL Parameter

A common parameter is the LEVEL parameter, which is set to an integer value. This parameter will define
exactly which model of the given type is to be used. For example, there are many different available
MOSFET models. All of them will be specified using the same possible names and types. The way to
differentiate (for example) between the BSIM3 model and the PSP model is by setting the appropriate
LEVEL.

2.1.21.2. Model Binning

Model binning is supported for MOSFET models. For model binning, the netlist contains a set of similar
.MODEL cards which correspond to different sizing information (length and width). They are similar in that
they are for the same model (and same LEVEL number), and have the same prefix. They are different in that
they have different lmin,lmax,wmin,wmax parameters, and the name suffix will be the bin number. For a
MOSFET device instance, Xyce will automatically select the appropriate binned model, based on the L and
W parameters of that instance. It will only seach the models with matching name prefixes, and can only work
if all the binned models have specified all the lmin,lmax,wmin,wmax parameters.

Model binning is enabled by default. To disable it, specify .options parser model_binning=false.

* Model binning example adapted from the BSIM4 tests

m1 2 1 0 b nch L=0.11u W=10.1u NF=5 rgeomod=1 geomod=0
vgs 1 0 1.2
vds 2 0 1.2
Vb b 0 0.0

.dc vds 0.0 1.21 0.02 vgs 0.2 1.21 0.2

.print dc v(2) v(1) i(vds)

* model binning
.model nch.1 nmos(level=14
+ lmin=0.1u lmax=20u
+ wmin=0.1u wmax=10u)
.model nch.2 nmos(level=14
+ lmin=0.1u lmax=20u
+ wmin=10u wmax=100u)

.end

Figure 2-1. Model Binning Example

2.1.21.3. Length Scaling

It can be convenient to specify the lengths and widths for MOSFET instances in scaled units. To enable this,
the netlist should include .options parser scale or .option scale. This feature is only supported in

88
MOSFET compact models. An example usage is given in figure 2-2. In this example, the scaled length and
width for transistor mn1 is l=5.0e-6 and w=175e-6.

mos level 1 model cmos inverter using scale]

.tran 20ns 6us
.print tran v(vout) v(in) v(1)
vdddev vdd 0 5v
rin in 1 1k
vin1 1 0 5v pulse (5v 0v 1.5us 5ns 5ns 1.5us 3us)
r1 vout 0 10k
c2 vout 0 0.1p
mn1 vout in 0 0 cd4012_nmos l=5 w=175
mp1 vout in vdd vdd cd4012_pmos l=5 w=270

.options parser scale=1.0e-6

* also valid:
*.option scale=1.0e-6

.model cd4012_pmos pmos (level=1 uo=310 vto=-1.6 tox=6e-08)

.model cd4012_nmos nmos (level=1 uo=190 vto=1.679 tox=6e-08)
.end

Figure 2-2. Scale Example

89
2.1.22. .NODESET (Approximate Initial Condition, Bias point)

The .NODESET command sets initial conditions for operating point calculations. It is similar to .IC
(Section 2.1.14), except it is applied as an initial guess, rather than as a firmly enforced condition. Like .IC,
.NODESET initial conditions can be specified for some or all of the circuit nodes.
Consult the Xyce Users’ Guide [1] for more guidance.

General Form .NODESET < V(<node>)=<value>

.NODESET <node> <value>

Examples .NODESET V(2)=3.1

.NODESET 2 3.1

Comments The Xyce .NODESET command uses a different strategy than either SPICE or HSPICE.
When .NODESET is specified, Xyce does two solves for the DC operating point. One
with the .NODESET values held as initial conditions (i.e., the same as if it was an .IC
solve). The second solve is then done without any conditions imposed, but with the
first solution as an initial guess.
The .NODESET capability can only set voltage values, not current values.
The .NODESET capability can not be used, within subcircuits, to set voltage values on
global nodes.

90
2.1.23. .NOISE (Noise Analysis)

Calculates the the small signal noise response of a circuit over a range of frequencies. The .NOISE
command can specify a linear sweep, decade logarithmic sweep, octave logarithmic sweep, or a data table
of multivariate values.

General Form .NOISE V(OUTPUT <, REF>) SRC <sweep type> <points value>
+ <start frequency value> <end frequency value>

Examples .NOISE V(5) VIN LIN 101 100Hz 200Hz

.NOISE V(5,3) V1 OCT 10 1kHz 16kHz
.NOISE V(4) V2 DEC 20 1MEG 100MEG
.NOISE V(4) V2 DATA=<table name>

Arguments and
Options V(OUTPUT <,REF>)
The node at which the total output noise is desired. If REF is specified, then
the noise voltage V(OUTPUT) - V(REF) is calculated. By default, REF is
assumed to be ground.

SRC The name of an independent source to which input noise is referred.

sweep type
Must be LIN, OCT, or DEC, as described below.
LIN Linear sweep
The sweep variable is swept linearly from the starting to the ending
value.
OCT Sweep by octaves
The sweep variable is swept logarithmically by octaves.
DEC Sweep by decades
The sweep variable is swept logarithmically by decades.
DATA Sweep values from a table
Sweep variables are applied based on the rows of a data table. This
format allows magnitude and phase to be swept in addition to frequency.
If using this format, then the V(OUTPUT <,REF>) and SRC arguments
are still needed on the .NOISE line.
points value
Specifies the number of points in the sweep, using an integer greater than or
equal to 1.
start frequency value
end frequency value
The end frequency value must not be less than the start frequency value, and
both must be greater than zero. The whole sweep must include at least one
point.

91
Comments Noise analysis is a linear analysis. The simulator calculates the noise response by
linearizing the circuit around the bias point.
If specifying the sweep points using the DATA type, one can also sweep the magnitude
and phase of an AC source, as well as the values of linear model parameters. However,
unlike the use of DATA for .STEP and .DC, it is not possible to sweep nonlinear device
parameters. This is because changing other nonlinear device parameters would alter
the correct DCOP solution, and the NOISE sweep happens after the DCOP calculation
in the analysis flow. To sweep a nonlinear device parameter on a NOISE problem, add
a .STEP command to the netlist to provide an outer parametric sweep around the
analysis.
If .DATA is used with .NOISE then the integrals for the total input noise and total
output noise will only be calculated, and sent to stdout, if the frequencies in the data
table are monotonically increasing.
A .PRINT NOISE must be used to get the results of the NOISE sweep analysis. See
Section 2.1.31.
Noise analysis is a relatively new feature to Xyce, so not all noise models have been
supported.
Power calculations (P(<device> and W(<device>) are not supported for any devices
for noise analysis. Current variables (e.g., I(<device>) are only supported for
devices that have “branch currents” that are part of the solution vector. This includes
the V, E, H and L devices. It also includes the voltage-form of the B device.

92
2.1.24. .OP (Bias Point Analysis)

The .OP command causes detailed information about the bias point to be printed.

General Form .OP

Comments This type of analysis can be specified by itself, in which case Xyce will run a nominal
operating point. However, if specified with another analysis type, no additional
operating point will be calculated, as most analyses require a DC operating point for
initialization.
.OP outputs the parameters for all the device models and all the device instances
present in the circuit. For large circuits, this can be a very large amount of output, so
use with caution.
If no analysis command is provided, .OP will run a DC Operating Point calculation
(i.e., a DC analysis) with all the voltage sources left at their nominal (instance line)
values.
The Xyce .OP statement may provide less, or different, output than other simulators.
For some of the missing quantities, a Xyce .PRINT line can give similar information.
Nodal voltages are always available on a .PRINT line. Device currents for many
devices are available on a .PRINT line using the lead current notation
(I(devicename)). Similarly, device power is available on a .PRINT line via
P(devicename) or W(devicename). However, these capabilities are not supported in
all devices. Table 2-36 shows which devices support these lead current and power
notations. Currently, there is no way to print out internal capacitances.

93
2.1.25. .OPTIONS Statements

Set various simulation limits, analysis control parameters and output parameters. In general, they use the
following format:

General Form .OPTIONS <pkg> [<name>=<value>]*

Examples .OPTIONS TIMEINT ABSTOL=1E-8

Arguments and
Options
pkg DEVICE Device Model
DIAGNOSTIC Diagnostic Simulation Output
TIMEINT Time Integration
NONLIN Nonlinear Solver
NONLIN-TRAN Transient Nonlinear Solver
NONLIN-HB HB Nonlinear Solver
LOCA Continuation/Bifurcation Tracking
LINSOL Linear Solver
LINSOL-HB HB Linear Solver
LINSOL-AC AC Linear Solver
OUTPUT Output
RESTART Restart
SAMPLES Sampling analysis and non-intrusive Polynomial
Chaos (PCE)
EMBEDDEDSAMPLES EmbeddedSampling and non-intrusive Polynomial
Chaos (PCE)
PCES Fully intrusive Polynomial Chaos (PCE)
SENSITIVITY Direct and Adjoint sensitivities
HBINT Harmonic Balance (HB)
DIST Distribution
MEASURE Measure
PARSER Parsing
name
value
The name of the parameter and the value it will be assigned.

94
Comments Exceptions to this format are the OUTPUT and RESTART options, which use their own
format. They are defined under their respective descriptions.
The designator pkg refers loosely to a module in the code. For most designators,
multiple .OPTIONS statements are allowed. So for example, the netlist may contain
multiple .OPTIONS DEVICE commands, and Xyce will parse and apply all of them. If
any parameters are specified more than once, Xyce will issue a warning. The warning
will include which parameter value is being used.
The OUTPUT and RESTART options have a different enough format that they can only
be specified once. Extra OUTPUT and RESTART option statements (beyond the first one)
will be ignored.
.OPTIONS statements are only permitted in the top level of the netlist. Any .OPTIONS
statemets inside of subcircuits will result in a warning, and otherwise will be ignored.

2.1.25.1. .OPTIONS DEVICE (Device Package Options)

The device package parameters listed in Table 2-3 outline the options available for specifying device
specific parameters. Some of these (DEFAS, DEFAD, TNOM etc.) have the same meaning as they do for the
.OPTION line from Berkeley SPICE (3f5). Parameters which apply globally to all device models will be
specified here. Parameters specific to a particular device instance or model are specified in section 2.3.

Table 2-3. Options for Device Package

Option Description Default
DEFAD MOS Drain Diffusion Area 0.0
DEFAS MOS Source Diffusion Area 0.0
DEFL MOS Default Channel Length 1.0E-4
DEFW MOS Default Channel Width 1.0E-4
This option controls the behavior of the Digital Latch
(DLTCH), D Flip-Flop (DFF), JK Flip-Flop (JKFF) and T
DIGINITSTATE Flip-Flop (TFF) behavioral digital devices during the DC 3
Operating Point (DCOP) calculations. See 2.3.28 for more
details.
GMIN Minimum Conductance 1.0E-12
This is a minimum resistance to be used in place of the default
zero value of semiconductor device internal resistances. It is
MINRES only used when model specifications (.MODEL cards) leave the 0.0
parameter at its default value of zero, and is not used if the
model explicitly sets the resistance to zero.
This is a minimum capacitance to be used in place of the default
zero value of semiconductor device internal capacitances. It is
MINCAP only used when model specifications (.MODEL cards) leave the 0.0
parameter at its default value of zero, and is not used if the
model explicitly sets the capacitance to zero.

95
 Table 2-3. Options for Device Package
Option Description Default
27.0 ◦ C
TEMP Temperature
(300.15K)
27.0 ◦ C
TNOM Nominal Temperature
(300.15K)
NUMJAC Numerical Jacobian flag (only use for small problems) 0 (FALSE)
VOLTLIM Voltage limiting 1 (TRUE)
BSIMSOI3 Voltage limiting. This flag is similar to VOLTLIM,
except that it only applies to the BSIMSOI version 3 (the newer
versions of the BSIM SOI do not have voltage limiting).
B3SOIVOLTLIM 1 (TRUE)
Turning this off will often improve numerical robustness.
Unlike VOLTLIM, turning this off does not disable the initial
condition code in the BSIMSOI model.
This is a multiplicative factor which is applied to right-hand
icFac side vector loads of .IC initial conditions during the DCOP 10000.0
phase.
MAXTIMESTEP Maximum time step size 1.0E+99
This flag enables smooth transitions by adding a RC network to
SMOOTHBSRC 0
the output of ABM devices
This option controls the smoothness of the transitions if the
RCCONST SMOOTHBSRC flag is enabled. This is done by specifying the 1e-9
RC constant of the RC network

MOSFET Homtopy parameters

VDSSCALEMIN Scaling factor for Vds 0.3
VGSTCONST Initial value for Vgst 4.5 Volt
LENGTH0 Initial value for length 5.0e-6
WIDTH0 Initial value for width 200.0e-6
TOX0 Initial value for oxide thickness 6.0e-8

Debug output parameters

DEBUGLEVEL The higher this number, the more info is output 1
DEBUGMINTIMESTEP First time-step debug information is output 0
DEBUGMAXTIMESTEP Last time-step of debug output 65536
Same as DEBUGMINTIMESTEP except controlled by time (sec.)
DEBUGMINTIME 0.0
instead of step number
Same as DEBUGMAXTIMESTEP except controlled by time (sec.)
DEBUGMAXTIME 100.0
instead of step number

96
2.1.25.2. .OPTIONS DIAGNOSTIC (Diagnostic Simulation Output)

This option enables the output of diagnostic data during a simulation to aid in debugging circuit problems.
There are three optional diagnostics that one can specify, Extrema Limit, Voltage Limit, Current Limit and
a Discontinuity limit. Their behavior is as follows:
• EXTREMALIMIT=value If value is given then Xyce will output the node name and absolute value of
the solution vector extrema when it exceeds value at each step.
• VOLTAGELIMIT=value If value is given then Xyce will output the voltage node names and values that
exceed ±value at each step.
• CURRENTLIMIT=value If value is given then Xyce will output the branch and lead current names and
values that exceed ±value at each step.
• DISCLIMIT=value If value is given then Xyce will output solution variables where the absolute value
of the difference between the solution and the predictor (from transient history) exceed value.
Note that EXTREMALIMIT will only output one name and value per step while VOLTAGELIMIT,
CURRENTLIMIT and DISCLIMIT can output many names and values per step. Xyce’s diagnostic output will
be sent to the diagnostic file set by DIAGFILENAME along with step information (time step in transient
simulation and step number otherwise) and the status of GMIN or source-stepping used to obtain the DC
operating point. If no filename is specified, then the default output file name will be the input netlist name
plus the extenstion .dia. A user can examine the output to infer if the circuit simulation is performing as
expected. Further examples of using diagnostic options can be found in chapter 10 of the Xyce User’s
Guide.

Table 2-4. Options for Diagnostic Output

Option Description Default
EXTREMA Output extreme values occurring in the solution vector true
Output the absolute value of extrema that exceed
EXTREMALIMIT 0.0
EXTREMALIMIT
VOLTAGELIMIT Output voltages that exceed ±VOLTAGELIMIT 0.0
Output branch and lead currents that that exceed
CURRENTLIMIT 0.0
±CURRENTLIMIT
Output solution variables where the absolute value of the
DISCLIMIT difference between the solution and the transient predictor 0.0
exceed ±DISCLIMIT
[input
DIAGFILENAME Save diagnostic data to a file named by DIAFILENAME
netlist].dia

2.1.25.3. .OPTIONS TIMEINT (Time Integration Options)

The time integration parameters listed in Table 2-5 give the available options for helping control the time
integration algorithms for transient analysis.
Time integration options are set using the .OPTIONS TIMEINT command.

97
 Table 2-5. Options for Time Integration Package.
Option Description Default
Time integration method. This parameter is only relevant when
running Xyce in transient mode. Supported methods: trap or 7
(variable
METHOD • trap or 7 (variable order Trapezoid)
order
• gear or 8 (Gear method) Trapezoid)

RELTOL Relative error tolerance 1.0E-03

ABSTOL Absolute error tolerance 1.0E-06
This parameter is a scalar which determines how small the
initial time step out of a breakpoint should be. In the current
RESTARTSTEPSCALE version of the time integrator, the first step after a breakpoint 0.005
isn’t subjected to much error analysis, so for very stiff circuits,
this step can be problematic.
This flag sets if “soft” failures of the nonlinear solver, when the
convergence criteria are almost, but not quite, met, should
result in a "success" code being returned from the nonlinear
solver to the time integrator. If this is enabled, it is expected
NLNEARCONV 0 (FALSE)
that the error analysis performed by the time integrator will be
the sole determination of whether or not the time step is
considered a “pass” or a “fail”. This is on by default, but
occasionally circuits need tighter convergence criteria.
This flag is another “soft” nonlinear solver failure flag. In this
case, if the flag is set, time steps in which the nonlinear solver
stalls, and is using updates that are numerically tiny, can be
NLSMALLUPDATE 1 (TRUE)
considered to have converged by the nonlinear solver. If this
flag is set, the time integrator is responsible for determining if a
step should be accepted or not.
The nonlinear solver resets its settings for the transient part of
the run to something more efficient (basically a simpler set of
RESETTRANNLS options with smaller numbers for things like max Newton step). 1 (TRUE)
If this is set to false, this resetting is turned off. Normally
should be left as default.
This parameter determines the maximum order of integration
that time integrators will attempt. Setting this option does not 2 for variable
guarantee that the integrator will integrate at this order, it just order
MAXORD
sets the maximum order the integrator will attempt. In order to Trapezoid
guarantee a particular order is used, see the option MINORD and Gear
below.
This parameter determines the minimum order of integration
that time integrators will attempt to maintain. The integrator
will start at Backward Euler and move up in order as quickly as
MINORD possible to achive MINORD and then it will keep the order above 1
this. If MINORD is set at 2 and MAXORD is set at 2, then the
integrator will move to second order as quickly as possible and
stay there.

98
 Table 2-5. Options for Time Integration Package.
Option Description Default
This parameter determines the reference value for relative
convergence criterion in the local truncation error based time
step control. The supported choices
• 0. The reference value is the current value on each node.
• 1. The reference value is the maximum of all the signals
NEWLTE at the current time. 1
• 2. The reference value is the maximum of all the signals
over all past time.
• 3. The reference value is the maximum value on each
signal over all past time.

This flag sets a new time stepping method after a break point.
Previously, Xyce treats each breakpoint identically to the
DCOP point, in which the intitial time step out of the DCOP is
made to be very very small, because the LTE calculation is
unreliable. As a result, Xyce takes an incredibly small step out
of each breakpoint and then tries to grow the stepsize from
there. When NEWBPSTEPPING is set, Xyce can take a reasonable
NEWBPSTEPPING large step out of every non-DCOP breakpoint, and then just 1 (TRUE)
relies on the step control to ensure that the step is small enough.
Note that the new time stepping method after a break point
does not work well with the old LTE calculation since the old
LTE calculation is conservative and it tends to reject the first
time step out of a break point. We recommend to use newlte if
you choose to use the new time stepping method out of a break
point.
This parameter masks out current variables in the local
MASKIVARS 0 (FALSE)
truncation error (LTE) based time step control.
This parameter determines if Local Truncation Error (LTE)
control is turned on or not. If ERROPTION is on, then step-size
selection is based on the number of Newton iterations
nonlinear solve. For Trapezoid and Gear, if the number of
nonlinear iterations is below NLMIN then the step is doubled. If
0 (Local
the number of nonlinear iterations is above NLMAX then the step
ERROPTION Truncation
is cut by one eighth. In between, the step-size is left alone.
Error is used)
Because this option can lead to very large time-steps, it is very
important to specify an appropriate DELMAX option. If the
circuit has breakpoints, then the option MINTIMESTEPSBP can
also help to adjust the maximum time-step by specifying the
minimum number of time points between breakpoints.
This parameter determines the lower bound for the desired
NLMIN number of nonlinear iterations during a Trapezoid time or Gear 3
integration solve with ERROPTION=1.
This parameter determines the upper bound for the desired
NLMAX number of nonlinear iterations during a Trapezoid time or Gear 8
integration solve with ERROPTION=1.

99
 Table 2-5. Options for Time Integration Package.
Option Description Default
This parameter determines the maximum time step-size used
with ERROPTION=1. If a maximum time-step is also specified
DELMAX 1e99
on the .TRAN line, then the minimum of that value and DELMAX
is used.
This parameter determines the minimum number of time-steps
to use between breakpoints. This enforces a maximum
MINTIMESTEPSBP time-step between breakpoints equal to the distance between 10
the last breakpoint and the next breakpoint divided by
MINTIMESTEPSBP.
This parameter determines whether time-steps are rejected
based upon the step-size selection strategy in ERROPTION=1. If
it is set to 0, then a step will be accepted with successful
nonlinear solves independent of whether the number of
0 (do not
TIMESTEPSREVERSAL nonlinear iterations is between NLMIN and NLMAX. If it is set to
reject steps)
1, then when the number of nonlinear iterations is above
NLMAX, the step will be rejected and the step-size cut by one
eighth and retried. If ERROPTION=0 (use LTE) then
TIMESTEPSREVERSAL=1 (reject steps) is set.
Default
value, for
TCAD
circuits, is a
TCAD devices by default will solve an extra "setup" problem to
combination:
mitigate some of the convergence problems that TCAD devices
nl_poisson,
often exhibit. This extra setup problem solves a nonlinear
drift_diffusion.
Poisson equation first to establish an initial guess for the full
Default
drift-diffusion(DD) problem. The name of this parameter refers
value, for
to the fact that the code is solving two DC operating point steps
DOUBLEDCOPSTEP non-TCAD
instead of one. To solve only the nonlinear Poisson problem,
circuits is a
then set DOUBLEDCOP=nl_poisson. To solve only the
moot point.
drift-diffusion problem (skipping the nonlinear Poisson), set
If no TCAD
DOUBLEDCOP=drift_diffusion. To explicitly set the default
devices are
behavior, then set DOUBLEDCOP=nl_poisson,
present in the
drift_diffusion.
circuit, then
there will not
be an extra
DCOP solve.
This parameter specifies a comma-separated list of timepoints
that should be used as breakpoints. They do not replace the
BREAKPOINTS N/A
existing breakpoints already being set internally by Xyce, but
instead will add to them.
Flag for turning on/off breakpoints (1 = ON, 0 = OFF). It is
BPENABLE unlikely anyone would ever set this to FALSE, except to help 1 (TRUE)
debug the breakpoint capability.

100
 Table 2-5. Options for Time Integration Package.
Option Description Default
If this is set to nonzero, the code will check the simulation time
at the end of each step. If the total time exceeds the exittime,
the code will ungracefully exit. This is a debugging option, the
EXITTIME -
point of which is the have the code stop at a certain time during
a run without affecting the step size control. If not set by the
user, it isn’t activated.
Same as EXITTIME, only applied to step number. The code will
EXITSTEP -
exit at the specified step. If not set by the user, it isn’t activated.

2.1.25.4. .OPTIONS NONLIN (Nonlinear Solver Options)

The nonlinear solver parameters listed in Table 2-6 provide methods for controlling the nonlinear solver for
DC, transient and harmonic balance. Note that the nonlinear solver options for DCOP, transient and
harmonic balance are specified in separate options statements, using .OPTIONS NONLIN, .OPTIONS
NONLIN-TRAN and .OPTIONS NONLIN-HB, respectively. The defaults for .OPTIONS NONLIN and
.OPTIONS NONLIN-TRAN are specified in the third and fourth columns of Table 2-6. The defaults for
.OPTIONS NONLIN-HB are the same as the default settings given for NONLIN-TRAN with two exceptions.
For NONLIN-HB, the default for ABSTOL is 1e-9 and the default for RHSTOL is 1e-4.

Table 2-6. Options for Nonlinear Solver Package.

NONLIN-
NONLIN
Option Description TRAN
Default
Default
NOX Use NOX nonlinear solver. 1 (TRUE) 0 (FALSE)
Nonlinear solution strategy. Supported Strategies:
• 0 (Newton)
NLSTRATEGY • 1 (Gradient) 0 (Newton) 0 (Newton)

• 2 (Trust Region)

Line-search method used by the nonlinear solver. Supported

line-search methods:
• 0 (Full Newton - no line search)
• 1 (Interval Halving) 0 (Full New- 0 (Full
SEARCHMETHOD
• 2 (Quadratic Interpolation) ton) Newton)

• 3 (Cubic Interpolation)
• 4 (More’-Thuente)

101
 Table 2-6. Options for Nonlinear Solver Package.
NONLIN-
NONLIN
Option Description TRAN
Default
Default
Enables the use of Homotopy/Continuation algorithms for the
nonlinear solve. Options are:
• 0 (Standard nonlinear solve)
• 1 (Natural parameter homotopy. See LOCA options list)
0 (Standard 0 (Standard
• 2/mos (Specialized dual parameter homotopy for
CONTINUATION nonlinear nonlinear
MOSFET circuits)
solve) solve)
• 3/gmin (GMIN stepping, similar to that of SPICE)
• 34/sourcestep (Simultaneous source stepping)
• 35/sourcestep2 (Sequential source stepping)

ABSTOL Absolute residual vector tolerance 1.0E-12 1.0E-06

RELTOL Relative residual vector tolerance 1.0E-03 1.0E-02
Weighted nonlinear-solution update norm convergence
DELTAXTOL 1.0 0.33
tolerance
RHSTOL Residual convergence tolerance (unweighted 2-norm) 1.0E-06 1.0E-02
Minimum acceptable norm for weighted nonlinear-solution
SMALLUPDATETOL 1.0E-06 1.0E-06
update
MAXSTEP Maximum number of Newton steps 200 20
MAXSEARCHSTEP Maximum number of line-search steps 2 2
IN_FORCING Inexact Newton-Krylov forcing flag 0 (FALSE) 0 (FALSE)
Sets the minimum allowed linear solver tolerance. Valid only if
AZ_TOL 1.0E-12 1.0E-12
IN_FORCING=1.
If using a line search, this option determines the type of step to
take if the line search fails. Supported strategies:
• 0 (Take the last computed step size in the line search
RECOVERYSTEPTYPE 0 0
algorithm)
• 1 (Take a constant step size set by RECOVERYSTEP)

RECOVERYSTEP Value of the recovery step if a constant step length is selected 1.0 1.0
DEBUGLEVEL The higher this number, the more info is output 1 1
DEBUGMINTIMESTEP First time-step debug information is output 0 0
DEBUGMAXTIMESTEP Last time-step of debug output 99999999 99999999
Same as DEBUGMINTIMESTEP except controlled by time (sec.)
DEBUGMINTIME 0.0 0.0
instead of step number
Same as DEBUGMAXTIMESTEP except controlled by time (sec.)
DEBUGMAXTIME 1.0E+99 1.0E+99
instead of step number

102
2.1.25.5. .OPTIONS LOCA (Continuation and Bifurcation Tracking Package Options)

The continuation selections listed in Table 2-7 provide methods for controlling continuation and bifurcation
analysis. These override the default settings that were given in the continuation package. This option block
is only used if the nonlinear solver or transient nonlinear solver enable continuation through the
CONTINUATION flag.
There are several specialized homotopy methods that can be set in the nonlinear solver options line,
.options nonlin. One is MOSFET-based homotopy, which is specific to MOSFET circuits. This is
specified using continuation=2 or continuation=mos. Another is GMIN stepping, which is specified
using continuation=3 or continuation=gmin. Lastly, source stepping can be done simultaneously,
using continuation=34 or continuation=sourcestep, or sequentially using continuation=35 or
continuation=sourcestep2. For any of these methods, while it is possible to modify their default
LOCA options, it is generally not necessary to do so. Note that Xyce automatically attempts GMIN stepping
if the inital attempt to find the DC operating point fails. In addition, Xyce will attempt simultaneous source
stepping if GMIN fails to find a DC operating point. If any of the specialized homotopy methods are
specified in the netlist, Xyce will attempt to find a DC operating point only using that method.
LOCA options are set using the .OPTIONS LOCA command.

Table 2-7. Options for Continuation and Bifurcation Tracking Package.

Option Description Default
Stepping algorithm to use:
• 0 (Natural or Zero order continuation)
STEPPER 0 (Natural)
• 1 (Arc-length continuation)

Predictor algorithm to use:

• 0 (Tangent)
• 1 (Secant)
PREDICTOR 0 (Tangent)
• 2 (Random)
• 3 (Constant)

Algorithm used to adjust the step size between continuation

steps:
STEPCONTROL • 0 (Constant) 0 (Constant)
• 1 (Adaptive)

CONPARAM Parameter in which to step during a continuation run VA:V0

INITIALVALUE Starting value of the continuation parameter 0.0
MINVALUE Minimum value of the continuation parameter -1.0E20
MAXVALUE Maximum value of the continuation parameter 1.0E20
BIFPARAM Parameter to compute during bifurcation tracking runs VA:V0
MAXSTEPS Maximum number of continuation steps (includes failed steps) 20

103
 Table 2-7. Options for Continuation and Bifurcation Tracking Package.
Option Description Default
Maximum number of nonlinear iterations allowed (set this
MAXNLITERS parameter equal to the MAXSTEP parameter in the NONLIN 20
option block
INITIALSTEPSIZE Starting value of the step size 1.0
MINSTEPSIZE Minimum value of the step size 1.0E20
MAXSTEPSIZE Maximum value of the step size 1.0E-4
Value between 0.0 and 1.0 that determines how aggressive the
AGGRESSIVENESS step size control algorithm should be when increasing the step 0.0
size. 0.0 is a constant step size while 1.0 is the most aggressive.
If set to a nonzero (small) number, this parameter will force the
GMIN stepping algorithm to stop and declare victory once the
artificial resistors have a conductance that is smaller than this
RESIDUALCONDUCTANCEnumber. This should only be used in transient simulations, and 0.0
ONLY if it is absolutely necessary to get past the DC operating
point calculation. It is almost always better to fix the circuit so
that residual conductance is not necessary.

2.1.25.6. .OPTIONS LINSOL (Linear Solver Options)

Xyce uses both sparse direct solvers as well as Krylov iterative methods for the solution of the linear
equations generated by Newton’s method. For the advanced users, there are a variety of options that can be
set to help improve these solvers. Transformations of the linear system have a “TR_” prefix on the flag.
Many of the options for the Krylov solvers are simply passed through to the underlying Trilinos/AztecOO
solution settings and thus have an “AZ_” prefix on the flag.
Linear solver options are set using the .OPTIONS LINSOL command.

Table 2-8. Options for Linear Solver Package.

Option Description Default
Determines which linear solver will be used.
• KLU
• KSparse
KLU (Serial,
• SuperLU (optional) Parallel
• AztecOO < 104
unknowns)
type • Belos AztecOO,
• ShyLU (optional) (Parallel,
≥ 104
Note that while KLU, KSparse, and SuperLU (optional) are unknowns)
available for parallel execution they will solve the linear system
in serial. Therefore they will be useful for moderate problem
sizes but will not scale in memory or performance for large
problems

104
 Table 2-8. Options for Linear Solver Package.
Option Description Default
Determines which preconditioner will be used with an iterative
linear solver
Ifpack (If-
prec_type • Ifpack
pack_IlukGraph)
A preconditioner will not be used if a direct solver (KLU,
KSparse, SuperLU) is specified.
Triggers use of native AztecOO preconditioners for the
use_aztec_precond 0 (FALSE)
iterative linear solves
Use Ifpack factory to create preconditioner instead of using
use_ifpack_factory 0 (FALSE)
Ifpack_IlukGraph
If using the Ifpack factory, use_ifpack_factory=1, which
preconditioner is created
• Amesos (Additive Schwarz w/ KLU subdomain solve)
ifpack_type Amesos
• ILU
• ILUT

Relative dropping threshold for Schur complement

ShyLU_rthresh 1.0E-03
preconditioner (ShyLU only)
Transformation parameters
0 (NONE,
Serial)
TR_partition Perform load-balance partitioning on the linear system
1 (Isorropia,
Parallel)
TR_partition_type Type of load-balance partitioning on the linear system HYPERGRAPH
0 (FALSE,
Serial) 1
TR_singleton_filterTriggers use of singleton filter for linear system
(TRUE,
Parallel)
0 (FALSE,
Triggers use of approximate minimum-degree (AMD) ordering Serial)
TR_amd
for linear system 1 (TRUE,
Parallel)
Triggers use of block triangular form (BTF) ordering for linear
TR_global_btf 0 (FALSE)
system, requires TR_amd=0 and TR_partition=0
Reindexes linear system parallel global indices in lexigraphical
TR_reindex 1 (TRUE)
order, recommended with singleton filter
Triggers remapping of column indices for parallel runs,
TR_solvermap 1 (TRUE)
recommended with singleton filter
Iterative linear solver parameters
Triggers use of AztecOO adaptive solve algorithm for
adaptive_solve 0 (FALSE)
preconditioning of iterative linear solves
AZ_max_iter Maximum number of iterative solver iterations 200

105
 Table 2-8. Options for Linear Solver Package.
Option Description Default
AztecOO iterative solver preconditioner flag (used only when AZ_dom_decomp
AZ_precond
use_aztec_precond=1) (14)
AZ_gmres
AZ_solver Iterative solver type
(1)
AZ_conv Convergence type AZ_r0 (0)
AZ_recalc
AZ_pre_calc Type of precalculation
(1)
AZ_true
AZ_keep_info Retain calculation info
(1)
AZ_modified
AZ_orthog Type of orthogonalization
(1)
AZ_ilut
AZ_subdomain_solve Subdomain solution for domain decomposition preconditioners
(9)
AZ_ilut_fill Approximate allowed fill-in factor for the ILUT preconditioner 2.0
Specifies drop tolerance used in conjunction with LU or ILUT
AZ_drop 1.0E-03
preconditioners
AZ_none
AZ_reorder Reordering type
(0)
AZ_none
AZ_scaling Type of scaling
(0)
AZ_kspace Maximum size of Krylov subspace 50
AZ_tol Convergence tolerance 1.0E-9
AZ_output Output level AZ_none (0)
50 (if
verbose
build)
AZ_diagnostics Diagnostic information level AZ_none (0)
AZ_overlap Schwarz overlap level for ILU preconditioners 0
AZ_rthresh Diagonal shifting relative threshold for ILU preconditioners 1.0001
AZ_athresh Diagonal shifting absolute threshold for ILU preconditioners 1.0E-04
Write out linear system matrix and right-hand-side vector,
output_ls 0 (no output)
post-transformation, to Matrix Market file every # solves
Write out linear system matrix and right-hand-side vector,
output_base_ls 0 (no output)
pre-transformation, to Matrix Market file every # solves
Write out linear system matrix and right-hand-side vector to
output_failed_ls Matrix Market file every # solves when linear solver fails (only 0 (no output)
available for direct solvers)

106
2.1.25.7. .OPTIONS LINSOL-HB (Linear Solver Options)

For harmonic balance (HB) analysis, Xyce provides both iterative and direct methods for the solution of the
steady state. Only matrix-free techniques are available for preconditioning the HB Jacobian with an
iterative linear solver. The direct linear solver explicitly forms the HB Jacobian and solves the
complex-valued linear system with the requested solver. For HB analysis, a reduced number of linear solver
options are available, and are set using the .OPTIONS LINSOL-HB command.

Table 2-9. Options for Linear Solver Package for HB.

Option Description Default

Determines which linear solver will be used

• AztecOO
type • Belos AztecOO

• Direct

Determines which preconditioner will be used with an iterative

linear solver
prec_type block_jacobi
• block_jacobi
A preconditioner will not be used if type=Direct is specified
Enable one-step correction to the block_jacobi
block_jacobi_corrected 0 (FALSE)
preconditioner.
Determines which direct linear solver will be used if
type=Direct is specified
direct_solver LAPACK
• LAPACK

AZ_kspace Maximum size of Krylov subspace 50

AZ_max_iter Maximum number of iterative solver iterations 200
AZ_tol Convergence tolerance 1.0E-9
Write out linear system matrix and right-hand-side vector to
output_ls 0 (no output)
Matrix Market file every # solves

2.1.25.8. .OPTIONS LINSOL-AC (Linear Solver Options)

For AC analysis, Xyce provides both iterative and direct methods for the solution of the linear equations.
For the advanced users, there are a variety of options that can be set to help improve these solvers.
Transformations of the linear system have a “TR_” prefix on the flag. Many of the options for the Krylov
solvers are simply passed through to the underlying Trilinos/AztecOO solution settings and thus have an
“AZ_” prefix on the flag.

Linear solver options are set using the .OPTIONS LINSOL-AC command. The available options are the
same as those for .OPTIONS LINSOL.

107
2.1.25.9. .OPTIONS OUTPUT (Output Options)

The .OPTIONS OUTPUT command can be used to allow control of the output frequency of data to files
specified by .PRINT TRAN commands.
One method is to specify output intervals. The format for this method is:
.OPTIONS OUTPUT INITIAL_INTERVAL=<interval> [<t0> <i0> [<t1> <i1>]* ]
where INITIAL_INTERVAL=<interval> specifies the starting interval time for output and <tx> <ix>
specifies later simulation times <tx> where the output interval will change to <ix>. The solution is output
at the exact intervals requested; this is done by interpolating the solution to the requested time points.
Another useful method for controlling the output frequency is to specify discrete output points.
.OPTIONS OUTPUT OUTPUTTIMEPOINTS=<t0>,<t1>,*
If this option is used, then only the specified time points will appear in the output file. No other points will
be output, so files using this method can be very sparse. For this type of output, the output values are not
interpolated. Instead, the specified output points are set as breakpoints in the time integrator, so the output
values are computed directly.
In addition to controlling the frequency of output, it is also possible to use output options to suppress the
header from standard format output files, and the footer from both standard and tecplot format output
files.
.OPTIONS OUTPUT PRINTHEADER=<boolean> PRINTFOOTER=<boolean>
where setting the PRINTHEADER variable to “false” will suppress the header and PRINTFOOTER variable to
“false” will suppress the footer. The PRINTHEADER option is only applicable to .PRINT <analysis>
FORMAT=<STD|GNUPLOT|SPLOT> files. The PRINTFOOTER option is only applicable to .PRINT
<analysis> FORMAT=<STD|GNUPLOT|SPLOT|TECPLOT> files.
It can be convenient to have all the solution variables output to file during a transient run without specifying
all of them on a .PRINT TRAN line. This can be accomplished with the SNAPSHOTS option available on the
.OPTIONS OUTPUT line.
.OPTIONS OUTPUT SNAPSHOTS=<boolean>
where setting the SNAPSHOTS variable to “true” will print all solution variables to the output file, ignoring
any solution nodes specified in the .PRINT TRAN line.
It is possible to add a STEPNUM column as the first column in the output file.
.OPTIONS OUTPUT ADD_STEPNUM_COL=<boolean>
where setting the ADD_STEPNUM_COL variable to “true” will add the STEPNUM column. The default is
“false”. This option is applicable to FORMAT=<STD|NOINDEX|GNUPLOT|SPLOT> for any .PRINT line that
supports FORMAT=STD output.
The default Xyce output for phase operators, such as VP(), IP(), SP(), YP() and ZP(), is in degrees. For
compatibility with other simulators like Spice3f5 and ngspice, it is possible to change that operator output
to use radians instead:
.OPTIONS OUTPUT PHASE_OUTPUT_RADIANS=<boolean>

108
The default value for this option is FALSE. If set to TRUE then the phase output will be in radians instead
of degrees. This option also applies to the format for AC sensitivity output. It does not affect the output
from a .FOUR analysis or a .FOUR measure though. Those two outputs are always in degrees.

2.1.25.10. .OPTIONS RESTART (Checkpointing Options)

The .OPTIONS RESTART command is used to control all checkpoint output and restarting.

The checkpointing form of the .OPTIONS RESTART command takes the following format:

General Format:

.OPTIONS RESTART [PACK=<0|1>] JOB=<job prefix>

+ [INITIAL_INTERVAL=<initial interval time> [<t0> <i0> [<t1> <i1>]* ]]

PACK=<0|1> indicates whether the restart data will be byte packed or not. Parallel restarts must always be
packed while Windows/MingW runs are always not packed. Otherwise, data will be packed by default
unless explicitly specified. JOB=<job prefix> identifies the prefix for restart files. The actual restart files
will be the job name with the current simulation time appended (e.g. name1e-05 for JOB=name and
simulation time 1e-05 seconds). Furthermore, INITIAL_INTERVAL=<initial interval time>
identifies the initial interval time used for restart output. The <tx> <ix> intervals identify times <tx> at
which the output interval (ix) should change. This functionality is identical to that described for the
.OPTIONS OUTPUT command.

Examples To generate checkpoints at every time step (default):

Example: .OPTIONS RESTART JOB=checkpt

To generate checkpoints every 0.1 𝜇𝑠:

Example: .OPTIONS RESTART JOB=checkpt INITIAL_INTERVAL=0.1us

To generate unpacked checkpoints every 0.1 𝜇𝑠:

Example: .OPTIONS RESTART PACK=0 JOB=checkpt INITIAL_INTERVAL=0.1us

To specify an initial interval of 0.1 𝜇𝑠, at 1 𝜇𝑠 change to interval of 0.5 𝜇𝑠, and at 10 𝜇𝑠 change to interval
of 0.1 𝜇𝑠:

Example:

.OPTIONS RESTART JOB=checkpt INITIAL_INTERVAL=0.1us 1.0us

+ 0.5us 10us 0.1us

109
2.1.25.11. .OPTIONS RESTART (Restarting Options)

To restart from an existing restart file, specify the file by either FILE=<restart file name> to explicitly
use a restart file or by JOB=<job name> START_TIME=<specified name> to specify a file prefix and a
specified time. The time must exactly match an output file time for the simulator to correctly identify the
correct file. To continue generating restart output files, INITIAL_INTERVAL=<interval> and following
intervals can be appended to the command in the same format as described above. New restart files will be
packed according to the previous restart file read in.
The restarting form of the .OPTIONS RESTART command takes the following format:

General Format:
.OPTIONS RESTART FILE=<restart file name>|JOB=<job name> START_TIME=<time>
+ [ INITIAL_INTERVAL=<interval> [<t0> <i0> [<t1> <i1>]* ]]

Examples Example restarting from checkpoint file at 0.133 𝜇𝑠:

Example: .OPTIONS RESTART JOB=checkpt START_TIME=0.133us

To restart from checkpoint file at 0.133 𝜇𝑠:

Example: .OPTIONS RESTART FILE=checkpt0.000000133

Restarting from 0.133 𝜇𝑠 and continue checkpointing at 0.1 𝜇𝑠 intervals:

Example:
.OPTIONS RESTART FILE=checkpt0.000000133 JOB=checkpt_again
+ INITIAL_INTERVAL=0.1us

2.1.25.12. .OPTIONS RESTART: special notes for use with two-level-Newton

Large parallel problems which involve power supply parasitics often require a two-level solve, in which
different parts of the problem are handled separately. In most respects, restarting a two-level simulation is
similar to restarting a conventional simulation. However, there are a few differences:

• When running with a two-level algorithm, Xyce requires (at least) two different input files. In order to
do a restart of a two-level Xyce simulation, it is necessary to have an .OPTIONS RESTART statement
in each file.
• It is necessary for the statements to be consistent. For example, the output times must be exactly the
same, meaning the initial intervals must be exactly the same.
• Xyce will not check to make sure that the restart options used in different files match, so it is up to the
user to ensure matching options.
• Finally, as each netlist that is part of a two-level solve will have its own .OPTIONS RESTART
statement, that means that each netlist will generate and/or use its own set of restart files. As a result,
the restart file name used by each netlist must be unique.

110
2.1.25.13. .OPTIONS SAMPLES (Sampling options)

The sampling selections listed in Table 2-10 provide methods for controlling Monte Carlo and Latin
Hypercube Sampling methods.
SAMPLES options are set using the .OPTIONS SAMPLES command. They are only used if the netlist also
includes a .SAMPLING statement.

Table 2-10. Options for Sampling Package.

Option Description Default
NUMSAMPLES Total number of samples 0
SAMPLE_TYPE Sampling type (MC or LHS) MC
Comma separated list of outputs (anything that would be a
OUTPUTS –
valid .RESULT output command)
Comma separated list of measure names (must refer to
MEASURES –
.MEASURE commands in the netlist)
Covariance matrix specified in row major form as
COVMATRIX –
comma-separated double precision numbers.
See
SEED Random seed
footnote.1
Compute and outputs statistics for specified outputs and/or
OUTPUT_SAMPLE_STATS –
measures.
Enable regression based PCE. If this is enabled, the randomly
REGRESSION_PCE sampled points will be used to produce a PCE approximation –
using regression methodss.
Enable projection based PCE (quadrature). If this is enabled, a
PCE approximation will be created using quadrature methods.
PROJECTION_PCE –
The NUMSAMPLES parameter will be ignored, and the
samples will be the quadrature points used by projection PCE.
Once the PCE coefficients are obtained, perform sampling on
RESAMPLE –
the PCE approximation
OUTPUT_PCE_COEFFS Output the PCE coefficients –
SPARSE_GRID Use sparse grid methods if using projection PCE. –
STDOUTPUT Send sampling and PCE output to the terminal –

2.1.25.14. .OPTIONS EMBEDDEDSAMPLES (Embedded Sampling options)

The sampling selections listed in Table 2-11 provide methods for controlling Embedded Sampling
methods.
EMBEDDEDSAMPLES options are set using the .OPTIONS EMBEDDEDSAMPLES command. They are only
used if the netlist also includes a .EMBEDDEDSAMPLING statement.
1 The seed can also be set using command line option, -randseed. The command line seed will override the netlist seed value. If the
seed is not set in either the netlist or on the command line, then Xyce generates a seed internally. In all cases, Xyce will output
text to the console indicating what seed is being used.

111
 Table 2-11. Options for Embedded Sampling Package.
Option Description Default
NUMSAMPLES Total number of samples 0
SAMPLE_TYPE Sampling type (MC or LHS) MC
Comma separated list of outputs (anything that would be a
OUTPUTS –
valid .PRINT output variable)
Covariance matrix specified in row major form as
COVMATRIX –
comma-separated double precision numbers.
See
SEED Random seed
footnote.1
OUTPUT_SAMPLE_STATSCompute and outputs statistics for specified outputs. –
Enable regression based PCE. If this is enabled, the randomly
REGRESSION_PCE sampled points will be used to produce a PCE approximation –
using regression methods.
Enable projection based PCE (quadrature). If this is enabled, a
PCE approximation will be created using quadrature methods.
PROJECTION_PCE –
The NUMSAMPLES parameter will be ignored, and the
samples will be the quadrature points used by projection PCE.
Once the PCE coefficients are obtained, perform sampling on
RESAMPLE –
the PCE approximation
OUTPUT_PCE_COEFFS Output the PCE coefficients –
SPARSE_GRID Use sparse grid methods if using projection PCE. –
STDOUTPUT Send sampling and PCE output to the terminal –

2.1.25.15. .OPTIONS PCES (Fully intrusive PCE options)

The sampling selections listed in Table 2-12 provide methods for controlling Embedded Sampling
methods.
PCES options are set using the .OPTIONS PCES command. They are only used if the netlist also includes a
.PCE statement.

Table 2-12. Options for PCE Package.

Option Description Default
Comma separated list of outputs (anything that would be a
OUTPUTS –
valid .PRINT output variable)
Covariance matrix specified in row major form as
COVMATRIX –
comma-separated double precision numbers.

1 The seed can also be set using command line option, -randseed. The command line seed will override the netlist seed value. If the
seed is not set in either the netlist or on the command line, then Xyce generates a seed internally. In all cases, Xyce will output
text to the console indicating what seed is being used.

112
 Table 2-12. Options for PCE Package.
Option Description Default
Sampling type (MC or LHS). This is only used if resampling is
SAMPLE_TYPE MC
enabled.
See
SEED Random seed. This is only used if resampling is enabled.
footnote.1
OUTPUT_SAMPLE_STATSCompute and outputs statistics for specified outputs. –
Once the PCE coefficients are obtained, perform sampling on
RESAMPLE –
the PCE approximation
OUTPUT_PCE_COEFFS Output the PCE coefficients –
SPARSE_GRID Use sparse grid methods if using projection PCE. –
STDOUTPUT Send sampling and PCE output to the terminal –

2.1.25.16. .OPTIONS SENSITIVITY (Direct and Adjoint Sensitivity Options)

The sensitivity selections listed in Table 2-13 provide methods for controlling direct and adjoint sensitivity
analysis.

SENSITIVITY options are set using the .OPTIONS SENSITIVITY command. They are only used if the
netlist also includes a .SENS statement.

Table 2-13. Options for Sensitivity Package.

Option Description Default
ADJOINT Flag to enable adjoint sensitivity calculation false
DIRECT Flag to enable direct sensitivity calculation false
OUTPUTSCALED Flag to enable output of scaled sensitivities false
OUTPUTUNSCALED Flag to enable output of unscaled sensitivities true
STDOUTPUT Flag to enable output of sensitivies to std output false
Start time for set of time steps over which to compute transient
ADJOINTBEGINTIME 0.0
adjoints.
End time for set of time steps over which to compute transient
ADJOINTFINALTIME 1.0e+99
adjoints.
List of comma-separated time points at which to compute
ADJOINTTIMEPOINTS –
transient adjoints.

1 The seed can also be set using command line option, -randseed. The command line seed will override the netlist seed value. If the
seed is not set in either the netlist or on the command line, then Xyce generates a seed internally. In all cases, Xyce will output
text to the console indicating what seed is being used.

113
2.1.25.17. .OPTIONS HBINT (Harmonic Balance Options)

The Harmonic Balance parameters listed in Table 2-14 give the available options for helping control the
algorithm for harmonic balance analysis.

Harmonic Balance options are set using the .OPTIONS HBINT command.

Table 2-14. Options for HB.

Option Description Default
Number of harmonics to be calculated for each tone. It must
10
NUMFREQ have the same number of entries as .HB statement
Number of periods to integrate through before calculating the
STARTUPPERIODS 0
initial conditions. This option is only used when TAHB=1.
SAVEICDATA Write out the initial conditions to a file. 0
This flag sets transient assisted HB. When TAHB=0, transient
analysis is not performed to get an initial guess. When
TAHB=1, it uses transient analysis to get an initial guess. For
multi-tone HB simulation, the initial guess is generated by a
TAHB single tone transient simulation. The first tone following .HB is 1
used to determine the period for the transient simulation. For
multi-tone HB simulation, it should be set to the frequency that
produces the most nonlinear response by the circuit. When tahb
= 2, the DC op is used as an initial guess
This flag sets voltage limiting for HB. During the initial guess
calculation, which normally uses transient simulation, the
VOLTLIM voltage limiting flag is determined by .options device voltlim. 1
During the HB phase, the voltage limiting flag is determined by
.options hbint voltlim.
the largest
The maximum intermodulation product order used in the value in the
INTMODMAX
spectrum. NUMFREQ
list.
The total
number of
frequencies
NUMTPTS Number of time points in the output
(positive,
negative and
DC).
The truncation method used in multi-tone HB to select
SELECTHARMS harmonics. Box, diamond and hybrid truncation methods are hybrid
supported

2.1.25.18. .OPTIONS DIST (Parallel Distribution Options)

The parameters listed in Table 2-15 give the available options for controlling the parallel distribution used
in Xyce. There are three choices for distribution strategy.

114
The default distribution strategy is “first-come, first-served” (STRATEGY=0), which divides the devices
found in the netlist into equal sized groups (in the order they are parsed) and distributes a group to each
processor. This does not take into account the connectivity of the circuit or balance device model
computation, and therefore can exhibit parallel imbalance for post-layout circuits that have a substantial
portion of parasitic devices.
The “flat round-robin” strategy (STRATEGY=1) will generate the same distribution as the default strategy, but
every parallel processor will participate in reading its portion of the netlist. This strategy provides a more
scalable setup than the default strategy, but can only be applied to flattened (non-hierarchical) netlists.
The “device balanced” strategy (STRATEGY=2) will evenly divide each of the device types over the number
of parallel processors, so each processor will have a balanced number of each model type. This allieviates
the parallel imbalance in the device model computation that can be experienced with post-layout circuits.
However, it does not take into account the circuit connectivity, so the communication will not be minimized
by this strategy.

Table 2-15. Options for Parallel Distribution.

Option Description Default
Parallel device distribution strategy
• 0 (First-Come, First-Served)
• 1 (Flat Round-Robin) 0
STRATEGY
• 2 (Device Balanced)

2.1.25.19. .OPTIONS FFT (FFT Options)

The parameters listed in Table 2-16 give the available options for controlling all of the .FFT statements in a
given Xyce netlist.
If FFT_ACCURATE is set to 1 (true), which is the default, then Xyce will insert breakpoints at the sample
times requested by the collection of .FFT lines in the netlist. This has been found to improve the accuracy
of the .FFT analyses, at the possible expense of simulation speed. If FFT_ACCURATE is set to 0 (false), then
interpolation is used to determine the output variable values at the specified sample times. If the
-remeasure command line option is used to recalculate the .MEASURE FFT and/or .FFT statements for a
.TRAN analysis, then FFT_ACCURATE is set to 0 during the re-measure operation. Finally, if .OPTIONS
OUTPUT INITIAL_INTERVAL is used in the netlist then .OPTIONS FFT FFT_ACCURATE will also be set to
0.
If FFTOUT is set to 1 then additional metrics are output to both stdout and the <netlistName>.fft0 file
for each .FFT line. In addition a sorted list of the 30 largest harmonics is output to stdout. Those additional
metrics are as follows, where Section 2.1.8.3 provides detailed definitions for these metrics:

• Effective Number of Bits (ENOB)

• Spurious Free Dynamic Range (SFDR)
• Signal to Noise Ratio (SNR)
• Signal to Noise-and-Distortion Ratio (SNDR)

115
 • Total Harmonic Distorion (THD)

The setting for FFT_MODE is used to control whether the Xyce FFT processing and output are more
compatible with HSPICE (0) or Spectre (1). This setting affects the format of the window functions, the
conversion from two-sided to one-sided results, and whether the default output for the magnitude values is
normalized, or not. Section 2.1.8.1 gives more details.

Table 2-16. Options for FFT.

Option Description Default
Insert breakpoints at the sample times requested by the
1 (true)
FFT_ACCURATE collection of .FFT lines in the netlist
Output additional metrics (ENOB, SFDR, SNR, SNDR and
FFTOUT THD) and a sorted list of the 30 largest harmonics for each 0 (false)
.FFT line
Controls whether the FFT calculations and output format are
FFT_MODE 0
more compatible with HSPICE (0) or Spectre (1)

2.1.25.20. .OPTIONS MEASURE (Measure Options)

The parameters listed in Table 2-17 give the available options for controlling all of the .MEASURE
statements in a given Xyce netlist. The MEASDGT, MEASFAIL and MEASOUT options are included for HSPICE
compatibility.
If given in the netlist, the setting for MEASOUT controls whether the .mt# (or .ms# or .ma#) files are made
(1) or not (0). The MEASOUT setting takes precedence over the MEASPRINT setting (which is a Xyce-specific
option) if both are given in the netlist. See Section 2.1.18.7 for more details then on how the MEASPRINT
option interacts with the individual .MEASURE statements and the -remeasure command line option.
If given in the netlist, the setting for the MEASDGT overrides the PRECISION qualifiers given on individual
.MEASURE lines. The default value for the MEASDGT option is different from in HSPICE.
The Xyce behavior for failed measures can be controlled via the MEASFAIL and DEFAULT_VAL options, as
well as with the DEFAULT_VAL qualifiers on individual .MEASURE lines. The order of precedence is the
DEFAULT_VAL option and then the DEFAULT_VAL qualifier on individual .MEASURE lines. If MEASFAIL=0
then Xyce outputs the default value in the .mt# ( or .ms# or .ma#) files for a failed measure. If
MEASFAIL=1 (or any other non-zero value) then Xyce outputs “FAILED” in the .mt# ( or .ms# or .ma#)
files for a failed measure. If given in the netlist, the setting for the DEFAULT_VAL option overrides the
DEFAULT_VAL qualifiers given on individual .MEASURE lines. The DEFAULT_VAL option and the
DEFAULT_VAL qualifiers can be set to any real number. For all of these cases, Xyce will print “FAILED” to
the standard output for a failed measure. As a final note, the FOUR measure is a special case since it
produces multiline output. Failed FOUR measures will be reported as “FAILED” in the .mt# ( or .ms# or
.ma#) files, irrespective of the various MEASFAIL and DEFAULT_VAL settings.
The USE_CONT_FILES option controls whether each AC_CONT, DC_CONT, NOISE_CONT or TRAN_CONT
mode measure uses a separate output file for its results, or not. Section 2.1.19.1 provides more details and
an example netlist for this options setting.
For backwards compatibility with previous Xyce versions, USE_LTTM has been added. This option defaults
to 0, which uses the new version of the TRIG-TARG measure; while setting it to 1 will use the old version of

116
the TRIG-TARG measure for all TRIG-TARG measures in the netlist. If the FRAC_MAX qualifier is used on a
TRIG-TARG line then Xyce will automatically default to USE_LTTM=1 for that particular measure line. It is
anticipated that this option setting will be removed at some point.

Table 2-17. Options for MEASURE.

Option Description Default
Default value for “failed measures” in the .mt# ( or .ms# or
-1
DEFAULT_VAL .ma#) files.
Precision for all .MEASURE statements. This value applies to
MEASDGT the output to both the .mt# ( or .ms# or .ma#) files and the 6
standard output.
MEASFAIL Specify output format for failed measures 1
MEASOUT Control whether the .mt0 file is made or not 1
Measure Output
• ALL (Output measure information to both file(s) and
stdout)
MEASPRINT ALL
• STDOUT (Output measure information to stdout only)
• NONE (Suppress all measure output)

Specifies whether “continuous” mode measures use separate

USE_CONT_FILES 1 (TRUE)
output files for each such measure
Use the “Legacy Trig-Trag Mode“. This option is included for
USE_LTTM backwards compatibility with previous Xyce versions. It may 0 (FALSE).
be removed in the future

2.1.25.21. .OPTIONS PARSER (Parser Options)

The parameter listed in Table 2-18 gives the available option for netlist parsing.

Table 2-18. Options for Parsing.

Option Description Default
Enable model binning during netlist parsing. See Section
MODEL_BINNING TRUE
2.1.21 for more details on how model binning works in Xyce.
Scale factor for geometric parameters such as MOSFET length
and width. This can also be specified as .option scale
SCALE (singular .OPTION and omitting the keyword PARSER) for 1.0
compatibility with other simulators. See section 2.1.21 for an
example usage.

117
2.1.26. .PARAM (Parameter)

User defined parameter that can be used in expressions throughout the netlist.

General Form .PARAM [<name>=<value>]*

Examples .PARAM A_Param=1K

.PARAM B_Param={A_Param*3.1415926535}

Arguments and
Options name
Parameter name. Parameters may be redefined. If the same parameter name
is used on multiple parameters, Xyce by default will use the last parameter of
that name. By default, no warning will be emitted. To change this behavior,
one can use the -redefined_param command line option, described in
section 3-1.
value
The value may be a number or an expression.

Comments Parameters defined using .PARAM only have a few restrictions on their usage. In earlier
versions of Xyce they were handled as constants that were evaluated during parsing.
This is no longer the case, and parameters can now have their values change
throughout the calculation. A .PARAM defined in the top level netlist is equivalent to a
.GLOBAL_PARAM, and they can be combined as needed. Thus, you may use parameters
defined by .PARAM in expressions used to define global parameters, and you may also
use global parameters in .PARAM definitions.
It is legal for parameters to depend on special variables such as TIME, FREQ, TEMP
and VT variables. However, it is not legal for parameters to depend on solution
variables such as voltage nodes or independent source currents.
Parameters defined using .PARAM can be modified directly by various analyses, such
as .DC, .STEP, .SAMPLING and .EMBEDDEDSAMPLING, subject to scoping rules.
To load an external data file with time voltage pairs of data on each line into a global
parameter, use this syntax:
.GLOBAL_PARAM extdata = {tablefile("filename")}
or
.GLOBAL_PARAM extdata = {table("filename")}
where filename would be the name of the file to load. Other interpolators that can
read in a data table from a file include fasttable,spline, akima, cubic, wodicka
and bli. See 2.2 for further information.
There are several reserved words that may not be used as names for parameters. These
reserved words are:
• Time

118
 • Freq
• Hertz
• Vt
• Temp
• Temper
• GMIN
The scoping rules for parameters are:
• If a .PARAM, statement is included in the main circuit netlist, then it is accessible
from the main circuit and all subcircuits.
• .PARAM statements defined within a subcircuit are scoped to that subciruit
definition. So, their parameters are only accessible within that subcircuit
definition, as well as within “nested subcircuits” also defined within that
subcircuit definition.
• Parameters defined via .PARAM statements can be modified by the various UQ
analysis techniques (.STEP, .SAMPLING, etc) but this only works for .PARAM
that have been defined in the top level netlist. Parameters defined inside of
subcircuits cannot be modified directly by these analyes, but they can be
modified indirectly via dependence on other globally scoped parameter.
Additional illustative examples of scoping are given in the “Working with Subcircuits
and Models” section of the Xyce Users’ Guide [1] .

119
2.1.27. .PCE (Fully Intrusive Polynomial Chaos Expansion (PCE) Analysis)

Calculates a fully intrusive Polynommial Chaos Expansion (PCE) analysis (for .DC or .TRAN only) to
propagate uncertainty from a set of uncertain inputs to uncertain outputs. This involves evaluating the
circuit at a set of parameter values corresponding to quadrature points used by the PCE algorithm. The loop
over parameter values happens at the inner-most part of the calculation, so all samples are propagated
simultaneously.

This fully-intrusive form of PCE is an experimental analysis method. Non-intrusive methods of PCE are
also available in Xyce and will usually be a better choice. The non-intrusive methods are used in
combination with .SAMPLING and/or .EMBEDDEDSAMPLING. Of those other methods, the behavior of .PCE
most closely resembles that of .EMBEDDEDSAMPLING with projection_pce=true.

General Form .PCE

+ param=<parameter name>,[parameter name]*
+ type=<parameter type>,[parameter type]*
+ means=<mean>,[mean]*
+ std_deviations=<standard deviation>,[standard deviation]*
+ lower_bounds=<lower bound>,[lower bound]*
+ upper_bounds=<upper bound>,[upper bound]*
+ alpha=<alpha>,[alpha]*
+ beta=<beta>,[beta]*

Examples .PCE
+ param=R1
+ type=normal
+ means=3K
+ std_deviations=1K

.PCE
+ param=R1,R2
+ type=uniform,uniform
+ lower_bounds=1K,2K
+ upper_bounds=5K,6K

.PCE useExpr=true

.options PCES
+ OUTPUTS={R1:R},{V(1)}

Arguments and
Options param
Names of the parameters to be sampled. This may be any of the parameters
that are valid for .STEP, including device instance, device model, or global
parameters. If more than one parameter, then specify as a comma-separated
list.

120
 type
Distribution type for each parameter. This may be uniform, normal or
gamma. If more than one parameter, then specify as a comma-separated list.
means
If using normal distributions, the mean for each parameter must be specified.
If more than one parameter, then specify as a comma-separated list.
std_deviations
If using normal distributions, the standard deviation for each parameter must
be specified. If more than one parameter, then specify as a comma-separated
list.
lower_bounds
If using uniform distributions, the lower bound must be specified. This is
optional for normal distributions. If used with normal distributions, may
alter the mean and standard deviation. If more than one parameter, then
specify as a comma-separated list.
upper_bounds
If using uniform distributions, the upper bound must be specified. This is
optional for normal distributions. If used with normal distributions, may
alter the mean and standard deviation. If more than one parameter, then
specify as a comma-separated list.
alpha
If using gamma distributions, the alpha value for each parameter must be
specified. If more than one parameter, then specify as a comma-separated list.
beta
If using gamma distributions, the beta value for each parameter must be
specified. If more than one parameter, then specify as a comma-separated list.
useExpr
If this argument is set to true, then the sampling algorithm will set up
random inputs from expression operators such as AGAUSS and AUNIF. In this
case it will also ignore the list of parameters on the .PCE command line. For
a complete description of expression-based random operators, see the
expression documentation in section 2.2.

Comments In addition to the .PCE command, this analysis requires a .options PCES command
as well. The .PCE command specifies parameters and their attributes, either using the
useExpr option, or with comma-separated lists. The .options PCES command
specifies the outputs for which to compute statistics. To see the details of the
.options PCES command , see table 2-12.

On the .PCE command line, if not using useExpr, parameters and their attributes
must be specified using comma-separated lists. The comma-separated lists must all be
the same length.

The .PRINT PCE command provides output based on the contents of those print-lines,
and also the OUTPUT arguments on the .OPTIONS PCES line.

121
If the OUTPUT_SAMPLE_STATS argument on a .PRINT PCE line is set to “true” then
the statistics for the MEAN, MEANPLUS, MEANMINUS, STDDEV and VARIANCE will be
output for each variable in the OUTPUT argument. If the OUTPUT_ALL_SAMPLES
argument on a .PRINT PCE line is set to “true” then the values of all quadrature
points, for each variable requested in the OUTPUTS argument, will be output.

122
2.1.28. .PREPROCESS REPLACEGROUND (Ground Synonym)

The purpose of ground synonym replacement is to treat nodes with the names GND, GND!, GROUND or any
capital/lowercase variant thereof as synonyms for node 0. The general invocation is

General Form .PREPROCESS REPLACEGROUND <bool>

Arguments and
Options bool
If TRUE, Xyce will treat all instances of GND, GND!, GROUND, or any
capital/lowercase variant thereof, as synonyms for node 0. If FALSE,
Xyce will consider each term as a separate node. Only one .PREPROCESS
REPLACEGROUND statement is permissible in a given netlist file.

123
2.1.29. .PREPROCESS REMOVEUNUSED (Removal of Unused Components)

If a given netlist file contains devices whose terminals are all connected to the same node
(e.g., R2 1 1 1M), it may be desirable to remove such components from the netlist before simulation
begins. This is the purpose of the command

General Form .PREPROCESS REMOVEUNUSED [<value>]

Arguments and
Options value
is a list of components separated by commas. The allowed values are
C Capacitor
D Diode
I Independent Current Source
L Inductor
M MOSFET
Q BJT
R Resistor
V Independent Voltage Source

Examples .PREPROCESS REMOVEUNUSED R,C

.PREPROCESS will attempt to search for all resistors and capacitors in a given netlist
file whose individual device terminals are connected to the same node and remove
these components from the netlist before simulation ensues. A list of components that
are supported for removal is given above. Note that for MOSFETS and BJTs, three
terminals on each device (the gate, source, and drain in the case of a MOSFET and the
collector, base, and emitter in the case of a BJT) must be the same for the device to be
removed from the netlist. As before, only one .PREPROCESS REMOVEUNUSED line is
allowed in a given netlist file.

124
2.1.30. .PREPROCESS ADDRESISTORS (Adding Resistors to Dangling Nodes)

We refer to a dangling node as a circuit node in one of the following two scenarios: either the node is
connected to only one device terminal, and/or the node has no DC path to ground. If several such nodes
exist in a given netlist file, it may be desirable to automatically append a resistor of a specified value
between the dangling node and ground. To add resistors to nodes which are connected to only one device
terminal, one may use the command

General Form .PREPROCESS ADDRESISTORS ONETERMINAL <value>

Arguments and
Options value
is the value of the resistor to be placed between nodes with only one device
terminal connection and ground. For instance, the command

Examples .PREPROCESS ADDRESISTORS ONETERMINAL 1G

will add resistors of value 1G between ground and nodes with only one device
terminal connection and ground. The command

Examples .PREPROCESS ADDRESISTORS NODCPATH <value>

acts similarly, adding resistors of value <VALUE> between ground and all nodes which
have no DC path to ground.
The .PREPROCESS ADDRESISTORS command is functionally different from either of
the prior .PREPROCESS commands in the following way: while the other commands
augment the netlist file for the current simulation, a .PREPROCESS ADDRESISTORS
statement creates an auxiliary netlist file which explicitly contains a set of resistors
that connect dangling nodes to ground. If the original netlist file containing a
.PREPROCESS ADDRESISTORS statement is called filename, invoking Xyce on this
file will produce a file filename_xyce.cir which contains the resistors that connect
dangling nodes to ground. One can then run Xyce on this file to run a simulation in
which the dangling nodes are tied to ground. Note that, in the original run on the file
filename, Xyce will continue to run a simulation as usual after producing the file
filename_xyce.cir, but this simulation will not include the effects of adding
resistors between the dangling nodes and ground. Refer to the Xyce Users’
Guide [1] for more detailed examples on the use of .PREPROCESS ADDRESISTOR
statements.
Note that it is possible for a node to have one device terminal connection and,
simultaneously, have no DC path to ground. In this case, if both a ONETERMINAL and
NODCPATH command are invoked, only the resistor for the ONETERMINAL connection is
added to the netlist; the NODCPATH connection is omitted.
As before, each netlist file is allowed to contain only one .PREPROCESS
ADDRESISTORS ONETERMINAL and one .PREPROCESS ADDRESISTORS NODCPATH
line each, or else Xyce will exit in error.

125
2.1.31. .PRINT (Print output)

Send analysis results to an output file.

Xyce allows multiple output files to be created during the run and supports several options for each.

General Form .PRINT <print type> [FILE=<output filename>]

+ [FORMAT=<STD|NOINDEX|PROBE|TECPLOT|RAW|CSV|GNUPLOT|SPLOT>]
+ [WIDTH=<print field width>]
+ [PRECISION=<floating point output precision>]
+ [FILTER=<absolute value below which a number outputs as 0.0>]
+ [DELIMITER=<TAB|COMMA>] [TIMESCALEFACTOR=<real scale factor>]
+ [OUTPUT_SAMPLE_STATS=<boolean>] [OUTPUT_ALL_SAMPLES=<boolean>]
+ <output variable> [output variable]*

Examples .print tran format=tecplot V(1) I(Vsrc) {V(1)*(I(Vsrc)**2.0)}

.PRINT TRAN FORMAT=PROBE FILE=foobar.csd V(1) {abs(V(1))-5.0}

.PRINT DC FILE=foobar.txt WIDTH=19 PRECISION=15 FILTER=1.0e-10

+ I(VSOURCE5) I(VSOURCE6)

.print tran FORMAT=RAW V(1) I(Vsrc)

R1 1 0 100
X1 1 2 3 MySubcircuit
V1 3 0 1V
.SUBCKT MYSUBCIRCUIT 1 2 3
R1 1 2 100K
R2 2 4 50K
R3 4 3 1K
.ENDS

.PRINT DC V(X1:4) V(2) I(V1)

Arguments and
Options
print type
A print type is the name of an analysis, one of the analysis specific print
subtypes, or a specialized output command.

126
 Analysis Print Type Description
Sets default variable list and formats
.AC AC
for print subtypes
Overrides variable list and format for
.AC AC_IC
AC initial conditions
.DC DC
.EMBDEDDEDSAMPLING ES
.HB HB
Overrides variable list and format for
.HB HB_FD
HB frequency domain
Overrides variable list and format for
.HB HB_IC
HB initial conditions
Overrides variable list and format for
.HB HB_STARTUP
HB start up
Overrides variable list and format for
.HB HB_TD
HB time domain
.NOISE Noise Outputs Noise spectral density curves
.TRAN TRAN
Specialized Output Commands
Sets variable list and format for ho-
Homotopy HOMOTOPY
motopy
Sets variable list and format for sen-
.SENS SENS
sitivity

A netlist may contain many .PRINT commands, but only commands with
analysis types which are appropriate for the analysis being performed are
processed. This feature allows you to generate multiple formats and variable
sets in a single analysis run.
For analysis types that generate multiple output files, the print subtype allows
you to specify variables and output parameters for each of those output files.
If there is no .PRINT <subtype> provided in the net list, the variables and
parameters from the analysis type will be used.

FORMAT=<STD|NOINDEX|PROBE|TECPLOT|RAW|CSV|GNUPLOT|SPLOT>
The output format may be specified using the FORMAT option. The STD
format outputs the data divided up into data columns. The NOINDEX format is
the same as the STD format except that the index column is omitted. The
PROBE format specifies that the output should be formatted to be compatible
with the PSpice Probe plotting utility. The TECPLOT format specifies that the
output should be formatted to be compatible with the Tecplot plotting
program. The RAW format specifies that the output should comply with the
SPICE binary rawfile format. The -a command line option, in conjunction
with FORMAT=RAW on the .PRINT line, can then be used to output an ASCII
rawfile. The CSV format specifies that the output file should be a

127
 comma-separated value file with a header indicating the variables printed in
the file. It is similar to, but not identical to using DELIMITER=COMMA; the
latter will also print a footer that is not compatible with most software that
requires CSV format. The GNUPLOT (or SPLOT) format is the same as the STD
format except that if .STEP is used then two (or one) blank lines are inserted
before the data for steps 1,2,3,... where the first step is step 0. The SPLOT
format is useful for when the “splot” command in gnuplot is used to produce
3D perspective plots.
FILE=<output filename>
Specifies the name of the file to which the output will be written. See the
“Results Output and Evaluation Options” section of the Xyce Users’
Guide [1] for more information on how this feature works for analysis types
(e.g., AC and HB) that can produce multiple output files.
WIDTH=<print field width>
Controls the output width used in formatting the output.
PRECISION=<floating point precision>
Number of floating point digits past the decimal for output data.
FILTER=<filter floor value>
Used to specify the absolute value below which output variables will be
printed as 0.0.
DELIMITER=<TAB|COMMA>
Used to specify an alternate delimiter in the STD or NOINDEX format
output.
TIMESCALEFACTOR=<real scale factor>
Specify a constant scaling factor for time. Time is normally printed in units
of seconds, but if one would like the units to be milliseconds, then set
TIMESCALEFACTOR=1000.
OUTPUT_SAMPLE_STATS=<boolean>
Output the sample statistics for an EMBEDDEDSAMPLING analysis.This
argument is only supported for .PRINT ES. Its default value is true. Section
2.1.5 has more details.
OUTPUT_ALL_SAMPLES=<boolean>
Output all of the sample values for an EMBEDDEDSAMPLING analysis. This
argument is only supported for .PRINT ES. Its default value is false. Section
2.1.5 has more details.
<output variable>
Following the analysis type and other options is a list of output variables.
There is no upper bound on the number of output variables. The output is
divided up into data columns and output according to any specified options
(see options given above). Output variables can be specified as:
• V(<circuit node>) to output the voltage at <circuit node>
• V(<circuit node>,<circuit node>) to output the voltage
difference between the first <circuit node> and second <circuit

128
 node>
• I(<device>) to output current through a two terminal device
• I<lead abbreviation>(<device>) to output current into a
particular lead of a three or more terminal device (see the Comments,
below, for details)
• P(<device>) or W(<device>) to output the power
dissipated/generated in a device. At this time, not all devices support
power calculations. In addition, the results for semiconductor devices
(D, J, M, Q and Z devices) and the lossless transmission device (T
device) may differ from other simulators. Consult the Features
Supported by Xyce Device Models table in section 2.3 and the individual
sections on each device for more details. Finally, power calculations are
not supported for any devices for .AC and .NOISE analyses.
• N(<device internal variable>) to output a specific device’s
internal variable. (The comments section below has more detail on this
syntax.)
• {expression} to output an expression
• <device>:<parameter> to output a device parameter
• <model>:<parameter> to output a model parameter
When the analysis type is AC, HB or Noise, additional output variable
formats are available:
• VR(<circuit node>) to output the real component of voltage
response at a point in the circuit
• VI(<circuit node>) to output the imaginary component of voltage
response at a point in the circuit
• VM(<circuit node>) to output the magnitude of voltage response
• VP(<circuit node>) to output the phase of voltage response in
degrees
• VDB(<circuit node>) to output the magnitude of voltage response in
decibels.
• VR(<circuit node>,<circuit node>) to output the real
component of voltage response between two nodes in the circuit
• VI(<circuit node>,<circuit node>) to output the imaginary
component of voltage response between two nodes in the circuit
• VM(<circuit node>,<circuit node>) to output the magnitude of
voltage response between two nodes in the circuit
• VP(<circuit node>,<circuit node>) to output the phase of
voltage response between two nodes in the circuit in degrees
• VDB(<circuit node>,<circuit node>) to output the magnitude of
voltage response between two nodes in the circuit, in decibels

129
 • IR(<device>) to output the real component of the current through a
two terminal device
• II(<device>) to output the imaginary component of the current
through a two terminal device
• IM(<device>) to output the magnitude of the current through a two
terminal device
• IP(<device>) to output the phase of the current through a two
terminal device in degrees
• IDB(<device>) to output the magnitude of the current through a two
terminal device in decibels.
In AC and Noise analyses, outputting a voltage node without any of these
optional designators results in output of the real and imaginary parts of the
signal. Note that under AC and Noise analyses, current variables are only
supported for devices that have “branch currents”that are part of the solution
vector. This includes the V, E, H and L devices. It also includes the
voltage-form of the B device.
Note that when using the variable list for time domain output, usage of
frequency domain functions like VDB can result in -Inf output being written
to the output file. This is easily solved by defining the time domain
equivalent command to specify the correct output for time domain data.
Further explanation of the current specifications is given in comments
section below.
When a .LIN analysis is done then additional output variable formats are
available via the .PRINT AC line, where <index1> and <index2> must
both be greater than 0 and also both less than or equal to the number of ports
in the netlist:
• SR(<index1>,<index2>) to output the real component of an
S-parameter
• SI(<index1>,<index2>) to output the imaginary component of an
S-parameter
• SM(<index1>,<index2>) to output the magnitude of an S-parameter
• SP(<index1>,<index2>) to output the phase of an S-parameter in
degrees
• SDB(<index1>,<index2>) to output the magnitude of an S-parameter
in decibels.
• YR(<index1>,<index2>) to output the real component of a
Y-parameter
• YI(<index1>,<index2>) to output the imaginary component of a
Y-parameter
• YM(<index1>,<index2>) to output the magnitude of a Y-parameter
• YP(<index1>,<index2>) to output the phase of a Y-parameter in

130
 degrees
• YDB(<index1>,<index2>) to output the magnitude of a Y-parameter
in decibels.
• ZR(<index1>,<index2>) to output the real component of a
Z-parameter
• ZI(<index1>,<index2>) to output the imaginary component of a
Z-parameter
• ZM(<index1>,<index2>) to output the magnitude of a Z-parameter
• ZP(<index1>,<index2>) to output the phase of a Z-parameter in
degrees
• ZDB(<index1>,<index2>) to output the magnitude of a Z-parameter
in decibels.
When the analysis type is Noise, additional output variable formats are
available via the .PRINT NOISE line for devices that support stationary
noise.
• INOISE to output the input noise contributions
• ONOISE to output the output noise contributions
• DNI(<deviceName>) to output the input noise contribution from
device <deviceName>
• DNI(<deviceName>,<noiseSource>) to output the input noise
contribution from source <noiseSource> for device <deviceName>
• DNO(<deviceName>) to output the output noise contribution from
device <deviceName>
• DNO(<deviceName>,<noiseSource>) to output the output noise
contribution from source <noiseSource> for device <deviceName>

Comments • Currents are positive flowing from node 1 to node 2 for two node devices, and
currents are positive flowing into a particular lead for multi-terminal devices.
• <circuit node> is simply the name of any node in your top-level circuit, or
<subcircuit name>:<node> to reference nodes that are internal to a
subcircuit.
• <device> is the name of any device in your top-level circuit, or <subcircuit
name>:<device> to reference devices that are internal to a subcircuit.
• <lead abbreviation> is a single character designator for individual leads on
a device with three or more leads. For bipolar transistors these are: c (collector),
b (base), e (emitter), and s (substrate). For MOSFETs, JFETs, and MESFETs,
lead abbreviations are: d (drain), g (gate), s (source), and for MOSFETS and
JFETs, b (bulk). In addition to these standard leads, SOI and CMG MOSFETs
have e (bulk) nodes and SOI transistors have optional b (body) nodes whose lead
currents may also be printed in this manner. For PDE devices, the nodes are
numbered according to the order they appear, so lead currents are referenced like

131
 I1(<device>), I2(<device>), etc. In Xyce, a .PRINT line request like I(Q1) is
a parsing error for a multi-terminal device. Instead, an explicit lead current
designator like IC(Q1) must be used.

• The "lead current" method of printing from devices in Xyce is done at a low level
with special code added to each device; the method is therefore only supported in
specific devices that have this extra code. So, if .PRINT I(Y) does not work, for
a device called Y, then you will need to attach an ammeter (zero-volt voltage
source) in series with that device and print the ammeter’s current instead.

• Lead currents of subcircuit ports are not supported. However, access is provided
via specific node names (e.g., X1:internalNodeName) or specific devices (e.g.,
X1:V3) inside the subcircuit.

• For STD formatted output, the values of the output variables are output as a
series of columns (one for each output variable).

• When the command line option -r <raw-file-name> is used, all of the output
is diverted to the raw-file-name file as a concatenation of the plots, and each plot
includes all of the variables of the circuit instead of the variable list(s) given on
the .PRINT lines in the netlist. Using the -a options in conjunction with the -r
option results in a raw file that is output all in ASCII characters.

• Any output going to the same file from one simulation of Xyce results in the
concatenation of output. However, if a simulation is re-run then the original
output will we over-written.

• During analysis a number of output files may be generated. The selection of

which files are created depends on a variety of factors, most obvious of which is
the .PRINT command. See section 2.1 for more details.

• Frequency domain values are output as complex values for Raw, TecPlot and
Probe formats when a complex variable is printed. For STD and CSV formats,
the output appears in two columns, the real part followed by the imaginary part.
The print variables VR, VI, VM, VDB and VP print the scalar values for the real
part, imaginary part, magnitude, magnitude in decibels, and phase, respectively.

• When outputting a device or model parameter, it is usually necessary to specify

both the device name and the parameter name, separated by a colon. For
example, the saturation current of a diode model DMOD would be requested as
DMOD:IS. Section 2.1.31.12 on “Device Parameters and Internal Variables”
below gives more details and provides an example.

• The N() syntax is used to access internal solution variables that are not normally
visible from the netlist, such as voltages on internal nodes and/or branch currents
within a given device. The internal solution variables for each Xyce device are
not given in the Reference Guide sections on those devices. However, if the user
runs Xyce -namesfile <filename> <netlist> then Xyce will output into
the first filename a list of all solution variables generated by that netlist.
Section 2.1.31.12 on “Device Parameters and Internal Variables” below gives
more details and provides an example.

132
 • The DNI() and DNO() syntax is used to print out the individual input and output
noise contributions for each noise source within a device. The user can get a
listing of the noise source names for each device in a netlist by running Xyce
-noise_names_file <filename> <netlist>. The Xyce Users’
Guide [1] provides an example.

• If multiple .PRINT lines are given for the same analysis type, the same output
file name, and the same format, the variable lists of all matching .PRINT lines
are merged together in the order found, and the resulting output is the same as if
all the print line variable lists had been specified on a single .PRINT line.

• Attempting to specify multiple .PRINT lines for the same analysis type to the
same file with different specifications of FORMAT is an error.

• Xyce should emit a warning or error message, similar to “Could not open
filename” if: 1) the name of the output file is actually a directory name; or 2) the
output file is in a subdirectory that does not already exist. Xyce will not create
new subdirectories.

• The output filename specified with the -r command line option, to produce raw
file output, should take precedence over a FILE= parameter specified on a
.PRINT line.

• The print statements for some analysis types could result in multiple output files.
For example, .PRINT HB will produce both frequency- and time-domain output,
and place these in different files. The default name of these files is the name of
the netlist followed by a data type suffix, followed by a format-specific extension.

In Xyce, if a FILE option is given to such a print statement, only the “primary”
data for that analysis type is sent to the named file. The secondary data is still
sent to the default file name. This behavior may be subject to change in future
releases.

For analysis types that can produce multiple files, special .PRINT lines have
been provided to allow the user to control the handling of the additional files.
These additional print line specifiers are enumerated in the analysis-specific
sections below.

If one desires that all outputs for a given analysis type be given user-defined file
names, it is necessary to use additional print lines with additional FILE options.
For example, if one uses a FILE option to a .PRINT HB line, only
frequency-domain data will be sent to the named file. To redirect the
time-domain data to a file with a user-defined name, add a .PRINT HB_TD line.
See the individual analysis types below for details of what additional print
statements are available.

2.1.31.1. Print AC Analysis

AC Analysis generates two output files, the primary output is in the frequency domain and the initial
conditions output is in the time domain.

133
Note that when using the .PRINT AC to create the variable list for DC type output, usage of frequency
domain functions like VDB can result in -Inf output being written to the output file. This is easily solved by
defining a .PRINT AC_IC command to specify the correct output for initial condition data.

Homotopy output can also be generated.

Table 2-19. Print AC Analysis Type

Trigger Files Columns/Description
.PRINT AC circuit-file.FD.prn INDEX FREQ
.PRINT AC FORMAT=GNUPLOT circuit-file.FD.prn INDEX FREQ
.PRINT AC FORMAT=SPLOT circuit-file.FD.prn INDEX FREQ
.PRINT AC FORMAT=NOINDEX circuit-file.FD.prn FREQ
.PRINT AC FORMAT=CSV circuit-file.FD.csv FREQ
.PRINT AC FORMAT=RAW circuit-file.raw FREQ
Xyce -a
circuit-file.raw FREQ
.PRINT AC FORMAT=RAW
.PRINT AC FORMAT=TECPLOT circuit-file.FD.dat FREQ
.PRINT AC FORMAT=PROBE circuit-file.csd –
Add .OP To Netlist To Enable AC_IC Output

.PRINT AC_IC circuit-file.TD.prn INDEX TIME

.PRINT AC_IC FORMAT=GNUPLOT circuit-file.TD.prn INDEX TIME
.PRINT AC_IC FORMAT=SPLOT circuit-file.TD.prn INDEX TIME
.PRINT AC_IC FORMAT=NOINDEX circuit-file.TD.prn TIME
.PRINT AC_IC FORMAT=CSV circuit-file.TD.csv TIME
.PRINT AC_IC FORMAT=RAW circuit-file.raw TIME
Xyce -a
circuit-file.raw TIME
.PRINT AC_IC FORMAT=RAW
.PRINT AC_IC FORMAT=TECPLOT circuit-file.TD.dat TIME
.PRINT AC_IC FORMAT=PROBE circuit-file.TD.csd –
Command Line Raw Override Output

Xyce -r raw-file-name raw-file-name All circuit variables printed

Xyce -r raw-file-name -a raw-file-name All circuit variables printed
Additional Output Available

.OP log file Operating point data

.SENS
see Print Sensitivity
.PRINT SENS
.OPTIONS NONLIN CONTINUATION=<method>
see Print Homotopy
.PRINT HOMOTOPY

134
2.1.31.2. Print DC Analysis

DC Analysis generates output based on the format specified by the .PRINT command.
Homotopy and sensitivity output can also be generated.

Table 2-20. Print DC Analysis Type

Trigger Files Columns/Description
.PRINT DC circuit-file.prn INDEX
.PRINT DC FORMAT=GNUPLOT circuit-file.prn INDEX
.PRINT DC FORMAT=SPLOT circuit-file.prn INDEX
.PRINT DC FORMAT=NOINDEX circuit-file.prn –
.PRINT DC FORMAT=CSV circuit-file.csv –
.PRINT DC FORMAT=RAW circuit-file.raw –
Xyce -a
circuit-file.raw –
.PRINT DC FORMAT=RAW
.PRINT DC FORMAT=TECPLOT circuit-file.dat –
.PRINT DC FORMAT=PROBE circuit-file.csd –
Command Line Raw Override Output
Xyce -r raw-file-name raw-file-name All circuit variables
Xyce -r raw-file-name -a raw-file-name All circuit variables
Additional Output Available
.OP log file Operating point data
.SENS
see Print Sensitivity
.PRINT SENS
.OPTIONS NONLIN CONTINUATION=<method>
see Print Homotopy
.PRINT HOMOTOPY

2.1.31.3. Print Harmonic Balance Analysis

HB Analysis generates one output file in the frequency domain and one in the time domain based on the
format specified by the .PRINT command. Additional startup and initial conditions output can be generated
based on .OPTIONS commands.
Note that when using the .PRINT HB to create the variable list for time domain output, usage of frequency
domain functions like VDB can result in -Inf output being written to the output file. This is easily solved by
defining a .PRINT HB_TD, .PRINT HB_IC and .PRINT HB_STARTUP commands to specify the correct
output for the time domain data.
If .STEP is used with HB then the Initial Condition (IC) data will initially be output to a “tmp file” (e.g.,
<netlist-name>.hb_ic.prn.tmp). If that IC data meets the required tolerance then it will be copied to
the end of the <netlist-name>.hb_ic.prn file, and the tmp file will be deleted.

135
Homotopy output can also be generated.

Table 2-21. Print HB Analysis Type

Trigger Files Columns/Description
circuit-file.HB.TD.prn INDEX TIME
.PRINT HB circuit-file.HB.FD.prn INDEX FREQ
circuit-file.hb_ic.prn INDEX TIME
circuit-file.HB.TD.prn INDEX TIME
.PRINT HB FORMAT=GNUPLOT circuit-file.HB.FD.prn INDEX FREQ
circuit-file.hb_ic.prn INDEX TIME
circuit-file.HB.TD.prn INDEX TIME
.PRINT HB FORMAT=SPLOT circuit-file.HB.FD.prn INDEX FREQ
circuit-file.hb_ic.prn INDEX TIME
circuit-file.HB.TD.prn TIME
.PRINT HB FORMAT=NOINDEX circuit-file.HB.FD.prn FREQ
circuit-file.hb_ic.prn TIME
circuit-file.HB.TD.csv TIME
.PRINT HB FORMAT=CSV circuit-file.HB.FD.csv FREQ
circuit-file.hb_ic.csv TIME
circuit-file.HB.TD.dat TIME
.PRINT HB FORMAT=TECPLOT circuit-file.HB.FD.dat FREQ
circuit-file.hb_ic.dat TIME
.PRINT HB_FD circuit-file.HB.FD.prn INDEX FREQ
.PRINT HB_FD FORMAT=GNUPLOT circuit-file.HB.FD.prn INDEX FREQ
.PRINT HB_FD FORMAT=SPLOT circuit-file.HB.FD.prn INDEX FREQ
.PRINT HB_FD FORMAT=NOINDEX circuit-file.HB.FD.prn FREQ
.PRINT HB_FD FORMAT=CSV circuit-file.HB.FD.csv FREQ
.PRINT HB_FD FORMAT=TECPLOT circuit-file.HB.FD.dat FREQ
.PRINT HB_TD circuit-file.HB.TD.prn INDEX TIME
.PRINT HB_TD FORMAT=GNUPLOT circuit-file.HB.TD.prn INDEX TIME
.PRINT HB_TD FORMAT=SPLOT circuit-file.HB.TD.prn INDEX TIME
.PRINT HB_TD FORMAT=NOINDEX circuit-file.HB.TD.prn TIME
.PRINT HB_TD FORMAT=CSV circuit-file.HB.TD.csv TIME
.PRINT HB_TD FORMAT=TECPLOT circuit-file.HB.TD.dat TIME
Startup Period
.OPTIONS HBINT STARTUPPERIODS=<n>
circuit-file.startup.prn INDEX TIME
.PRINT HB_STARTUP
.OPTIONS HBINT STARTUPPERIODS=<n>
circuit-file.startup.prn INDEX TIME
.PRINT HB_STARTUP FORMAT=GNUPLOT
.OPTIONS HBINT STARTUPPERIODS=<n>
circuit-file.startup.prn INDEX TIME
.PRINT HB_STARTUP FORMAT=SPLOT
.OPTIONS HBINT STARTUPPERIODS=<n>
circuit-file.startup.prn TIME
.PRINT HB_STARTUP FORMAT=NOINDEX

136
 Table 2-21. Print HB Analysis Type
Trigger Files Columns/Description
.OPTIONS HBINT STARTUPPERIODS=<n>
.PRINT HB_STARTUP FORMAT=CSV circuit-file.startup.csv TIME
.OPTIONS HBINT STARTUPPERIODS=<n>
.OPTIONS HBINT STARTUPPERIODS=<n>
.PRINT HB_STARTUP FORMAT=TECPLOT circuit-file.startup.dat TIME
.OPTIONS HBINT STARTUPPERIODS=<n>
Initial Conditions
.OPTIONS HBINT SAVEICDATA=1
circuit-file.hb_ic.prn INDEX TIME
.PRINT HB_IC
.OPTIONS HBINT SAVEICDATA=1
circuit-file.hb_ic.prn INDEX TIME
.PRINT HB_IC FORMAT=GNUPLOT
.OPTIONS HBINT SAVEICDATA=1
circuit-file.hb_ic.prn INDEX TIME
.PRINT HB_IC FORMAT=SPLOT
.OPTIONS HBINT SAVEICDATA=1
circuit-file.hb_ic.prn TIME
.PRINT HB_IC FORMAT=NOINDEX
.OPTIONS HBINT SAVEICDATA=1
circuit-file.hb_ic.csv TIME
.PRINT HB_IC FORMAT=CSV
.OPTIONS HBINT SAVEICDATA=1
circuit-file.hb_ic.dat TIME
.PRINT HB_IC FORMAT=TECPLOT
Additional Output Available
.OP log file Operating point data
.SENS
see Print Sensitivity
.PRINT SENS
.OPTIONS NONLIN CONTINUATION=<method>
see Print Homotopy
.PRINT HOMOTOPY

2.1.31.4. Print Noise Analysis

NOISE Analysis generates two output files, the primary output is in the frequency domain and the initial
conditions output is in the time domain.

Table 2-22. Print NOISE Analysis Type

Trigger Files Columns/Description
.PRINT NOISE circuit-file.NOISE.prn INDEX FREQ
.PRINT NOISE FORMAT=GNUPLOT circuit-file.NOISE.prn INDEX FREQ
.PRINT NOISE FORMAT=SPLOT circuit-file.NOISE.prn INDEX FREQ
.PRINT NOISE FORMAT=NOINDEX circuit-file.NOISE.prn FREQ
.PRINT NOISE FORMAT=CSV circuit-file.NOISE.csv FREQ
.PRINT NOISE FORMAT=TECPLOT circuit-file.NOISE.dat FREQ

137
 Table 2-22. Print NOISE Analysis Type
Trigger Files Columns/Description
Additional Output Available
.OP log file Operating point data
.OPTIONS NONLIN CONTINUATION=<method>
see Print Homotopy
.PRINT HOMOTOPY

2.1.31.5. Print Transient Analysis

Transient Analysis generates time domain output based on the format specified by the .PRINT command.
Homotopy and sensitivty output can also be generated.

Table 2-23. Print Transient Analysis Type

Trigger Files Columns/Description
.PRINT TRAN circuit-file.prn INDEX TIME
.PRINT TRAN FORMAT=GNUPLOT circuit-file.prn INDEX TIME
.PRINT TRAN FORMAT=SPLOT circuit-file.prn INDEX TIME
.PRINT TRAN FORMAT=NOINDEX circuit-file.prn TIME
.PRINT TRAN FORMAT=CSV circuit-file.csv TIME
.PRINT TRAN FORMAT=RAW circuit-file.raw TIME
Xyce -a
circuit-file.raw TIME
.PRINT TRAN FORMAT=RAW
.PRINT TRAN FORMAT=TECPLOT circuit-file.dat TIME
.PRINT TRAN FORMAT=PROBE circuit-file.csd –
Command Line Raw Override Output
Xyce -r raw-file-name raw-file-name All circuit variables printed
Xyce -r raw-file-name -a raw-file-name All circuit variables printed
Additional Output Available
.OP log file Operating point data
.SENS
see Print Sensitivity
.PRINT SENS
.OPTIONS NONLIN CONTINUATION=<method>
see Print Homotopy
.PRINT HOMOTOPY

2.1.31.6. Print Homotopy

Homotopy output is generated by the inclusion of the

.OPTIONS NONLIN CONTINUATION=<method> command.

138
 Table 2-24. Print Homotopy
Trigger Files Columns/Description
.OPTIONS NONLIN CONTINUATION=<method>
circuit-file.HOMOTOPY.prn INDEX TIME
.PRINT HOMOTOPY
.OPTIONS NONLIN CONTINUATION=<method>
circuit-file.HOMOTOPY.prn INDEX TIME
.PRINT HOMOTOPY FORMAT=GNUPLOT
.OPTIONS NONLIN CONTINUATION=<method>
circuit-file.HOMOTOPY.prn INDEX TIME
.PRINT HOMOTOPY FORMAT=SPLOT
.OPTIONS NONLIN CONTINUATION=<method>
circuit-file.HOMOTOPY.prn TIME
.PRINT HOMOTOPY FORMAT=NOINDEX
.OPTIONS NONLIN CONTINUATION=<method>
circuit-file.HOMOTOPY.csv TIME
.PRINT HOMOTOPY FORMAT=CSV
.OPTIONS NONLIN CONTINUATION=<method>
circuit-file.HOMOTOPY.dat TIME
.PRINT HOMOTOPY FORMAT=TECPLOT

2.1.31.7. Print Sensitivity

Sensitivity is enabled by inclusion of the

.SENS command.
Steady-state sensitivities (adjoint or direct) and transient direct sensitivities will be handled by the .PRINT
SENS command. Transient adjoint, on the other hand, is handled by the .PRINT TRANADJOINT
command.
For transient sensitivity output, a TIME column will be included for the STD, GNUPLOT, SPLOT,
NOINDEX and CSV formats. For AC sensitivity output, a FREQ column will be included for the STD,
GNUPLOT, SPLOT, NOINDEX and CSV formats.

Table 2-25. Print Sensitivities for .TRAN and .DC

Trigger Files Columns/Description
{𝑜𝑏 𝑗},
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]* d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
circuit-file.SENS.prn
.PRINT SENS d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]

{𝑜𝑏 𝑗},
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]* d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
circuit-file.SENS.prn
.PRINT SENS FORMAT=GNUPLOT d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]

{𝑜𝑏 𝑗},
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]* d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
circuit-file.SENS.prn
.PRINT SENS FORMAT=SPLOT d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]

{𝑜𝑏 𝑗},
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]* d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
circuit-file.SENS.prn
.PRINT SENS FORMAT=NOINDEX d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]

139
 Table 2-25. Print Sensitivities for .TRAN and .DC
Trigger Files Columns/Description
{𝑜𝑏 𝑗},
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]* d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
circuit-file.SENS.csv
.PRINT SENS FORMAT=CSV d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]

{𝑜𝑏 𝑗},
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 ][,𝑝 𝑛 ]* d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
circuit-file.SENS.dat
.PRINT SENS FORMAT=TECPLOT d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]

Table 2-26. Print Sensitivities for .AC

Trigger Files Columns/Description
{𝑜𝑏 𝑗},
.SENS [objvars|acobjfunc]=<𝑜𝑏 𝑗>
d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
+ param=[𝑝 1 1][,𝑝 𝑛 ]* circuit-file.FD.SENS.prn
d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]
.PRINT SENS

{𝑜𝑏 𝑗},
.SENS [objvars|acobjfunc]=<𝑜𝑏 𝑗>
d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
+ param=[𝑝 1 1][,𝑝 𝑛 ]* circuit-file.FD.SENS.prn
d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]
.PRINT SENS FORMAT=GNUPLOT

{𝑜𝑏 𝑗},
.SENS [objvars|acobjfunc]=<𝑜𝑏 𝑗>
d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
+ param=[𝑝 1 1][,𝑝 𝑛 ]* circuit-file.FD.SENS.prn
d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]
.PRINT SENS FORMAT=SPLOT

{𝑜𝑏 𝑗},
.SENS [objvars|acobjfunc]=<𝑜𝑏 𝑗>
d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
+ param=[𝑝 1 1][,𝑝 𝑛 ]* circuit-file.FD.SENS.prn
d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]
.PRINT SENS FORMAT=NOINDEX

{𝑜𝑏 𝑗},
.SENS [objvars|acobjfunc]=<𝑜𝑏 𝑗>
d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
+ param=[𝑝 1 1][,𝑝 𝑛 ]* circuit-file.FD.SENS.csv
d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]
.PRINT SENS FORMAT=CSV

{𝑜𝑏 𝑗},
.SENS [objvars|acobjfunc]=<𝑜𝑏 𝑗>
d_{𝑜𝑏 𝑗}/d_𝑝 1 _[dir|adj],
+ param=[𝑝 1 ][,𝑝 𝑛 ]* circuit-file.FD.SENS.dat
d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _[dir|adj]
.PRINT SENS FORMAT=TECPLOT

Table 2-27. Print Transient Adjoint Sensitivities

Trigger Files Columns/Description
{𝑜𝑏 𝑗}, d_{𝑜𝑏 𝑗}/d_𝑝 1 _adj,
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]*
circuit-file.TRADJ.prn d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _adj
.PRINT TRANADJOINT

140
 Table 2-27. Print Transient Adjoint Sensitivities
Trigger Files Columns/Description
{𝑜𝑏 𝑗}, d_{𝑜𝑏 𝑗}/d_𝑝 1 _adj,
.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]*
circuit-file.TRADJ.prn d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _adj
.PRINT TRANADJOINT FORMAT=NOINDEX

{𝑜𝑏 𝑗}, d_{𝑜𝑏 𝑗}/d_𝑝 1 _adj,

.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 1][,𝑝 𝑛 ]*
circuit-file.TRADJ.csv d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _adj
.PRINT TRANADJOINT FORMAT=CSV

{𝑜𝑏 𝑗}, d_{𝑜𝑏 𝑗}/d_𝑝 1 _adj,

.SENS objfunc=<𝑜𝑏 𝑗> param=[𝑝 1 ][,𝑝 𝑛 ]*
circuit-file.TRADJ.dat d_{𝑜𝑏 𝑗}/d_𝑝 𝑛 _adj
.PRINT TRANADJOINT FORMAT=TECPLOT

2.1.31.8. Print Embedded Sampling Analysis

EMBEDDEDSAMPLING Analysis generates one output file, with the information for each output variable
grouped in a set of contiguous columns, based on the format specified by the .PRINT command. The
arguments OUTPUT_SAMPLE_STATS and OUTPUT_ALL_SAMPLES are specific to .PRINT ES lines. Section
2.1.5 has more details on their usage.
For a transient analysis, a TIME column will also be included for the STD, GNUPLOT, SPLOT, NOINDEX
and CSV formats. The TIME variable will also be included in the TECPLOT format for that case.

Table 2-28. Print EMBEDDEDSAMPLING Analysis Type

Trigger Files Columns/Description
.PRINT ES circuit-file.ES.prn INDEX
.PRINT ES FORMAT=GNUPLOT circuit-file.ES.prn INDEX
.PRINT ES FORMAT=SPLOT circuit-file.ES.prn INDEX
.PRINT ES FORMAT=NOINDEX circuit-file.ES.prn –
.PRINT ES FORMAT=CSV circuit-file.ES.csv –
.PRINT ES FORMAT=TECPLOT circuit-file.ES.dat –

2.1.31.9. Print Intrusive PCE Analysis

PCE Analysis generates one output file, with the information for each output variable grouped in a set of
contiguous columns, based on the format specified by the .PRINT command. The arguments
OUTPUT_SAMPLE_STATS and OUTPUT_ALL_SAMPLES are specific to .PRINT PCE lines. Section 2.1.27 has
more details on their usage.
For a transient analysis, a TIME column will also be included for the STD, GNUPLOT, SPLOT, NOINDEX
and CSV formats. The TIME variable will also be included in the TECPLOT format for that case.

141
 Table 2-29. Print PCE Analysis Type
Trigger Files Columns/Description
.PRINT PCE circuit-file.PCE.prn INDEX
.PRINT PCE FORMAT=GNUPLOT circuit-file.PCE.prn INDEX
.PRINT PCE FORMAT=SPLOT circuit-file.PCE.prn INDEX
.PRINT PCE FORMAT=NOINDEX circuit-file.PCE.prn –
.PRINT PCE FORMAT=CSV circuit-file.PCE.csv –
.PRINT PCE FORMAT=TECPLOT circuit-file.PCE.dat –

2.1.31.10. Parameter Stepping

During parameter stepping, enabled with the .STEP command, the output generated by each of analysis
types varies. Generally the FORMAT indicates this variation, however some combinations of analysis and
format can result in additional variation.
The following table lists how the output differs for each analysis type and format.

Print Type Format Description

AC STD 1, 3, 4, 11, 12, 13
AC GNUPLOT 1, 3, 4, 11, 12, 13, 20
AC SPLOT 1, 3, 4, 11, 12, 13, 21
AC CSV 4, 11
AC PROBE 16
AC TECPLOT 4, 12, 13, 18
AC RAW 19
AC RAW (Xyce -a) 19
AC_IC STD 1, 4, 11, 12, 13
AC_IC GNUPLOT 1, 4, 11, 12, 13, 20
AC_IC SPLOT 1, 4, 11, 12, 13, 21
AC_IC CSV 4, 11
AC_IC PROBE 16
AC_IC TECPLOT 12, 13, 18
AC_IC RAW 19
AC_IC RAW (Xyce -a) 19
DC STD 1, 11, 12
DC GNUPLOT 1, 11, 12, 20
DC SPLOT 1, 11, 12, 21
DC CSV 11
DC PROBE 17
DC TECPLOT 4, 12, 13, 18

142
Print Type Format Description
DC RAW 19
DC RAW (Xyce -a) 19
HB_TD STD 1, 2, 4, 11, 12, 13
HB_TD GNUPLOT 1, 2, 4, 11, 12, 13, 20
HB_TD SPLOT 1, 2, 4, 11, 12, 13, 21
HB_TD CSV 11
HB_TD TECPLOT 12, 13, 18
HB_FD STD 1, 3, 4, 11, 12, 13
HB_FD GNUPLOT 1, 3, 4, 11, 12, 13, 20
HB_FD SPLOT 1, 3, 4, 11, 12, 13, 21
HB_FD CSV 4, 11
HB_FD TECPLOT 4, 12, 13, 18
HB_IC STD 1, 2, 4, 11, 12, 13
HB_IC GNUPLOT 1, 2, 4, 11, 12, 13, 20
HB_IC SPLOT 1, 2, 4, 11, 12, 13, 21
HB_IC CSV 11
HB_IC TECPLOT 12, 13, 18
HB_STARTUP STD 1, 2, 4, 11, 12, 13
HB_STARTUP GNUPLOT 1, 2, 4, 11, 12, 13, 20
HB_STARTUP SPLOT 1, 2, 4, 11, 12, 13, 21
HB_STARTUP CSV 11
HB_STARTUP TECPLOT 12, 13, 18
TRAN STD 1, 2, 11, 12
TRAN GNUPLOT 1, 2, 11, 12, 20
TRAN SPLOT 1, 2, 11, 12, 21
TRAN CSV 2, 11
TRAN PROBE 17
TRAN TECPLOT 2, 4, 12, 13, 18
TRAN RAW 2, 19
TRAN RAW (Xyce -a) 2, 19
Specialized Output Commands
HOMOTOPY STD 1, 2, 4, 11, 15
HOMOTOPY GNUPLOT 1, 2, 4, 11, 15, 20
HOMOTOPY SPLOT 1, 2, 4, 11, 15, 21
HOMOTOPY CSV 2, 11
HOMOTOPY PROBE 17
HOMOTOPY TECPLOT 2, 4, 15, 18

143
 Print Type Format Description
SENSITIVITY STD 1, 2, 11, 14
SENSITIVITY GNUPLOT 1, 2, 11, 14, 20
SENSITIVITY SPLOT 1, 2, 11, 14, 21
SENSITIVITY CSV 2, 11
SENSITIVITY TECPLOT 2, 14, 18

Description
1 INDEX column added to output variable list
2 TIME column added to output variable list
3 FREQ column added to output variable list
Frequency domain data written as Re(𝑣𝑎𝑟) and
4
Im(𝑣𝑎𝑟)
11 INDEX resets to zero at start of each .STEP
Prints ’End of Xyce(TM) Parameter Sweep’ at end of
12
.STEP simulation
Prints ’End of Xyce(TM) Simulation’ at end of
13
non-.STEP simulation
Prints ’End of Xyce(TM) Sensitivity Simulation’ at
14
end of simulation
Prints ’End of Xyce(TM) Homotopy Simulation’ at end
15
of simulation
16 Two ’#;’ at the end of each .STEP (BUG)
17 One ’#;’ at end of each .STEP
New ZONE for each .STEP, and AUXDATA for each .STEP
18
parameter
Prints ’Plotname: Step Analysis: Step 𝑠 of 𝑛
19
params’ at the start of each .STEP
Inserts two blank lines before the data for steps
20
1,2,3,... where the first step is step 0
Inserts one blank line before the data for steps
21
1,2,3,... where the first step is step 0

2.1.31.11. Print Wildcards

Wildcards are supported on .PRINT lines, as described below. In particular, V(*) will print all of the node
voltages in the circuit for all analysis modes. The P(*) and W(*) wildcards are supported for analysis
modes (TRAN and DC) that support power calculations.
For TRAN and DC analysis modes, I(*) will print all of the currents. This includes both solution
variables, which generally means those associated with voltage sources and inductors that are not coupled
through a mutual inductance device, and the lead currents associated with most other devices. For TRAN
and DC, the I(*) wildcard also supports lead currents for the multi-terminal J, M and Z devices via IB(*),
ID(*), IG(*) and IS(*), and for the multi-terminal Q device via IB(*), IC(*), IE(*) and IS(*). The

144
IE(*) wildcard is also supported for SOI and CMG devices. A request for I(*) will not return any of the
lead currents for J, M, Q or Z devices. Wildcards of the form I1(*), that use numerical designators, are
only supported for the T and YGENEXT devices. Finally, as an example, a request for IC(*) in a netlist that
does not contain any Q devices will be silently ignored.

For AC and NOISE analysis modes, the I(*) operator will only output the branch currents, since lead
currents are not supported for those two analysis modes. The VR(*), VI(*), VP(*), VM(*), VDB(*),
IR(*), II(*), IP(*), IM(*) and IDB(*) wildcards are also supported for these two analysis modes.

There is also support for the * character (meaning “zero or more characters”) and the ? character (meaning
“any one character”) in more complex wildcards, where the * and/or ? characters can be in any positions in
the wildcard specification. For example, V(X1*) will output the voltage at all nodes in subcircuit X1 for all
analysis modes. As another example, V(1?) will output the voltage at all nodes that have two-character
names that start with the character 1. These more complex wildcards should work for all supported voltage
operators.

Similarly, P(X1*) or W(X1*) will output the power for all devices, that support power calculations, in
subcircuit X1. Devices that don’t support power calculations will be silently omitted. Alternately, P(R?) or
W(R?) will output the power for all resistors that have two-character names.

More complex wildcards are also supported for all valid current operators. The caveats are that for DC and
TRAN analyses, the wildcard will include both branch and lead currents. For AC and NOISE analyses, the
wildcard will only include branch currents.

2.1.31.12. Device Parameters and Internal Variables

This subsection describes how to print out device parameters and device internal variables, via a simple V-R
circuit example. In particular, the example given below gives illustrative examples of how to print out the
voltage at a node (V(1)), the current through a device (I(V1)), the current through a device using using an
internal solution variable (N(V1_branch)), a device parameter (R1:R) and the power dissipated by a device
(P(R1)). It also shows how device parameters and internal variables can be used in a Xyce expression.

* filename is example.cir
.DC V1 1 2 1
V1 1 0 1
R1 1 0 2
.PRINT DC FORMAT=NOINDEX PRECISION=2 WIDTH=8
+ V(1) I(V1) N(V1_branch) R1:R P(R1) {R1:R*N(V1_branch)*I(V1)}
.END

The Xyce output would then be (where the NOFORMAT, WIDTH and PRECISION arguments were used mainly
to format the example output for this guide):

V(1) I(V1) N(V1_BRANCH) R1:R P(R1) {R1:R*N(V1_BRANCH)*I(V1)}

1.00e+00 -5.00e-01 -5.00e-01 2.00e+00 5.00e-01 5.00e-01
2.00e+00 -1.00e+00 -1.00e+00 2.00e+00 2.00e+00 2.00e+00

145
The internal solution variables for each Xyce device are typically not given in the Reference Guide sections
on those devices. However, if for the example given above, the user runs Xyce -namesfile
example_names example.cir then the file example_names would contain a list of the two solution
variables that are accessible with the N() syntax on a .PRINT line. In this simple example, they are the
voltage at Node 1 and the branch current through the voltage source V1. If V1 was in a subcircuit then the
example_names file would have shown the “fully-qualified” device name, including the subcircuit
names.

HEADER
0 v1_branch
1 1

Additional (and more useful) examples for using the N() syntax to print out:
• The 𝑀, 𝑅, 𝐵 and 𝐻 internal variables for mutual inductors are given in Section 2.3.6. This includes
an example where the mutual inductor is in a sub-circuit.
• The 𝑔𝑚 (tranconductance), 𝑉𝑡 ℎ , 𝑉𝑑𝑠 , 𝑉𝑔𝑠 , 𝑉𝑏𝑠 , and 𝑉𝑑𝑠𝑎𝑡 internal variables for the BSIM3 and BSIM4
models for the MOSFET are given in Section 2.3.20.
In these two cases, only the 𝑀 and 𝑅 variables for the mutual inductors are actually solution variables.
However, the -namesfile approach can still be used to determine the fully-qualified Xyce device names
required to use the N() syntax.

146
2.1.32. .RESULT (Print results)

Outputs the value of user-specified expressions at the end of a simulation.

General Form .RESULT {output variable}

Examples .RESULT {V(a)}

.RESULT {V(a)+V(b)}

Comments The .RESULT line must use an expression. The line .RESULT V(a) will result in a
parse error.
Each .RESULT line must have only one expression. Multiple .RESULT lines can be
used though to output multiple columns in the output .res file.
Xyce will not produce output for .RESULT statements if there are no .STEP statements
in the netlist.

2.1.32.1. Example Netlist

.RESULT lines can be combined with .STEP lines to output the ending values of multiple simulation runs in
one .res file, as shown in the following usage example. The resultant .res file will have four lines that
give the final values of the expressions {v(b)} and {v(b)*v(b)/2} at time=0.75 seconds for all four
requested combinations of R2 and v_amplitude.

Simple Example of .RESULT capability with .STEP

R1 a b 10.0
R2 b 0 2.0

.GLOBAL_PARAM v_amplitude=2.0
Va a 0 sin (5.0 {v_amplitude} 1.0 0.0 0.0)

.PRINT TRAN v(b) {v(b)*v(b)/2}

.TRAN 0 0.75

.STEP R2 1.0 2.0 1.0

.STEP v_amplitude 1.0 2.0 1.0

.RESULT {v(b)}
.RESULT {v(b)*v(b)/2}

.END

147
2.1.33. .SAMPLING (Sampling UQ Analysis)

Calculates a full analysis (.DC, .TRAN, .AC, etc.) over a distribution of parameter values. Sampling operates
similarly to .STEP, except that the parameter values are generated from random distributions rather than
sweeps. If used in conjunction with projection-based PCE methods, then the sample points are not based on
random samples. Instead they are based on the quadrature points.

General Form .SAMPLING

+ param=<parameter name>,[parameter name]*
+ type=<parameter type>,[parameter type]*
+ means=<mean>,[mean]*
+ std_deviations=<standard deviation>,[standard deviation]*

Examples .SAMPLING
+ param=R1
+ type=normal
+ means=3K
+ std_deviations=1K

.SAMPLING
+ param=R1,R2
+ type=uniform,uniform
+ lower_bounds=1K,2K
+ upper_bounds=5K,6K

.SAMPLING
+ useExpr=true

.options SAMPLES numsamples=10000

.options SAMPLES numsamples=25000

+ OUTPUTS={R1:R},{V(1)}
+ SAMPLE_TYPE=MC

.options SAMPLES numsamples=1000

+ MEASURES=maxSine
+ SAMPLE_TYPE=LHS

.options samples numsamples=30

+ covmatrix=1e6,1.0e-3,1.0e-3,4e-14
+ OUTPUTS={V(1)},{R1:R},{C1:C}

Arguments and
Options param
Names of the parameters to be sampled. This may be any of the parameters
that are valid for .STEP, including device instance, device model, or global

148
 parameters. If more than one parameter, then specify as a comma-separated
list.
type
Distribution type for each parameter. This may be uniform or normal. If
more than one parameter, then specify as a comma-separated list.
means
If using normal distributions, the mean for each parameter must be specified.
If more than one parameter, then specify as a comma-separated list.
std_deviations
If using normal distributions, the standard deviation for each parameter must
be specified. If more than one parameter, then specify as a comma-separated
list.
lower_bounds
If using uniform distributions, the lower bound must be specified. This is
optional for normal distributions. If used with normal distributions, may
alter the mean and standard deviation. If more than one parameter, then
specify as a comma-separated list.
upper_bounds
If using uniform distributions, the upper bound must be specified. This is
optional for normal distributions. If used with normal distributions, may
alter the mean and standard deviation. If more than one parameter, then
specify as a comma-separated list.
useExpr
If this argument is set to true, then the sampling algorithm will set up
random inputs from expression operators such as AGAUSS and AUNIF. In this
case it will also ignore the list of parameters on the .SAMPLING command
line. For a complete description of expression-based random operators, see
the expression documentation in section 2.2.

Comments In addition to the .SAMPLING command, this analysis requires a .options SAMPLES
command as well. The .SAMPLING command specifies parameters and their attributes,
either using the useExpr option, or with comma-separated lists. The
.options SAMPLES command specifies analysis options, including the number of
samples, the type of sampling (LHS or MC) and the outputs and/or measures for
which to compute statistics. This line also allows one to specify a non-intrusive
Polynomial Chaos Expansion (PCE) method (either regression or projection PCE). To
see the details of the .options SAMPLES command , see table 2-10.
On the .SAMPLING command line, if not using useExpr, parameters and their
attributes must be specified using comma-separated lists. The comma-separated lists
must all be the same length.

149
2.1.34. .SAVE (Save operating point conditions)

Stores the operating point of a circuit in the specified file for use in subsequent simulations. The data may
be saved as .IC or .NODESET lines.

General Form .SAVE [TYPE=<IC|NODESET>] [FILE=<filename>] [LEVEL=<all|none>]

+ [TIME=<save_time>]

Examples .SAVE TYPE=IC FILE=mycircuit.ic

.SAVE TYPE=NODESET FILE=myothercircuit.ic

.include mycircuit.ic

Comments The file created by .SAVE will contain .IC or .NODESET lines containing all the
voltage node values at the DC operating point of the circuit. The default TYPE is
NODESET. The default filename is netlist.cir.ic.
The resulting file may be used in subsequent simulations to obtain quick DC
convergence simply by including it in the netlist, as in the third example line above.
Xyce has no corresponding .LOAD statement.
The LEVEL parameter is included for compatibility with HSPICE netlists. If none is
specified, then no save file is created. The default LEVEL is all.
TIME is also an HSPICE compatibility parameter. This is unsupported in Xyce. Xyce
outputs the save file only at time=0.0.

150
2.1.35. .SENS (Compute DC, AC or transient sensitivities)

Computes sensitivies for a user-specificed objective function with respect to a user-specified list of circuit
parameters.

General Form .SENS objfunc=<output expression(s)> param=<circuit parameter(s)>

Examples .SENS objfunc={0.5*(V(B)-3.0)**2.0} param=R1:R,R2:R

.options SENSITIVITY direct=1 adjoint=1

.SENS objfunc={I(VM)},{V(3)*V(3)} param= Q2N2222:bf

.param RES=1k
.SENS objfunc={RES*V(3)*V(3)} param=C1:C

.param res=2
.func powerTestFunc(I) {res*I*I}
.SENS objfunc={powerTestFunc(I(V1))} param=R1:R

.global_param res=2
.SENS objfunc={res*I(V1)} param=R1:R

.global_param res=3.0k
.SENS objfunc={res*I(V1)} param=res

* AC example using objvars

.sens objvars=2,3 param=r1:r,c1:c,v1:acmag

* AC example using acobjfunc

.sens acobjfunc={2.0*V(2)},{I(VM)} param=r1:r,c1:c

Comments This capability can be applied to either DC, transient or AC analysis. Both direct and
adjoint sensitivities are supported. The user can optionally request either direct or
adjoint sensitivities, or both.
Although Xyce will allow the user to specify both direct and adjoint, one would
generally not choose to do both. The best choice of sensitivity method depends on the
problem. For problems with a small number of parameters, and (possibly) lots of
objective functions, then the direct method is a more efficient choice. For problems
with large numbers of parameters, but a small number of objective functions, the
adjoint method is more efficient.
For all variants of sensitivity analysis, it is necessary to specify circuit parameters on
the .SENS line in a comma-separated list. Unlike the SPICE version, this capability
will not automatically use every parameter in the circuit. It is also necessary for all
variations of sensitivity analysis to specify at least one objective function. This
capability will not assume any particular objective function. Also, it is possible to
specify multiple objective functions, in a comma-separated list.

151
As noted, for transient analysis, both types of sensitivities are supported. Direct
sensitivities are computed at each time step during the forward calculation. Transient
adjoint sensitivities, in contrast, must be computed using a reverse time integration
method. The reverse time integration must be performed after the original forward
calculation is complete. As such, transient adjoint sensitivity calculations can be
thought of as a post-processing step. One consequence of this is that transient adjoint
output must be specified using the .PRINT TRANADJOINT type, rather than the
.PRINT SENS type.
If transient adjoints are specified, the default behavior for the capability is for a
transient sensitivity calculation be performed for each time step, even if the forward
transient simulation consists of millions of steps. For adjoint calculations, this can be
problematic, as adjoint methods (noted above) are not very efficient when applied to
problems with a large number of objective functions. Each time step, from the point of
view of transient adjoints, is effectively a separate objective function. As such, this
isn’t the best use of adjoints. One can specify a list of time points for which to compute
transient adjoint sensitivities. For many practical problems, the sensitivies at only one
or a handful of points is needed, so this is a good way to mitigate the computational
cost of adjoints. The Xyce Users’ Guide [1] provides an example.
If performing a sensitivity calculation with AC analysis, there are two options for the
specification of the objective function. These options are both different from the DC
and TRAN method. Instead of specifying objective functions with the parameter
objfunc, one should either use objvars or acobjfunc. The parameter objvars
should be followed by a comma separated list of voltage nodes. The parameter
acobjfunc should be followed by a comma separated list of objective functions. It is
also possible to use both specifications in the same netlist.

152
2.1.36. .STEP (Step Parametric Analysis)

Calculates a full analysis (.DC, .TRAN, .AC, etc.) over a range of parameter values. This type of analysis is
very similar to .DC analysis. Similar to .DC analysis, .STEP supports sweeps which are linear, decade
logarithmic, octave logarithmic, a list of values, or over a multivariate data table.

LIN Linear sweep

The sweep variable is swept linearly from the starting to the ending value.

OCT Sweep by octaves

The sweep variable is swept logarithmically by octaves.

DEC Sweep by decades

The sweep variable is swept logarithmically by decades.

LIST Sweep over specified values

The sweep variable is swept over an enumerated list of values.

DATA Sweep over table of multivariate values

The sweep variables are swept over the rows of a table.

2.1.36.1. Linear Sweeps

General Form .STEP [LIN] <parameter name> <initial> <final> <step>

Examples .STEP R1 45 50 5
.STEP V1 20 10 -1
.STEP LIN V1 20 10 -1
.STEP TEMP -45 -55 -10
.STEP C101:C 45 50 5
.STEP DLEAK:IS 1.0e-12 1.0e-11 1.0e-12

.global_param v1_val=10
V1 1 0 DC {v1_val}
.STEP v1_val 20 10 -1

.param v2_val=10
V2 2 0 DC {v2_val}
.STEP v2_val 20 10 -1

.data table
+ c1 r1
+ 1e-8 1k
+ 2e-8 0.5k
+ 3e-8 0.25k
.enddata
.STEP data=table

153
Arguments and
Options parameter name
Name of the parameter to be swept. This may be the special parameter name
TEMP (the ambient simulation temperature), a device name, device instance
or model parameter name, or global parameter name as defined in a
.global_param or globally-scoped .param statement.
If a device name is given, the primary parameter for that device is taken as
the parameter; in the first two examples above, the primary parameters of the
devices R1 and V1 are stepped (resistance and DC voltage, respectively).
The C, L and I devices are then the other devices with primary parameters,
which are the capacitance, inductance and DC current, respectively.
To specify a device instance parameter other than the device’s primary
parameter, or if the device has no primary parameter, use the syntax
<device name>:<parameter name>, as in the fourth example above.
To sweep a device model parameter, use the syntax <model
name>:<parameter name>, as in the fifth example above.
initial
Initial value for the parameter.
final
Final value for the parameter.
step
Value that the parameter is incremented at each step.

Comments For linear sweeps, the LIN keyword is optional.

STEP parameter analysis will sweep a parameter from its initial value to its final value,
at increments of the step size. At each step of this sweep, it will conduct a full analysis
(.DC, .TRAN, .AC, etc.) of the circuit.
The specification is similar to that of a .DC sweep, except that unlike .DC, only one
parameter may be swept on each .STEP line. Multiple .STEP lines may be specified,
forming nested step loops. The variables will be stepped in order such that the first
.STEP line that appears in the netlist will be the innermost loop, and the last .STEP
line will be the outermost.
Output, as designated by a .PRINT statement, is slightly more complicated in the case
of a .STEP simulation. If the user has specified a .PRINT line in the input file, Xyce
will output two files. All steps of the sweep will be output to a single file as usual, but
with the results of each step appearing one after another with the “Index” column
starting over at zero. Additionally, a file with a “.res” suffix will be produced
indicating what parameters were used for each iteration of the step loops; this file will
always be in columnar text format, irrespective of any FORMAT= option specified on
.PRINT lines. If .RESULT lines (see section 2.1.32) appear in the netlist, the “.res” file
will also contain columns for each expression given on the .RESULT lines, and the
value of the result expression will be printed for each step taken.

154
 Note that analysis lines in Xyce do not currently support use of expressions to define
their parameters (e.g., end times for .TRAN analysis, or fundamental frequencies for
.HB analysis), and so it is not possible to use stepped parameters to vary how the
analysis will be run at each step. If each step requires different analysis parameters,
this would have to be accomplished by performing separate runs of Xyce.
If the stop value is smaller than the start value, the step value should be negative. If a
positive step value is given in this case, only a single point (at the start value) will be
performed, and a warning will be emitted.

2.1.36.2. Decade Sweeps

General Form .STEP DEC <sweep variable name> <start> <stop> <points>

Examples .STEP DEC VIN 1 100 2

.STEP DEC R1 100 10000 3
.STEP DEC TEMP 1.0 10.0 3

Comments The stop value should be larger than the start value. If a stop value smaller than the
start value is given, only a single point at the start value will be performed, and a
warning will be emitted. The points value must be an integer.

2.1.36.3. Octave Sweeps

General Form .STEP OCT <sweep variable name> <start> <stop> <points>

Examples .STEP OCT VIN 0.125 64 2

.STEP OCT TEMP 0.125 16.0 2
.STEP OCT R1 0.015625 512 3

Comments The stop value should be larger than the start value. If a stop value smaller than the
start value is given, only a single point at the start value will be performed, and a
warning will be emitted. The points value must be an integer.

2.1.36.4. List Sweeps

General Form .STEP <sweep variable name> LIST <val> <val> <val>...

Examples .STEP VIN LIST 1.0 2.0 10. 12.0

.STEP TEMP LIST 8.0 21.0

2.1.36.5. Data Sweeps

155
General Form .STEP DATA=<data table name>

Examples .STEP data=resistorValues

.data resistorValues
+ r1 r2
+ 8.0000e+00 4.0000e+00
+ 9.0000e+00 4.0000e+00
.enddata

156
2.1.37. .SUBCKT (Subcircuit)

The .SUBCKT statement begins a subcircuit definition by giving its name, the number and order of its nodes
and the names and default parameters that direct its behavior. The .ENDS statement signifies the end of the
subcircuit definition. See Section 2.3.33 for more information on using subcircuits with the X device.

General Form .SUBCKT <name> [node]*

+ [PARAMS:] [<name>=<value>]*
...
.ENDS

Examples .SUBCKT OPAMP 10 12 111 112 13

...
.ENDS

.SUBCKT FILTER1 INPUT OUTPUT PARAMS: CENTER=200kHz,

+ BANDWIDTH=20kHz
...
.ENDS

.SUBCKT PLRD IN1 IN2 IN3 OUT1

+ PARAMS: MNTYMXDELY=0 IO_LEVEL=1
...
.ENDS

.SUBCKT 74LS01 A B Y
+ PARAMS: MNTYMXDELY=0 IO_LEVEL=1
...
.ENDS

Arguments and
Options
name
The name used to reference a subcircuit.
node
An optional list of nodes. This is not mandatory since it is feasible to define a
subcircuit without any interface nodes.

PARAMS:
Optional keyword that precedes the list of subcircuit parameters. Parameters
specified on the subcircuit instance line are treated as being local to
individual subcircuit instances. They can be used inside a subcircuit in the
same manner as a .param inside the subcircuit definition. Parameters defined
on the instance line override identically named parameters in the subcircuit
definition.

157
Comments A subcircuit designation ends with a .ENDS command. The entire netlist between
.SUBCKT and .ENDS is part of the definition. Each time the subcircuit is called via an
X device, the entire netlist in the subcircuit definition replaces the X device.
There must be an equal number of nodes in the subcircuit call and in its definition. As
soon as the subcircuit is called, the actual nodes (those in the calling statement)
substitute for the argument nodes (those in the defining statement).
Node zero cannot be used in this node list, as it is the global ground node.
Subcircuit references may be nested to any level. Subcircuits definitions may also be
nested; a .SUBCKT statement and its closing .ENDS may appear between another
.SUBCKT/.ENDS pair. A subcircuit defined inside another subcircuit definition is local
to the outer subcircuit and may not be used at higher levels of the circuit netlist.
Subcircuits should include only device instantiations and possibly these statements:
• .MODEL (model definition)
• .PARAM (parameter)
• .FUNC (function)
Models, parameters, and functions defined within a subcircuit are scoped to that
definition. That is they are only accessible within the subcircuit definition in which
they are included. Further, if a .MODEL, .PARAM or a .FUNC statement is included in
the main circuit netlist, it is accessible from the main circuit as well as all subcircuits.
Node, device, and model names are scoped to the subcircuit in which they are defined.
It is allowable to use a name in a subcircuit that has been previously used in the main
circuit netlist. When the subcircuit is flattened (expanded into the main netlist), all of
its names are given a prefix via the subcircuit instance name. For example, Q17
becomes X3:Q17 after expansion. After expansion, all names are unique. The single
exception occurs in the use of global node names, which are not expanded.
Additional illustative examples of scoping are given in the “Working with Subcircuits
and Models” section of the Xyce Users’ Guide [1] . Those examples apply to models
and functions also.

158
2.1.38. .TRAN (Transient Analysis)

Calculates the time-domain response of a circuit for a specified duration.

General Form .TRAN <initial step value> <final time value>

+ [<start time value> [<step ceiling value>]] [NOOP] [UIC]
+ [{schedule( <time>, <maximum time step>, ... )}]

Examples .TRAN 1us 100ms

.TRAN 1ms 100ms 0ms .1ms
.TRAN 0 2.0e-3 {schedule( 0.5e-3, 0, 1.0e-3, 1.0e-6, 2.0e-3, 0 )}

.param initialStep=1ms, tstop=100ms, tstart=0ms, dtmax=0.1ms

.TRAN {initialStep} {tstop} {tstart} {dtmax}

Arguments and
Options
initial step value
Used to calculate the initial time step (see below).

final time value

Sets the end time (duration) for the analysis.

start time value

Sets the time at which output of the simulation results is to begin. Defaults to
zero.
step ceiling value
Sets a maximum time step. Defaults to ((final time value)-(start time
value))/10, unless there are breakpoints (see below).

NOOP or UIC
These two options are synonyms which specify that no operating point
calculation is to be performed, and that the specified initial condition (from
.IC lines or capacitor “IC” parameters) should be used as the transient initial
condition instead. Unspecified values are set to zero. Finally, the .IC
capability can only set voltage values, not current values.

schedule(<time>, <maximum time step>, ...)

Specifies a schedule for maximum allowed time steps. The list of arguments,
𝑡0 , Δ𝑡0 , 𝑡1 , Δ𝑡1 , etc. implies that a maximum time step of Δ𝑡0 will be used
while the simulation time is greater than or equal to 𝑡0 and less than 𝑡1 . A
maximum time step of Δ𝑡1 will be used when the simulation time is greater
or equal to than 𝑡1 and less than 𝑡2 . This sequence will continue for all pairs
of 𝑡𝑖 , Δ𝑡𝑖 that are given in the {schedule()}. If Δ𝑡 is zero or negative, then no
maximum time step is enforced (other than hardware limits of the host
computer).

159
Comments The transient analysis calculates the circuit’s response over an interval of time
beginning with TIME=0 and finishing at <final time value>. Use a .PRINT (print)
statement to get the results of the transient analysis.
Before calculating the transient response Xyce computes a bias point for the circuit
that is different from the regular bias point. This is necessary because at the start of a
transient analysis, the independent sources can have different values than their DC
values. Specifying NOOP on the .TRAN line causes Xyce to begin the transient analysis
without performing the usual bias point calculation.
The time integration algorithms within Xyce use adaptive time-stepping methods that
adjust the time-step size according to the activity in the analysis. The default ceiling
for the internal time step is (<final time value>-<start time value>)/10.
This default ceiling value is automatically adjusted if breakpoints are present, to
ensure that there are always at least 10 time steps between breakpoints. If the user
specifies a ceiling value, however, it overrides any internally generated ceiling values.
Xyce is not strictly compatible with SPICE in its use of the values on the .TRAN line.
In SPICE, the first number on the .TRAN line specifies the printing interval. In Xyce,
the first number is the <initial step value>, which is used in determining the
initial step size. The actual initial step size is chosen to be the smallest of three
quantities: the <inital step value>, the <step ceiling value>, or 1/200th of
the time until the next breakpoint.
The third argument to .TRAN simply determines the earliest time for which results are
to be output. Simulation of the circuit always begins at TIME=0 irrespective of the
setting of <start time value>.

160
2.1.39. Miscellaneous Commands

2.1.39.1. * (Comment)

A netlist comment line. Whitespace at the beginning of a line is also interpreted as a comment unless it is
followed by a + symbol, in which case it treats the line as a continuation.

2.1.39.2. ; (In-line Comment)

Add a netlist in-line comment.

2.1.39.3. + (Line Continuation)

Continue the text of the previous line.

161
2.2. Expressions

Xyce supports use of mathematical expressions in several contexts:

• for the values of device instance and model parameters.
• in definition of parameters in .PARAM and .GLOBAL_PARAM statements.
• for output on .PRINT lines.
In all contexts where expressions are allowed, it is best practice to enclose them in curly braces ({}). For
netlist compatibility with other simulators, expressions may be enclosed in single quotation marks instead
(’). Also, in some circumstances (such as .param and .global_param expressions), Xyce will accept
expressions that are not surrounded by either braces or single quotes. However, it is recommended that the
braces be used in netlists written specifically for Xyce, as this will be the most reliable option.
The expression package in Xyce supports all standard arithmetic operators, trigonometric functions, a
collection of arithmetic functions, and some functions to mimic the pulse, sine, exp, and sffm
time-dependent functions in the independent current and voltage sources. These functions are listed in
tables 2-33 and 2-34.

Operators

Table 2-32. Operators

Class of
Operator Meaning
Operator
arithmetic + addition or string concatenation
- subtraction
* multiplication
/ division
** exponentiation
% modulus
logical1 ~ unary NOT
| boolean OR
^ boolean XOR
& boolean AND
relational == equality
!= non-equality
> greater-than
>= greater-than or equal
< less-than
<= less-than or equal
conditional ? : Ternary conditional operator

1 Logical and relational operators are used only with the IF() function and the ternary operator for its conditional argument.

162
Special note on ternary operator Note that the ternary operator is available for use in Xyce. This
operator is the same as the ternary conditional operator in C, C++, Perl, and others. The ternary expression
t?a : b is equivalent to the function IF(t,a,b) described below. However, please be aware that the
ternary operator has extremely low precedence just as it has in these other languages, and if parentheses are
not used to make explicit which expressions are supposed to be part of the condition or true and false
values, the resolution of the expression may be surprising.
For example, the expression

1+a==b?1:0+1

is equivalent to the expression

IF(1+a==b,1,0+1)

because the “+” and “==” operators have higher precedence than either “?” or “:”. Similarly:

A==B?1:0 + A==C?2:0 + A==D?3:0

is equivalent to

if(A==B,1,IF(0 + A==C,2,IF(0 + A==D,3,0)))

Given the way the original expression is written, it appears that the intent was that the expression be
evaluated as:

If(A==B,1,0) + IF(A==C,2,0) + IF(A==D,3,0)

This is not how the expression will be evaluated. Fortunately, because of the use of “0” to the right of the
colons in each case, the expression just happens to give the desired result in either interpretation, but Xyce
is using the nested IF equivalent.
Finally, due to restrictions on the expression parser, it is essential that ternary operators never be written
so that a bare parameter is directly to the left of a colon. This is because colons are actually legal
characters in parameters — the colon represents hierarchy, so that R1:R means the R parameter of device
R1, and X1:A refers to the node A of subcircuit X1. Therefore, it is necessary to put at least one character
that is invalid in parameter names in between the colon and the parameter. It is sufficient to use a space.

{(A==B)?C:D} ; this expression will generate a syntax error

{(A==B)?C :D} ; this expression is acceptable
{(A==B)?C+0:D} ; this expression is acceptable
{(A==B)?(C):D} ; this expression is acceptable

Note that if using expressions without curly braces or single quotes (something that is allowed on .param
and .global_param lines) this becomes even more restrictive because the parser eats up whitespace. In
this case, the expression needs to be fixed another way, such as with parenthesis.

(A==B)?C:D ; this expression will generate a syntax error

(A==B)?C :D ; parser eliminates whitespace, causing syntax error
(A==B)?C+0:D ; this expression is acceptable
(A==B)?(C):D ; this expression is acceptable

163
Arithmetic Functions

Table 2-33. Arithmetic Functions

Function Meaning Explanation
Arithmetic functions
ABS(x) |𝑥| absolute value of 𝑥
least integer greater or equal to variable
CEIL(x) ⌈𝑥⌉
𝑥
𝑑
DDT(x) 𝑑𝑡 𝑥(𝑡) time derivative of 𝑥
𝜕 partial derivative of 𝑓 (𝑥) with respect
DDX(f(x),x) 𝑓 (𝑥)
𝜕𝑥 to 𝑥
greatest integer less than or equal to
FLOOR(x) ⌊𝑥⌋
variable 𝑥
returns the remainder of x/y as a real
FMOD(x,y)
number
𝑡 is an expression using the relational
IF(t,x,y) 𝑥 if 𝑡 is true,
operators in Table 2-32.2
𝑦 otherwise
INT(x) sgn(𝑥) ⌊|𝑥|⌋ integer part of the real variable 𝑥
LIMIT(x,y,z) 𝑦 if 𝑥 < 𝑦
𝑥 if 𝑦 < 𝑥 < 𝑧 𝑥 limited to range 𝑦 to 𝑧
Note: The two-argument version of
𝑧 if 𝑥 > 𝑧 limit, documented below, is a random
operator.
M(x) |𝑥| absolute value or magnitude of 𝑥
MIN(x,y) min(𝑥, 𝑦) minimum of 𝑥 and 𝑦
MAX(x,y) max(𝑥, 𝑦) maximum of 𝑥 and 𝑦
rounds 𝑥 up or down, to the nearest
NINT(x)
integer
PWR(x,y) 𝑥𝑦 𝑥 raised to 𝑦 power
POW(x,y) 𝑥𝑦 𝑥 raised to 𝑦 power
PWRS(x,y) 𝑥𝑦 if 𝑥 > 0
0 if 𝑥 = 0 sign corrected 𝑥 raised to 𝑦 power
−(−𝑥) 𝑦 if 𝑥 < 0
∫
SDT(x) 𝑥(𝑡)𝑑𝑡 time integral of 𝑥
SGN(x) +1 if 𝑥 > 0
0 if 𝑥 = 0 sign value of 𝑥
-1 if 𝑥 < 0
SIGN(x,y) sgn(𝑦)|𝑥| sign of 𝑦 times absolute value of 𝑥
STP(x) 1 if 𝑥 > 0 step function
0 otherwise

164
 Table 2-33. Arithmetic Functions
Function Meaning Explanation
√
SQRT(x) 𝑥 square root of 𝑥
URAMP(x) 𝑥 if 𝑥 > 0 ramp function
0 otherwise
Operators related to interpolating tabular data
𝑓 (𝑥) where piecewise linear interpolation, multiple
TABLE(x,y,z,*)
𝑓 (𝑦) = 𝑧 (𝑦,𝑧) pairs can be specified
Alternate specification for TABLE, in
which data is read in from a file.
filename consists of 𝑥, 𝑦 data pairs,
TABLE(“filename”, [𝑁], [𝑙𝑜𝑔]) one pair per line, space separated.
Synonymous with TABLEFILE.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
piecewise linear interpolation without
𝑓 (𝑥) where
FASTTABLE(x,y,z,*) breakpoints, multiple (𝑦,𝑧) pairs can be
𝑓 (𝑦) = 𝑧
specified
Alternate specification for FASTTABLE,
in which data is read in from a file.
FASTTABLE(“filename”, [𝑁], filename consists of 𝑥, 𝑦 data pairs,
[log]) one pair per line, space separated.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
Akima spline interpolation, multiple
SPLINE(x,y,z,*)
(𝑦,𝑧) pairs should be specified
Alternate specification for SPLINE, in
which data is read in from a file.
filename consists of 𝑥, 𝑦 data pairs,
SPLINE(“filename”, 𝑁)
one pair per line, space separated.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
Akima spline interpolation [8],
AKIMA(x,y,z,*) multiple (𝑦,𝑧) pairs should be specified
Synonymous with SPLINE.
Alternate specification for AKIMA, in
which data is read in from a file.
filename consists of 𝑥, 𝑦 data pairs,
AKIMA(“filename”, [𝑁], [log])
one pair per line, space separated.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
Cubic spline interpolation, multiple
CUBIC(x,y,z,*)
(𝑦,𝑧) pairs should be specified

165
 Table 2-33. Arithmetic Functions
Function Meaning Explanation
Alternate specification for CUBIC, in
which data is read in from a file.
filename consists of 𝑥, 𝑦 data pairs,
CUBIC(“filename”, [𝑁], [log])
one pair per line, space separated.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
Wodicka spline interpolation [9],
WODICKA(x,y,z,*)
multiple (𝑦,𝑧) should be specified
Alternate specification for WODICKA, in
which data is read in from a file.
filename consists of 𝑥, 𝑦 data pairs,
WODICKA(“filename”, [𝑁], [log])
one pair per line, space separated.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
Barycentric Lagrange
BLI(x,y,z,*) interpolation [10], multiple (𝑦,𝑧) pairs
should be specified
Alternate specification for BLI, in
which data is read in from a file.
filename consists of 𝑥, 𝑦 data pairs,
BLI(“filename”, [𝑁], [log])
one pair per line, space separated.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
Alternate specification for TABLE, in
which data is read in from a file.
TABLEFILE(“filename”, [𝑁], filename consists of 𝑥, 𝑦 data pairs,
[log]) one pair per line, space separated.
Arguments 𝑁 and 𝑙𝑜𝑔 are optional, and
related to sparsification.3
Operators related to complex numbers
DB (x) output the magnitude of 𝑥 in decibels
IMG(x) imaginary part of variable 𝑥
PH(x) phase of variable 𝑥
R(x) real part of variable 𝑥
RE(x) real part of variable 𝑥
M(x) |𝑥| absolute value or magnitude of 𝑥
Exponential, logarithmic, and trigonometric functions
ACOS(x) arccos(𝑥) result in radians
−1
ACOSH(x) cosh (𝑥) hyperbolic arccosine of 𝑥
ARCTAN(x) arctan(𝑥) result in radians
ASIN(x) arcsin(𝑥) result in radians
−1
ASINH(x) sinh (𝑥) hyperbolic arcsine of 𝑥

166
 Table 2-33. Arithmetic Functions
Function Meaning Explanation
ATAN(x) arctan(𝑥) result in radians
−1
ATANH(x) tanh (𝑥) hyperbolic arctangent of 𝑥
ATAN2(x,y) arctan(𝑥/𝑦) result in radians
COS(x) cos(𝑥) 𝑥 in radians
COSH(x) cosh(𝑥) hyperbolic cosine of 𝑥
EXP(x) 𝑒𝑥 𝑒 to the 𝑥 power
LN(x) ln(𝑥) log base 𝑒
LOG(x) log(𝑥) log base 10
LOG10(x) log(𝑥) log base 10
SIN(x) sin(𝑥) 𝑥 in radians
SINH(x) sinh(𝑥) hyperbolic sine of 𝑥
TAN(x) tan(𝑥) 𝑥 in radians
TANH(x) tanh(𝑥) hyperbolic tangent of 𝑥
Operators related to random distributions
Random number sampled from normal
AGAUSS(𝜇,𝛼,𝑛) distribution with mean 𝜇 and standard
deviation 𝛼/𝑛
A deviation 𝛼 will be 𝑛 standard
deviations from the mean. The
argument 𝑛 is optional, and defaults to
1.1
Random number sampled from normal
GAUSS(𝜇,𝛼,𝑛) distribution with mean 𝜇 and standard
deviation (𝛼 ∗ 𝜇)/𝑛
A deviation 𝛼 ∗ 𝜇 will be 𝑛 standard
deviations from the mean. The
argument 𝑛 is optional, and defaults to
1.1
Random number sampled from uniform
AUNIF(𝜇,𝛼) distribution with mean 𝜇 and standard
deviation 𝛼/𝑛
The number returned will differ from
the mean by at most 𝛼 1
Random number sampled from uniform
UNIF(𝜇,𝛼) distribution with mean 𝜇 and standard
deviation (𝛼 ∗ 𝜇)/𝑛
The number returned will differ from
the mean by at most 𝛼 ∗ 𝜇 1
where 𝑥 is the nominal value and 𝑦 is
the absolute variation. It will return x+y
LIMIT(x,y)
or x-y, depending on a random number
between -1 and +1.

167
 Table 2-33. Arithmetic Functions
Function Meaning Explanation
Note: The three-argument version of
limit, documented above, is not a
random operator.
random number between 0 and 1
RAND()
sampled from a uniform distribution1

1 The default behavior of the random number functions RAND, GAUSS, and AGAUSS, if there are not any UQ commands such as
.SAMPLING in the netlist, is to return the mean value of the operator. If a UQ command is present, then these operators can
be used to define the distribution of random inputs to the UQ analysis. However, this will only happen if the UQ analysis
specifically requests it using the command .SAMPLING USEEXPR=TRUE Unless a specific random seed is specified using either
the -randseed command line option, or from the netlist, the random number generator will be seeded internally. In all cases,
Xyce will output text to the console indicating what seed is being used.
2 Use of the IF function to create an expression that has step-function-like behavior as a function of a solution variable is highly
likely to produce convergence errors in simulation. IF statements that have step-like behavior with an explicit time dependence
are the exception, as the code will insert breakpoints at the discontinuities. Do not use step-function or other infinite-slope
transitions dependent on variables other than time. Smooth the transition so that it is more easily integrated through. See the
“Analog Behavioral Modeling” chapter of the Xyce Users’ Guide [1] for guidance on using the IF function with the B-source
device.
3 For all the interpolator functions, such as TABLE and SPLINE, the second and third arguments are optional. The second argument,
𝑁, toggles automatic sparsification of the data, where 𝑁 is the requested number of data points. The third argument, 𝑙𝑜𝑔 is a
boolean. If true, the gradient-based sparsification uses a log scale. To specify 𝑙𝑜𝑔, the second parameter 𝑁 must be specified
first.

168
Spice Compatable Functions Expressions can also use functions that are the equivalent of standard
sources.

Table 2-34. SPICE Compatibility Functions

Function Explanation
SPICE_EXP(V1,V2,TD1,TAU1,TD2,TAU2) SPICE style transient exponential
V1 = initial value
V2 = pulsed value
TD1 = rise delay time
TAU1 = rise time constant
TD2 = fall delay time
TAU2 = fall time constant
SPICE_PULSE(V1,V2,TD,TR,TF,PW,PER) SPICE style transient pulse
V1 = initial value
V2 = pulsed value
TD = delay
TR = rise time
TF = fall time
PW = pulse width
PER = period
SPICE_SFFM(V0,VA,FC,MDI,FS) SPICE style transient single frequency FM
V0 = offset
VA = amplitude
FC = carrier frequency
MDI = modulation index
FS = signal frequency
SPICE_SIN(V0,VA,FREQ,TD,THETA) SPICE style transient sine wave
V0 = offset
VA = amplitude
FREQ = frequency (hz)
TD = delay
THETA = damping factor

Information about the restrictions on expressions in specific contexts is given in the subsections that
follow.

2.2.1. Expressions in .PARAM or .GLOBAL_PARAM Statements

Parameters defined using .PARAM and .GLOBAL_PARAM are mostly synonymous and subject to most of the
same constraints. Both types are allowed to depend on parameters defined in .PARAM or .GLOBAL_PARAM

169
statements, and both may contain special variables such as TIME, FREQ, TEMP or VT. Neither type may
contain any references to solution variables or lead currents.

Example: .PARAM SQUARES=5.0

Example: .PARAM SHEETRES=25

Example: .PARAM RESISTANCE={SQUARES*SHEETRES}

Example: .PARAM a0 = 1.0+2.0J

Example: .PARAM dTdt=.01

Example: .GLOBAL_PARAM Temperature={27+dTdt*TIME}

Both types of parameter are subject to extra constraints if used inside of subcircuits. .PARAM parameters
may be defined inside of a subcircuit, while .GLOBAL_PARAM parameters may not. Parameters inside of
subcircuits may not be swept by analysis commands such as .STEP or .SAMPLING. Only globally scoped
user-defined parameters can be swept in this manner. Note, this restriction does not apply to device
parameters, just user-defined parameters.
Parameters can be complex-valued. The suffix letter used for the imaginary part of a complex number is the
letter J. For example, in the parameter statement .param a0=1.0+2.0J, the parameter a0 has 1.0 for the
real part and 2.0 is the imaginary part.
Although parameters are allowed to be complex, many uses of parameters are inherently real-valued. Most
device parameters, for example, are assumed to be double precision numbers. So, if a real-valued device
parameter is set equal to a complex-valued .param, the device parameter will use the real part and the
imaginary part will be ignored.

2.2.2. Expressions in .PRINT Lines

Expressions on .PRINT lines may contain references to parameters defined in either .PARAM or
.GLOBAL_PARAM statements, device parameters using the syntax <device name>:<parameter name>,
and may also contain solution variables.

*example with .print expressions

.PARAM RES=50
R1 1 0 {RES}
V1 1 0 sin(0 5 100khz)
.tran 1u 1m
*Print power dissipated through resistor,
*and actual resistance used in the R1
*device
.print tran {V(1)*V(1)/RES} {R1:R}
.end

170
2.2.3. Using Complex Values in Expressions

The Xyce expression library was written to work with complex quantities, and can perform complex
arithmetic on any set of complex inputs to produce a complex output. However, the default behavior on the
.PRINT line depends on the type of analysis being run.
If running a frequency domain analysis such as .AC or .HB, then outputs of complex valued expressions
will automatically include both the real and imaginary part. For example:
.param r0={log(-1)}
.print ac {r0} // automatically output real and imaginary parts
If running a real-valued analysis such as .DC or .TRAN, the output on the .PRINT line will, by default, only
include the real part. This is true even if the expression evaluates to a complex number. If the user is
running a real-valued analysis, but desires output of both the real and imaginary part, it is necessary to use
operators such as Re() and Img(). Here are some examples.

.param r0={log(-1)}
.print tran {r0} // only output the real part of {r0}
.print tran {re(r0)} {img(r0)} // output both real and imaginary parts

Note any complex-valued expression will internally be evaluaed using complex math. The only thing
different for {r0} in the above examples is the default outputs for .AC and .TRAN.

2.2.4. Expressions for Device Instance and Model Parameters

Expressions of constants, .PARAM and .GLOBAL_PARAM parameters may be used for the values of any
device parameters in instance and model lines.
Except in very specific devices, expressions used for device parameter values must evaluate to a
time-independent constant, and must not contain dependence on solution variables such as nodal voltages
or currents.
*example of use of expressions for device parameters
.PARAM RES=50
.GLOBAL_PARAM theSaturationCurrent=1.5e-14
R1 1 0 {RES}
V1 1 0 sin(0 5 100khz)
D1 1 0 DMODEL
.MODEL D DMODEL IS=theSaturationCurrent

.step theSaturationCurrent 1e-14 5e-14 1e-14

Some parameters of specific devices are exceptions to the general rule. These parameters have no
restrictions and may depend on any parameters, time, or solution variables in the netlist:
• The V or I instance parameters of the B source.
• The CONTROL instance parameter of the switch (S device).
• The C (capacitance) instance parameter for the capacitor.

171
 • The Q (charge) instance parameter for the capacitor.
• The coupling coefficient instance parameter for the LINEAR mutual inductor (K device with no model
card specified)
These specific instance parameters may be time-depdendent (i.e. they may reference the TIME special
variable), but may not depend on any solution variables:

• The TEMP instance parameter of all devices.

• The L (inductance) parameter of the inductor.
• The R (resistance) parameter of the resistor.
• The R, RESISTIVITY, DENSITY, HEATCAPACITY and THERMAL_HEATCAPACITY
parameters of the thermal resistor (resistor level 2).

2.2.5. POLY expressions

The POLY keyword is available in the E, F, G, H and B dependent sources. Based on the same keyword from
SPICE2, POLY provides a compact method of specifying polynomial expressions in which the variables in
the polynomial are specified followed by an ordered list of polynomial coefficients. All expressions
specified with POLY are ultimately translated by Xyce into an equivalent, straightforward polynomial
expression in a B source. Since a straightforward polynomial expression can be easier to read, there is no
real benefit to using POLY except to support netlists imported from other SPICE-based simulators.
There are three different syntax forms for POLY, which can be a source of confusion. The E and G sources
(voltage-dependent voltage or current sources) use one form, the F and H sources (current-dependent
voltage or current sources) use a second form, and the B source (general nonlinear source) a third form.
During input processing, any of the E,F,G or H sources that use nonlinear expressions are first converted
into an equivalent B source, and then any B sources that use the POLY shorthand are further converted into
standard polynomial expressions. This section describes how the compact form will be translated into the
final form that is used internally.
All three formats of POLY express the same three components: a number of variables involved in the
expression (𝑁, the number in parentheses after the POLY keyword), the variables themselves, and an ordered
list of coefficients for the polynomial terms. Where they differ is in how the variables are expressed.

2.2.5.1. Voltage-controlled sources

The E and G sources are both voltage-controlled, and so their POLY format requires specification of two
nodes for each voltage on which the source depends, i.e. the positive and negative nodes from which a
voltage drop is computed. There must therefore be twice as many nodes as the number of variables
specified in parentheses after the POLY keyword:
Epoly 1 2 POLY(3) n1p n1m n2p n2m n3p n3m ...
In this example, the voltage between nodes 1 and 2 is determined by a polynomial whose variables are
V(n1p,n1m), V(n2p,n2m), V(n3p,n3m). Not shown in this example are the polynomial coefficients,
which will be described later.

172
2.2.5.2. Current-controlled sources

The F and H sources are both current-controlled, and so their POLY format requires specification of one
voltage source name for each current on which the source depends. There must therefore be exactly as many
nodes as the number of variables specified in parentheses after the POLY keyword:
Fpoly 1 2 POLY(3) V1 V2 V3 ...
In this example, the voltage between nodes 1 and 2 is determined by a polynomial whose variables are
I(V1), I(V2), and I(V3). Not shown in this example are the polynomial coefficients, which will be
described later.

2.2.5.3. B sources

Finally, the most general form of POLY is that used in the general nonlinear dependent source, the B source.
In this variant, each specific variable must be named explicitly (i.e. not simply by node name or by voltage
source name), because currents and voltages may be mixed as needed.
Bpoly 1 2 V={POLY(3) I(V1) V(2,3) V(3) ...}
Bpoly2 1 2 I={POLY(3) I(V1) V(2,3) V(3) ...}
In these examples, the source between nodes 1 and 2 is determined by a polynomial whose variables are
I(V1), V(2,3), and V(3). In the first example, the polynomial value determines the voltage between nodes
1 and 2, and in the second the current.
The E, F, G and H formats are all converted internally in a first step to the B format. Thus the following
pairs of sources are exactly equivalent:
Epoly 1 2 POLY(3) n1p n1m n2p n2m n3p n3m ...
BEpoly 1 2 V={POLY(3) V(n1p,n1m) V(n2p,n2m) V(n3p,n3m) ...
Fpoly 1 2 POLY(3) V1 V2 V3 ...
BFpoly 1 2 V={POLY(3) I(V1) I(V2) I(V3) ...
After conversion to the B source form, the POLY form is finally converted to a normal expression using the
coefficients and variables given.
Coefficients are given in a standard order, and the polynomial is built up by terms until the list of
coefficients is exhausted. The first coefficient is the constant term of the polynomial, followed by the
coefficients of linear terms, then bi-linear, and so on. For example:
Epoly 1 2 POLY(3) n1p n1m n2p n2m n3p n3m 1 .5 .5 .5
In this example, the constant term is 1.0, and the coefficients of the three terms linear in the input variables
are 0.5. Thus, this E source is precisely equivalent to the general B source:
BEstandard 1 2 V={1.0 + .5*V(n1p,n1m) + .5*V(n2p,n2m) +.5*V(n3p,n3m)}
The standard ordering for coefficients is:
POLY(N) 𝑋1 . . . 𝑋 𝑁 𝐶0 𝐶1 . . . 𝐶 𝑁 𝐶11 . . . 𝐶1𝑁 𝐶21 . . . 𝐶 𝑁 1 . . . 𝐶 𝑁 𝑁 𝐶12 1 . . . 𝐶12 𝑁 . . .
with the polynomial then being:

173
 𝑁
∑︁ 𝑁 ∑︁
∑︁ 𝑁 𝑁 ∑︁
∑︁ 𝑁
𝑉 𝑎𝑙𝑢𝑒 = 𝐶0 + 𝐶𝑗 𝑋𝑗 + 𝐶𝑖 𝑗 𝑋𝑖 𝑋 𝑗 + 𝐶𝑖2 𝑗 𝑋𝑖2 𝑋 𝑗 + . . .
𝑗=1 𝑖=1 𝑗=1 𝑖=1 𝑗=1

Here we have used the general form 𝑋𝑖 for the 𝑖 𝑡 ℎ variable, which may be either a current or voltage
variable in the general case.
It should be reiterated that the POLY format is provided primarily for support of netlists from other
simulators, and that its compactness may be a disadvantage in readability of the netlist and may be more
prone to usage error. Xyce users are therefore advised that use of the more straightforward expression format
in the B source may be more appropriate when crafting original netlists for use in Xyce. Since Xyce converts
POLY format expressions to the simpler format internally, there is no performance benefit to use of POLY.

174
2.3. Devices

Xyce supports many devices, with an emphasis on analog devices, including sources, subcircuits and
behavioral models. This section serves as a reference for the devices supported by Xyce. Each device is
described separately and includes the following information, if applicable:
• a description and an example of the correct netlist syntax.
• the matching model types and their description.
• the matching list of model parameters and associated descriptions.
• the corresponding characteristic equations for the model (as required).
• references to publications on which the model is based.
User-defined models may be implemented using the .MODEL (model definition) statement, and
macromodels can be created as subcircuits using the .SUBCKT (subcircuit) statement.
Please note that the characteristic equations are provided to give a general representation of the device
behavior. The actual Xyce implementation of the device may be slightly different in order to improve, for
example, the robustness of the device.
Table 2-35 gives a summary of the device types and the form of their netlist formats. Each of these is
described below in detail.

Table 2-35. Analog Device Quick Reference.

Device Type Letter Typical Netlist Format
Nonlinear Dependent B<name> <+ node> <- node>
B
Source (B Source) + <I or V>={<expression>}
C<name> <+ node> <- node> [model name] <value>
Capacitor C
+ [IC=<initial value>]
D<name> <anode node> <cathode node>
Diode D
+ <model name> [area value]
Voltage Controlled E<name> <+ node> <- node> <+ controlling node>
E
Voltage Source + <- controlling node> <gain>
Current Controlled F<name> <+ node> <- node>
F
Current Source + <controlling V device name> <gain>
Voltage Controlled G<name> <+ node> <- node> <+ controlling node>
G
Current Source + <- controlling node> <transconductance>
Current Controlled H<name> <+ node> <- node>
H
Voltage Source + <controlling V device name> <gain>
I<name> <+ node> <- node> [[DC] <value>]
Independent Current
I + [AC [magnitude value [phase value] ] ]
Source
+ [transient specification]
K<name> <inductor 1> [<ind. n>*]
Mutual Inductor K
+ <linear coupling or model>
L<name> <+ node> <- node> [model name] <value>
Inductor L
+ [IC=<initial value>]

175
 Table 2-35. Analog Device Quick Reference.
Device Type Letter Typical Netlist Format
J<name> <drain node> <gate node> <source node>
JFET J
+ <model name> [area value]
M<name> <drain node> <gate node> <source node>
MOSFET M + <bulk/substrate node> [SOI node(s)]
+ <model name> [common model parameter]*
O<name> <A port (+) node> <A port (-) node>
Lossy Transmission Line
O + <B port (+) node> <B port (-) node>
(LTRA)
+ <model name>
Q<name> <collector node> <base node>
Bipolar Junction
Q + <emitter node> [substrate node]
Transistor (BJT)
+ <model name> [area value]
R<name> <+ node> <- node> [model name] <value>
Resistor R
+ [L=<length>] [W=<width>]
S<name> <+ switch node> <- switch node>
Voltage Controlled Switch S + <+ controlling node> <- controlling node>
+ <model name>
S<name> <+ switch node> <- switch node>
Generic Switch S
+ <model name> CONTROL=\{expression\}
T<name> <A port + node> <A port - node>
Transmission Line T + <B port + node> <B port - node>
+ <ideal specification>
U<name> <type> <digital power node>
Digital Devices U
+ <digital ground node> [node]* <model name>
V<name> <+ node> <- node> [[DC] <value>]
Independent Voltage
V + [AC [magnitude value [phase value] ] ]
Source
+ [transient specification]
P<name> <+ node> <- node> [[DC] <value>]
+ port=port number [Z0 = value]
Port Device P
+ [AC [magnitude value [phase value] ] ]
+ [transient specification]
X<name> [node]* <subcircuit name>
Subcircuit X
+ [PARAMS:[<name>=<value>]*]
W<name> <+ switch node> <- switch node>
Current Controlled Switch W
+ <controlling V device name> <model name>
Digital Devices, Y Type
Y<type> Y<type> <name> [node]* <model name>
(deprecated)
PDE Devices YPDE YPDE <name> [node]* <model name>
YACC <name> <acceleration> <velocity> <position>
Accelerated masses YACC
+ [x0=<initial position>] [v0=<initial velocity>]
Linear Device YLIN YLIN <name> <+ node> <- node> <model name>
Memristor Device YMEMRISTOR YMEMRISTOR <name> <+ node> <- node> <model name>
Z<name> <drain node> <gate node> <source node>
MESFET Z
+ <model name> [area value]

176
 Table 2-36. Features Supported by Xyce Device Models

Device Comments Analytic

Branch
Power Sensi- Stationary
Current
tivity Noise

Capacitor Age-aware, semiconductor

Y Y Y

Inductor Coupled mutual inductors (see below)

Y Y Y
Linear and Nonlinear Mutual
Inductor Y Y

Resistor (Level 1) Normal and Semiconductor

Y Y Y Y

Resistor (Level 2) Thermal Resistor

Y Y

Diode (Level 1)
Y Y Y Y

Diode (Level 2) Addition of PSpice enhancements

Y Y

Diode (Level 200) JUNCAP200 model

Y Y Y Y
Independent Voltage Source
(VSRC) Y Y Y
Independent Current Source
(ISRC) Y Y Y
Voltage Controlled Voltage
Source (VCVS) Y Y
Voltage Controlled Current
Source (VCCS) Y Y
Current Controlled Voltage
Source (CCVS) Y Y
Current Controlled Current
Source (CCCS) Y Y

Port Device
Y Y
Nonlinear Dependent Source
(B Source) Y Y

177
 Table 2-36. Features Supported by Xyce Device Models

Device Comments Analytic

Branch
Power Sensi- Stationary
Current
tivity Noise
Bipolar Junction Transistor
(BJT) (Level 1) Y Y Y Y
Bipolar Junction Transistor Vertical Bipolar Intercompany (VBIC)
(BJT) (Level 11) model, version 1.3 (3-terminal) Y Y Y Y
Bipolar Junction Transistor Vertical Bipolar Intercompany (VBIC)
(BJT) (Level 12) model, version 1.3 (4-terminal) Y Y Y Y
FBH (Ferdinand-Braun-Institut für
Bipolar Junction Transistor
Höchstfrequenztechnik) HBT model,
(BJT) (Level 23) Y Y Y N
version 2.1
Bipolar Junction Transistor
HICUM Level 0
(BJT) (Level 230) Y Y Y Y
Bipolar Junction Transistor
HICUM Level 2
(BJT) (Level 234) Y Y Y Y
Bipolar Junction Transistor
MEXTRAM version 504.12.1
(BJT) (Level 504) Y Y Y Y
Bipolar Junction Transistor MEXTRAM version 504.12.1 (with
(BJT) (Level 505) self-heating) Y Y Y Y
Junction Field Effect Transistor
SPICE-compatible JFET model
(JFET) (Level 1) Y Y
Junction Field Effect Transistor
Shockley JFET model
(JFET) (Level 2) Y Y

MESFET
Y Y

MOSFET (Level 1)
Y Y Y

MOSFET (Level 2) SPICE level 2 MOSFET

Y Y Y

MOSFET (Level 3)
Y Y Y

MOSFET (Level 6) SPICE level 6 MOSFET

Y Y Y

MOSFET (Level 9) BSIM3 model

Y Y Y Y

178
 Table 2-36. Features Supported by Xyce Device Models

Device Comments Analytic

Branch
Power Sensi- Stationary
Current
tivity Noise

MOSFET (Level 10) BSIM SOI model

Y Y

MOSFET (Level 14 or 54) BSIM4 model

Y Y

MOSFET (Level 18) VDMOS general model

Y Y

MOSFET (Level 77) BSIM6 model version 6.1.1

Y Y Y Y

MOSFET (Level 102) Legacy PSP model

Y Y Y Y

MOSFET (Level 103) PSP model

Y Y Y Y

MOSFET (Level 107) BSIM-CMG version 107.0.0

Y Y Y Y

MOSFET (Level 110) BSIM-CMG version 110.0.0

Y Y Y Y

MOSFET (Level 301) EKV model version 3.0.1

Y Y Y Y

MOSFET (Level 2000) MVS ETSOI model version 2.0.0

Y Y Y Y

MOSFET (Level 2001) MVS HEMT model version 2.0.0

Y Y Y Y

Transmission Line (TRA) Lossless

Y Y

Transmission Line (LTRA) Lossy

Lumped Transmission Line Lossy or Lossless

Controlled Switch (S,W)

Voltage or current controlled
(VSWITCH/ISWITCH) Y Y

Generic Switch (SW) Controlled by an expression

Y Y

179
 Table 2-36. Features Supported by Xyce Device Models

Device Comments Analytic

Branch
Power Sensi- Stationary
Current
tivity Noise

PDE Devices (Level 1) one-dimensional

Y

PDE Devices (Level 2) two-dimensional

Y

Digital (Level 1) Behavioral Digital

NA NA
Accelerated mass device, used for
ACC simulation of electromechanical and
magnetically-driven machines NA NA
Separate models for Branch, Bus
Shunt, Transformer and Generator
Power Grid
Bus. The Generator Bus model
supports reactive power (Q) limiting

Memristor TEAM formulation

Y Y

Memristor Yakopcic
Y Y

Memristor PEM Formulation

Y Y

180
2.3.1. Voltage Nodes

Devices in a netlist are connected between nodes, and all device types in Xyce require at least two nodes on
each instance line. Section 2.3.2 lists the characters that are legal and illegal in Xyce node and device
names.
A node is simply a named point in the circuit. The naming of nodes is mainly known within the level of
circuit hierarchy where they appear. Nodes can be passed into subcircuits through an argument list, and in
this manner subcircuits are given access to nodes from the upper-level circuit. Historicaly, this is how nodes
are passed thru the circuit hierarchy in most circuit simulators, and this is the convention used by most
circuit netlists. However, Xyce has two exceptions to this convention. Global nodes, described in
section 2.3.1.1 and fully resolved internal subcircuit nodes, described in section 2.3.1.2.

2.3.1.1. Global nodes

A special syntax is used to designate certain nodes as global nodes. Any node whose name starts with the
two characters “$G” is a global node, and such nodes are available to be used in any subcircuit. A typical
usage of such global nodes is to define a VDD or VSS signal that all subcircuits need to be able to access, but
without having to provide VSS and VDD input nodes to every subcircuit. In this case, a global $GVDD node
would be use for the VDD signal.
The node named 0 is a special global node. Node 0 is always ground, and is accessible to all levels of a
hierarchical netlist.
For compatibility with HSPICE, the .GLOBAL command can be used to define global nodes that do not start
with the two characters “$G”. See section 2.1.11 for more details.

2.3.1.2. Subcircuit Nodes

Hierarchical netlists may be created using .SUBCKT [2.1.37] to define common subcircuit types, and X
[2.3.33] lines to create instances of those subcircuits. There are two types of nodes associated with such
subcircuits, interface nodes and internal nodes.
Interface nodes are the nodes named on the .SUBCKT line. These are effectively local aliases internal to the
subcircuit definition for the node names used on the X instance lines. Internal nodes are nodes inside the
subcircuit definition that are strictly local to that subcircuit. Inside a subcircuit, these node names may be
used without restriction in device instance lines and expressions on B source lines.
There are some circumstances when it is desirable to access internal nodes of a subcircuit from outside that
subcircuit. Xyce provides a syntax that allows this to be done. The primary context in which this is
supported is on .PRINT lines, to allow the user to print out signals that are usually local to a subcircuit.
However, this syntax isn’t limited to .PRINT lines, and can work in other contexts.
The syntax used by Xyce to refer to nodes within a subcircuit is to prefix the name of the node with the full
path of subcircuit instances in which the node is contained, with colons (:) separating the instance names.
So, to reference a node “A” that is inside a subcircuit instance called “Xnot1” inside another subcircuit
instance called “Xmain”, one would refer to “Xmain:Xnot1:A”.
Note that while the default separator in Xyce is the colon (:), the period (.) is also optionally supported. For
more information about using a period separator, see section 3.

181
The same syntax works on .PRINT lines even if the subcircuit node is one of the interface nodes on the
.SUBCKT line, but those nodes can also be accessed by using the names of the nodes at the higher level of
circuit hierarchy that are used on its instance line.

* Netlist demonstrating subcircuit node .PRINT access

V1 1 0 1
X1 1 2 demosubc
X2 2 0 demosubc
.subckt demosubc A B
R1 A C 1
R2 C B 1
.ends

.dc V1 1 5 1

*V(X1:A) and V(1) are the same signal.

*V(X1:C) is the internal C node of the X1 instance
*V(X2:C) is the internal C node of the X2 instance
*V(X1:B), V(X2:A) and V(2) are the same signal
.print DC V(X1:C) V(X2:C) V(X1:A) V(1)
+ V(X1:B) V(X2:A) V(2)
.end

Internal subcircuit nodes may also be accessed from outside of the subcircuit if one uses the fully resolved
syntax. This works in B source voltage or current expressions, and also in standard netlist usage on device
instance lines. This type of usage is outside of the typically strict hierarchy required by most circuit
simulators, but it can be useful in some contexts.
The one difference between this usage and .PRINT usage is that it is not possible to use the subcircuit node
syntax to access interface nodes. These must be accessed using the node names being used on the instance
line, as in the “V(1)” example in the netlist fragment above. Two valid examples of internal subcircuit node
access are given by the fragment below.
* Netlist demonstrating resolved subcircuit nodes
Vin 1 0 1.0
X1 1 2 test
Rout 2 0 1.0
.subckt test A B
Rt1 A testNode 1.0
Rt2 testNode B 1.0
.ends

* this works:
Btest1 3 0 V = V(X1:testNode)
Rtest1 3 0 1.0
* this also works:
Itest2 0 4 1.0
Rtest2 X1:testNode 4 1.0
Rtest3 X1:testNode 0 1.0

182
2.3.2. Legal Characters in Node and Device Names

Xyce node names and device names can consist of any printable ASCII characters, with the following
exceptions and caveats which may be different than other SPICE-like circuit simulators. The exceptions
are:

• White space (space, tab, newline) is not allowed.

• Parentheses (“(” or “)”), braces (“{” or “}”), commas, colons, semi-colons, double quotes and single
quotes are also not allowed, since they do not work correctly in node names or device names in all
netlist contexts in Xyce.

The caveats are as follows:

• The star (*) and question mark (?) characters are allowed in both node names and device names.
However, those two characters also function as “print wildcards” in Xyce, per section 2.1.31.11. So,
that usage is discouraged.

• Global nodes in Xyce begin with the two characters “$G”.

• The node named 0 (“zero”) is a special global node, which is always the ground node.

• These arithmetic operators % ^ & ? : ~ * - + < > / | should not be used in node or device names that
will be used outside of a Xyce “operator”, such as V(), within a Xyce expresson. Examples of this
caveat are given below.

• The # character should not be used as the first character of a node name that will be used within an
expression. Examples of this caveat are also given below.

These are some examples of the caveats of the use of arithmetic operators and # character within
expressions:

* Okay since the + in the node name is enclosed within the V() operator.
.PRINT TRAN {V(1+) - V(+)}
* Okay since the R+ and R- device names are enclosed within the I() operator.
.PRINT TRAN {I(R+) * I(R-)}
* Okay, for printing the resistance value, since the R-1 device name
* is not used in an expression.
.PRINT TRAN R-1:R
* Will produce a parsing error, since the R-1 device name is used outside
* of an operator. That makes this statement ambiguous within an expression.
.PRINT TRAN {R-1:R}
* These uses of # are okay.
.PRINT TRAN V(#) {V(1#) -1}
* These usages of # are parsing errors, since # is the first character
* in the node names.
.PRINT TRAN {V(#) - 2} {V(#1) -1}

183
2.3.3. Lead Currents and Power Calculations

For some devices, such as independent voltage and current sources, the current through that device is a
“solution” variable. For other devices, the current through the device is a “lead current”, whose value is
calculated during a post-processing step. This approach has ramifications in Xyce for the availability and
accuracy of lead current values. In particular, both lead currents and power calculations need to have been
explicitly enabled for a given device, analysis type (e.g., .AC) or netlist command (e.g., .MEASURE).
For voltage sources, both V and I are solution variables. So, their accuracy is more likely to be limited by the
nonlinear solver tolerances (RELTOL and ABSTOL). The lead current accuracy, for a device like the resistor,
can also be limited by the right-hand side tolerance RHSTOL. So, the calculated lead currents through very
small resistances (e.g., 1e-12) may be inaccurate if the default solver tolerances for Xyce are used.
Lead currents have the following additional limitations:
• They are not enabled for .AC analyses.
• They are not allowed in the expression controlling a B-Source.
• They do not work for .RESULT statements.
Lead currents and power calculations are available in .MEASURE and .FOUR statements.
At this time the power calculations are only supported for .DC and .TRAN analysis types and for a limited
set of devices. In addition, the results for semiconductor devices (D, J, M, Q and Z devices) and the lossless
transmission device (T device) may differ from other simulators. Consult the Features Supported by Xyce
Device Models table in section 2.3 and the individual sections on each device for more details.
As an example, the power supplied or dissipated by the voltage source V is calculated as 𝐼 · Δ𝑉 where the
voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− . Dissipated power has a
positive sign, while supplied power has a negative sign.
An important note is that the power calculations are also a post-processing step, which places a limit on the
accuracy of circuit-wide “energy conservation” calculations (e.g., total power supplied by sources - total
power dissipated in non-source devices) in Xyce. The accuracy of the inputs (V and I) to the power
calculations is limited by the nonlinear solver and right-hand side tolerances, as noted above, and the error
in the power calculations is upper-bounded by the sum of the product-terms of V*(error in I) and
I*(error in V).

184
2.3.4. Capacitor

Symbol

Instance Form C<device name> <(+) node> <(-) node> [model name] [value]
+ [device parameters]

Model Form .MODEL <model name> C [model parameters]

.MODEL <model name> CAP [model parameters]

Examples CM12 2 4 5.288e-13

CLOAD 1 0 4.540pF IC=1.5V
CFEEDBACK 2 0 CMOD 1.0pF
CAGED 2 3 4.0uF D=0.0233 AGE=86200
CSOLDEP 3 0 C={ca*(c0+c1*tanh((V(3,0)-v0)/v1))}
CSOLDEPQ 3 0 Q={ca*(c1*v1*ln(cosh((v(3,0)-v0)/v1))+c0*v(3,0))}

Parameters and
Options
device name
The name of the device.
(+) node
(-) node
Polarity definition for a positive voltage across the capacitor. The first node is
defined as positive. Therefore, the voltage across the component is the first
node voltage minus the second node voltage.
model name
If model name is omitted, then value is the capacitance in farads. If [model
name] is given then the value is determined from the model parameters; see
the capacitor value formula below.
value
Positional specification of device parameter C (capacitance). Alternately,
this can be specified as a parameter, C=<value>, or in the (optional) model.
device parameters
Parameters listed in Table 2-37 may be provided as space separated
<parameter>=<value> specifications as needed. Any number of
parameters may be specified.
model parameters
Parameters listed in Table 2-38 may be provided as space separated
<parameter>=<value> specifications as needed. Any number of
parameters may be specified.

185
Comments Positive current flows through the capacitor from the (+) node to the (-) node. In
general, capacitors should have a positive capacitance value (<value> property). In
all cases, the capacitance must not be zero.

However, cases exist when a negative capacitance value may be used. This occurs
most often in filter designs that analyze an RLC circuit equivalent to a real circuit.
When transforming from the real to the RLC equivalent, the result may contain a
negative capacitance value.

In a transient run, negative capacitance values may cause the simulation to fail due to
instabilities they cause in the time integration algorithms.

The power stored or released from the capacitor is calculated with 𝐼 · Δ𝑉 where the
voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .

For compatibility with PSpice, either C or CAP can be used in a .MODEL statement for a
capacitor.

The Multiplicity Factor (M) can be used to specify multiple, identical capacitors in
parallel. The effective capacitance becomes C*M. The M value need not be an integer.
It can be any positive real number. M can not be used as a model parameter.

Device Parameters

Table 2-37. Capacitor Device Instance Parameters

Parameter Description Units Default
AGE Age of capacitor hour 0
C Capacitance F 1e-06
D Age degradation coefficient – 0.0233
IC Initial voltage drop across device V 0
L Semiconductor capacitor width m 1
M Multiplicity Factor – 1
Q Charge C 0
TC1 Linear Temperature Coefficient ◦ C −1 0
TC2 Quadratic Temperature Coefficient ◦ C −2 0
◦C Ambient
TEMP Device temperature
Temperature
W Semiconductor capacitor length m 1e-06

In addition to the parameters shown in the table, the capacitor supports a vector parameter for the
temperature correction coefficients. TC1=<linear coefficient> and TC2=<quadratic
coefficient> may therefore be specified compactly as TC=<linear coefficient>,<quadratic
coefficient>.

186
Model Parameters

Table 2-38. Capacitor Device Model Parameters

Parameter Description Units Default
C Capacitance multiplier – 1
CJ Junction bottom capacitance F/m2 0
CJSW Junction sidewall capacitance F/m 0
DEFW Default device width m 1e-06
NARROW Narrowing due to side etching m 0
TC1 Linear temperature coefficient ◦ C −1 0
TC2 Quadratic temperature coefficient ◦ C −2 0
◦C Ambient
TNOM Nominal device temperature
Temperature

Capacitor Equations

Capacitance Value Formula If [model name] is specified, then the capacitance is given by:

C · (1 + TC1 · (𝑇 − 𝑇0 ) + TC2 · (𝑇 − 𝑇0 ) 2 )
where C is the base capacitance specified on the device line and is normally positive (though it can be
negative, but not zero). 𝑇0 is the nominal temperature (set using TNOM option).

Age-aware Formula If AGE is given, then the capacitance is:

C[1 − D log(AGE)]

Semiconductor Formula If [model name] and L and W are given, then the capacitance is:

CJ(L − NARROW) (W − NARROW) + 2 · CJSW(L − W + 2 · NARROW)

Solution-Dependent Capacitor If the capacitance (C) is set equal to an expression then a

“solution-dependent” capacitor is used, where the capacitance is a function of other simulation variables.
The formulas for temperature-dependence and age-dependence, given above, then use that calculated C
value.

If the parameter Q is set equal to an expression instead of specifying a capacitance, this expression is used
to evaluate the charge on the capacitor instead of computing it from capacitance. Temperature and age
dependence are not computed in this case, as these effects are applied by modifying the capacitance.

Both solution-dependent charge and capacitance formulations are implemented to assure charge
conservation. The capacitor:

187
 c_mcap 1 2 q={ca*(c1*v1*ln(cosh((v(1,2)-v0)/v1))+c0*v(1,2))}

is exactly equivalent to the capacitor

c_mcap 1 2 c={ca*(c0+c1*tanh((V(1,2)-v0)/v1))}

because the capacitance is the derivative of the charge with respect to the voltage drop across the capacitor.
Similarly, both are equivalent to the behavioral source:

BC 1 2 I=ddt(V(1,2))*(ca*(c0+c1*tanh((V(1,2)-v0)/v1)))

because 𝐼 = 𝑑𝑄/𝑑𝑡 = 𝑑𝑄/𝑑𝑉 ∗ 𝑑𝑉/𝑑𝑡 = 𝐶 ∗ 𝑑𝑉/𝑑𝑡.

The restrictions for this formulation are:
• The expression used for C or Q must only use solution variables, which are node voltages and also
branch currents for source devices. It may not use device lead currents, which are post-processed
quantities that are not solution variables.
• The expression must not use time derivatives.
• Capacitance (C) and Charge (Q) are the only instance or model parameters that are allowed to be
solution-dependent.

Other Restrictions and Caveats A netlist parsing error will occur if:
• Neither the C, Q, nor L instance parameters are specified.
• Both C and Q are specified as expressions.
• Q is specified in addition to an IC=.
• The A instance parameter is specified for a semiconductor capacitor (which is specified via L, W and
CJSW).
If both the C and L instance parameters are specified then C will be used, rather than the semiconductor
formulation.

Special note on Initial Conditions: The IC parameter of the capacitor may be used to specify an initial
voltage drop on the capacitor. Unlike SPICE3F5, this parameter is never ignored (SPICE3F5 only respects
it if UIC is used on a transient line). The initial condition is applied differently depending on the analysis
specified.
If one is doing a transient with DC operating point calculation or a DC operating point analysis, the initial
condition is applied by inserting a voltage source across the capacitor to force the operating point to find a
solution with the capacitor charged to the specific voltage. The resulting operating point will be one that is
consistent with the capacitor having the given voltage in steady state.
If one specifies UIC or NOOP on the .TRAN line, then Xyce does not perform an operating point calculation,
but rather begins a transient simulation directly given an initial state for the solution. In this case, IC initial
conditions are applied only for the first iteration of the Newton solve of the first time step — the capacitor

188
uses the initial condition to compute its charge, and the nonlinear solver will therefore find a solution to the
circuit problem consistent with this charge, i.e., one with the correct voltage drop across the capacitor.
The caveats of this section apply only to initial conditions specified via IC= parameters on the capacitor,
and do not affect how initial conditions are applied when using .IC lines to specify initial conditions on
node values.
The three different ways of specifying initial conditions can lead to different circuit behaviors. Notably,
when applying initial conditions during a DC operating point with IC= on the capacitor line, the resulting
operating point will be a DC solution with currents everywhere consistent with there being a constant
charge on the capacitor, whereas in general a transient run from an initial condition without having
performed an operating point calculation will have a quiescent circuit at the first timestep.

189
2.3.5. Inductor

Symbol

Instance Form L<name> <(+) node> <(-) node> [model] <value> [device parameters]

Model Form .MODEL <model name> L [model parameters]

.MODEL <model name> IND [model parameters]

Examples L1 1 5 3.718e-08
LM 7 8 L=5e-3 M=2
LLOAD 3 6 4.540mH IC=2mA
Lmodded 3 6 indmod 4.540mH
.model indmod L (L=.5 TC1=0.010 TC2=0.0094)

Parameters and
Options (+) node
(-) node
Polarity definition for a positive voltage across the inductor. The first node is
defined as positive. Therefore, the voltage across the component is the first
node voltage minus the second node voltage.
initial value
The initial current through the inductor during the bias point calculation.

Comments In general, inductors should have a positive inductance value. The inductance must not
be zero. Also, a netlist parsing error will occur if no value is specified for the
inductance.
However, cases exist when a negative value may be used. This occurs most often in
filter designs that analyze an RLC circuit equivalent to a real circuit. When
transforming from the real to the RLC equivalent, the result may contain a negative
inductance value.
The power stored or released from the inductor is calculated with 𝐼 · Δ𝑉 where the
voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
If a model name is given, the inductance is modified from the value given on the
instance line by the parameters in the model card. See “Inductance Value Formula”
below.
When an inductor is named in the list of coupled inductors in a mutual inductor device
line (see page 193) , and that mutual inductor is of the nonlinear-core type, the
<value> is interpreted as a number of turns rather than as an inductance in Henries.
For compatibility with PSpice, either L or IND can be used in a .MODEL statement for
an inductor.

190
The Multiplicity Factor (M) can be used to specify multiple, identical inductors in
parallel. The effective inductance becomes L/M. However, the value for the IC instance
parameter is not multiplied by the M value. The M value need not be an integer. It can
be any positive real number. M can not be used as a model parameter.

191
Device Parameters

Table 2-39. Inductor Device Instance Parameters

Parameter Description Units Default
IC Initial current through device A 0
L Inductance henry 0
M Multiplicity Factor – 1
TC1 Linear Temperature Coefficient ◦ C −1 0
TC2 Quadratic Temperature Coefficient ◦ C −2 0
◦C Ambient
TEMP Device temperature
Temperature

Model Parameters

Table 2-40. Inductor Device Model Parameters

Parameter Description Units Default
IC Initial current through device A 0
L Inductance Multiplier – 1
TC1 First order temperature coeff. ◦ C −1 0
TC2 Second order temperature coeff. ◦ C −2 0
TNOM Reference temperature ◦C 27

In addition to the parameters shown in the table, the inductor supports a vector parameter for the
temperature correction coefficients. TC1=<linear coefficient> and TC2=<quadratic
coefficient> may therefore be specified compactly as TC=<linear coefficient>,<quadratic
coefficient>.

Inductor Equations

Inductance Value Formula If [model name] is specified, then the inductance is given by:

L𝑏𝑎𝑠𝑒 · L · (1 + TC1 · (𝑇 − 𝑇0 ) + TC2 · (𝑇 − 𝑇0 ) 2 )

where L𝑏𝑎𝑠𝑒 is the base inductance specified on the device line and is normally positive (though it can be
negative, but not zero). L is the inductance multiplier specified in the model card. 𝑇0 is the nominal
temperature (set using TNOM option).

192
2.3.6. Mutual Inductors

Symbol

Instance Form K<name> L<inductor name> [L<inductor name>*]

+ <coupling value> [model name]

Model Form .MODEL <model name> CORE [model parameters]

Examples ktran1 l1 l2 l3 1.0

KTUNED L3OUT L4IN .8
KTRNSFRM LPRIMARY LSECNDRY 1
KXFRM L1 L2 L3 L4 .98 KPOT_3C8

Parameters and
Options inductor name
Identifies the inductors to be coupled. The inductors are coupled and in the
dot notation the dot is placed on the first node of each inductor. The polarity
is determined by the order of the nodes in the L devices and not by the order
of the inductors in the K statement.
If more than two inductors are given on a single K line, each inductor is
coupled to all of the others using the same coupling value.
coupling value
The coefficient of mutual coupling, which must be between −1.0 and 1.0.
This coefficient is defined by the equation
𝑀𝑖 𝑗
<coupling value> = √
𝐿𝑖 𝐿 𝑗
where
𝐿 𝑖 is the inductance of the 𝑖th named inductor in the K-line
𝑀𝑖 𝑗 is the mutual inductance between 𝐿 𝑖 and 𝐿 𝑗
For transformers of normal geometry, use 1.0 as the value. Values less than
1.0 occur in air core transformers when the coils do not completely overlap.
model name
If model name is present, four things change:
• The mutual coupling inductor becomes a nonlinear, magnetic core
device.
• The inductors become windings, so the number specifying inductance
now specifies the number of turns.
• The list of coupled inductors could be just one inductor.

193
 • If two or more inductors are listed, each inductor is coupled to all others
through the magnetic core.
• A model statement is required to specify the model parameters.

Comments Lead currents and power calculations are supported for the component inductors in
both linear and nonlinear mutual inductors. They are not supported for the composite
mutual inductor though. So, if L1 is a component inductor for mutual inductor K1,
then requests for I(L1), P(L1) and W(L1) will return lead current and power values
as defined in Section 2.3.5. However, any usage of I(K1), P(K1) and W(K1) will
result in a Xyce netlist parsing error.

Model Parameters

Table 2-41. Nonlinear Mutual Inductor Device Model Parameters

Parameter Description Units Default
A Thermal energy parameter A/m 1000
ALPHA Domain coupling parameter – 5e-05
AREA Mean magnetic cross-sectional area cm2 0.1
BETAH Modeling constant – 0.0001
BETAM Modeling constant – 3.125e-05
BHSIUNITS Flag to report B and H in SI units – 0
C Domain flexing parameter – 0.2
Value below which domain flexing parameter will be
CLIM – 0.005
treated as zero.
Use constant scaling factor to smooth voltage
CONSTDELVSCALING V false
difference over first inductor
Smoothing coefficient for voltage difference over first
DELVSCALING V 1000
inductor
Flag to factor the saturation magnetization from the
FACTORMS – 0
magnetics equation.
GAP Effective air gap cm 0
INCLUDEMEQU Flag to include the magnetics in the solution. – true
K Domain anisotropy parameter A/m 500
KIRR Domain anisotropy parameter A/m 500
LEVEL for pspice compatibility – ignored – 0
MEQNSCALING M-equation scaling – 1
MS Saturation magnetization A/m 1e+06
MVARSCALING M-variable scaling. – 1
OUTPUTSTATEVARS Flag to save state variables – 0
PACK for pspice compatibility – ignored – 0
PATH Total mean magnetic path cm 1

194
 Table 2-41. Nonlinear Mutual Inductor Device Model Parameters
Parameter Description Units Default
PZEROTOL Tolerance for nonlinear zero crossing – 0.1
REQNSCALING R-equation scaling – 1
RVARSCALING R-variable scaling – 1
TC1 First order temperature coeff. – 0
TC2 Second order temperature coeff. – 0
TNOM Reference temperature ◦C 27

Note that Xyce’s default value for the GAP parameter as zero. Some simulators will use non-zero values of
the GAP as a default. When using netlists from other simulators in Xyce, ensure that the default parameters
are consistent.

Special Notes The coupling coefficient of the linear mutual inductor (i.e. a mutual inductor without a
core model) is permitted to be a time- or solution variable-dependent expression. This is intended to allow
simulation of electromechnical devices in which there might be moving coils that interact with fixed coils.
Additionally, for linear mutual inductors, different coupling terms can be applied to different pairs of
inductors with this syntax:

L1 1 2 2.0e-3
L2 0 3 8.1e-3
L3 3 4 8.1e-3
ktran1 l1 l2 0.7
ktran2 l2 l3 0.9
ktran3 l1 l3 0.99

Nonlinear mutual inductors can output 𝐵(𝑡) and 𝐻 (𝑡) variables so that one can plot 𝐵 − 𝐻 loops. On the
.print line the 𝐵 and 𝐻 variables are accessible using the node output syntax as in n(
non-linear-inductor-name_b ) for 𝐵 and n( non-linear-inductor-name_h ) for 𝐻. A
confusing aspect of this is that the non-linear inductor name is the internal name used by Xyce. For
example, consider this circuit which defines a nonlinear mutual inductor at both the top level of the circuit
and within a subcircuit:

* Test Circuit for Mutually Coupled Inductors

VS 0 1 SIN(0 169.7 60HZ)

R1 1 2 1K
R2 3 0 1K
L1 2 0 10
L2 3 0 20
K1 L1 L2 0.75 txmod
.model txmod core

195
.subckt mysub n1 n2 n3
r1s n1 n2 1000
r2s n3 0 1000
L1s n2 0 10
L2s n3 0 20
k1s L1s L2s 0.75 txmod
.ends

xtxs 1 4 5 mysub

.TRAN 100US 25MS

* output the current through each inductor and the B & H values.
.PRINT TRAN I(L1) I(L2) n(ymin!k1_b) n(ymin!k1_h)
+ I(xtxs:L1s) I(xtxs:l2s) n(xtxs:ymin!k1s_b) n(xtxs:ymin!k1s_h)

.END

The internal, Xyce name of the non-linear mutual inductor is YMIN!K1 or ymin!k1 as the name is not
case-sensitive. The device k1s is declared within a subcircuit called xtxs. Thus, its full name is
xtxs:ymin!k1s. The reason for this is that both the linear and non-linear mutual inductors are devices that
are collections of other devices, inductors in this case. Rather than use one of the few remaining single
characters left to signify a new device, Xyce uses Y devices as an indicator of a extended device set, where
the characters after the Y denote the device type and then the device name. Here, ymin means a min
device which is a mutual-inductor, non-linear device. Thus, to print the 𝐵 or 𝐻 variable of the non-linear
mutual inductor called k1 one would use n(ymin!k1_b) and n(ymin!k1_h) respectively for a .print
line that looks like this:

.PRINT TRAN I(L1) I(L2) n(ymin!k1_b) n(ymin!k1_h)

And if the mutual inductor is in a subcircuit called xtxs then the .print line would look like this:

.PRINT TRAN I(xtxs:L1s) I(xtxs:l2s) n(xtxs:ymin!k1s_b) n(xtxs:ymin!k1s_h)

The above example also demonstrates how one outputs the current through inductors that are part of mutual
inductors. The syntax is I( inductor name ).
Note that while MKS units are used internally in Xyce, 𝐵 and 𝐻 are output by default in the SI units of
Gauss for 𝐵 and Oersted for 𝐻. To convert 𝐵 to units of Tesla divide Xyce’s output by 10, 000. To convert 𝐻
to units of 𝐴/𝑚 divide Xyce’s output by 4𝜋/1000. Additionally, one can set the .model CORE parameter
BHSIUNITS to 1 to force 𝐵 and 𝐻 to be output in MKS units.
Finally, one can access the 𝐵 and 𝐻 data via the .model CORE line. On the nonlinear mutual inductor’s
.model line, set the option OUTPUTSTATEVARS=1. This will cause Xyce to create a unique file for each
nonlinear mutual inductor that uses this .model line with a name of the form Inductor_device_name.
There are five columns of data in this file: time (𝑡), magnetic moment (𝑀), total current flux (𝑅), flux
density (𝐵) and magnetic field strength (𝐻). As with data output on the .print line, SI units are used such
that 𝐵 is output with units of Gauss and 𝐻 in Oersted. As mentioned earlier, setting the model flag
BHSIUNITS to 1 causes the output of 𝐵 and 𝐻 uses MKS units of Tesla and 𝐴/𝑚 respectively.

196
Mutual Inductor Equations The voltage to current relationship for a set of linearly coupled inductors
is:
𝑁
∑︁ √︁ 𝑑𝐼 𝑗
𝑉𝑖 = 𝑐𝑖 𝑗 𝐿 𝑖 𝐿 𝑗 (2.3)
𝑗=1
𝑑𝑡

Here, 𝑉𝑖 is the voltage drop across the 𝑖th inductor in the coupled set. The coupling coefficient between a
pair of inductors is 𝑐 𝑖 𝑗 with a value typically near unity and 𝐿 is the inductance of a given inductor which
has units of Henry’s (1 Henry = 1𝐻 = 𝑉 𝑜𝑙𝑡 · 𝑠/𝐴𝑚 𝑝)

For nonlinearly coupled inductors, the above equation is expanded to the form:

   𝑁
 ∑︁
ℓ𝑔 𝑑𝐼 𝑗
𝑉𝑖 = 1 + 1 − 𝑃(𝑀, 𝐼1 ...𝐼 𝑁 ) 𝐿𝑜 𝑖 𝑗 (2.4)
ℓ𝑡 𝑗=1
𝑑𝑡

This is similar in form to the linearly coupled inductor equation. However, the coupling has become more
complicated as it now depends on the magnetic moment created by the current flow, 𝑀. Additionally, there
are geometric factors, ℓ𝑔 and ℓ𝑡 which are the effective air gap and total mean magnetic path for the coupled
inductors. The matrix of terms, 𝐿𝑜𝑖 𝑗 is defined as

𝜇0 𝐴𝑐 𝑁𝑖 𝑁 𝑗
𝐿𝑜𝑖 𝑗 = (2.5)
ℓ𝑡

and it represents the physical coupling between inductors 𝑖 and 𝑗. In this expression, 𝑁𝑖 is the number of
windings around the core of inductor 𝑖, 𝜇0 is the magnetic permeability of free space which has units of
Henries per meter and a value of 4𝜋 × 10−7 and 𝐴𝑐 is the mean magnetic cross-sectional area.

The magnetic moment, 𝑀 is defined by:

𝑁
𝑑𝑀 1 ∑︁ 𝑑𝐼𝑖
= 𝑃 𝑁𝑖 (2.6)
𝑑𝑡 ℓ𝑡 𝑖=1 𝑑𝑡

and the function 𝑃 is defined as:

′ + (1 − 𝑐)𝑀 ′
𝑐𝑀𝑎𝑛 𝑖𝑟𝑟
𝑃=   (2.7)
ℓ𝑔 ′ ℓ𝑔 ′
1+ ℓ𝑡 − 𝛼 𝑐𝑀𝑎𝑛 + ℓ𝑡 (1 − 𝑐)𝑀𝑖𝑟𝑟

If 𝑐 < CLIM, then 𝑐 is treated as zero in the above equation and Xyce simplifies the formulation. In this
case, the magnetic-moment equation will not be needed and it will be be dropped form the formulation.
One can controll this behavior by modifying the value of CLIM.

197
The remaining functions are:

′ 𝑀𝑠 𝐴
𝑀𝑎𝑛 = (2.8)
( 𝐴 + |𝐻𝑒 |) 2
𝐻𝑒 = 𝐻 + 𝛼𝑀 (2.9)
ℓ𝑔
𝐻 = 𝐻𝑎 𝑝 𝑝 − 𝑀 (2.10)
ℓ𝑡
𝑁
1 ∑︁
𝐻𝑎 𝑝 𝑝 = 𝑁 𝑖 𝐼𝑖 (2.11)
ℓ𝑡 𝑖=1
′ Δ𝑀 𝑠𝑔𝑛(𝑞) + |Δ𝑀 |
𝑀𝑖𝑟𝑟 = (2.12)
2 (𝐾𝑖𝑟𝑟 − 𝛼|Δ𝑀 |)
Δ𝑀 = 𝑀𝑎𝑛 − 𝑀 (2.13)
𝑀𝑠 𝐻 𝑒
𝑀𝑎𝑛 = (2.14)
𝐴 + |𝐻𝑒 |
𝑞 = DELVSCALINGΔ𝑉 (2.15)

Xyce dynamically modifies DELVSCALING to be 1000/ Maximum Voltage Drop over the first inductor.
This typically produces accurate results for both low voltage and high voltage applicaitons. However, it is
possible to use a fixed scaling by setting the model parameter CONSTDELVSCALING to true and then
setting DELVSCALING to the desired scaling value.

In Xyce’s formulation, we define 𝑅 as:

𝑁
𝑑𝐻 𝑎 𝑝 𝑝 1 ∑︁ 𝑑𝐼𝑖
𝑅= = 𝑁𝑖 (2.16)
𝑑𝑡 ℓ𝑡 𝑖=1 𝑑𝑡

This simplifies the 𝑀 equation to:

𝑑𝑀
= 𝑃𝑅 (2.17)
𝑑𝑡
Xyce then solves for the additional variables 𝑀 and 𝑅 when modeling a nonlinear mutual inductor device.

B-H Loop Calculations To calculate 𝐵-𝐻 loops, 𝐻 is used as defined above and 𝐵 is a derived quantity
calculated by:

𝐵 = 𝜇0 (𝐻 + 𝑀) (2.18)
   
ℓ𝑔
= 𝜇0 𝐻 𝑎 𝑝 𝑝 + 1 − 𝑀 (2.19)
ℓ𝑡

Converting Nonlinear to Linear Inductor Models At times one may have a model for nonlinear
mutual inductor, but wish to use a simpler linear model in a given circuit. To convert a non-linear model to
an equivalent linear form, one can start by equating the coupling components of equations 2.3 and 2.4 as:

   
√︁ ℓ𝑔
𝑐𝑖 𝑗 𝐿𝑖 𝐿 𝑗 = 1 + 1 − 𝑃(𝑀, 𝐼1 ...𝐼 𝑁 ) 𝐿𝑜𝑖 𝑗 (2.20)
ℓ𝑡

198
In the above relationship, 𝑖 and 𝑗 represent the 𝑖th and 𝑗th inductors. Since we would like to equate the 𝑖th
inductor’s nonliner properties to its linear properties, we will substitute 𝑖 → 𝑗 and simplify assuming steady
state where 𝑑/𝑑𝑡 = 0 and 𝑀 (𝑡) = 0.


  
1 



ℓ𝑔
  𝑐 𝑀𝑠 
 𝜇 𝐴𝑐 2

𝐿𝑖 = 1+ 1−  𝐴  (2.21)

𝑁
ℓ𝑡 𝑖
 
𝑐 𝑖𝑖 
 ℓ𝑡  1 + ℓ𝑔 − 𝛼 𝑀𝑠  

  ℓ𝑡 𝐴  
  
In the above equatin, 𝑐 𝑖𝑖 represents the coupling coefficient between the 𝑖th inductor with itself. This will
likely be 1 unless there are very unusual geometry considerations. Note, that the terms 𝐴, 𝑀 𝑠, 𝐴𝑐 , 𝜇, ℓ𝑔 and
ℓ𝑡 all have units of length within them and must use the same unit for this relationship to be valid.
Specifically, 𝜇 has units of Henery’s per meter and 𝐴 and 𝑀 𝑠 have units of Amps per meter. 𝐴𝑐 , ℓ𝑔 and ℓ 𝑝
have units of length2 and length respectively, but the length unit used in the model statement is 𝑐𝑚 2 and 𝑐𝑚
respectively. Thus, one must use consistent units such as meters for 𝐴𝑐 , ℓ𝑔 and ℓ 𝑝 in equation 2.21 for a
valid inductance approximation.

199
2.3.7. Resistor

Symbol

Instance Form R<name> <(+) node> <(-) node> [model name] [value] [device
parameters]

Model Form .MODEL <model name> R [model parameters]

.MODEL <model name> RES [model parameters]

Examples R1 1 2 2K TEMP=27
RM 4 5 R=4e3 M=2
RSOLDEP 2 0 R={1.0+scalar*V(1)}
RLOAD 3 6 RTCMOD 4.540 TEMP=85
.MODEL RTCMOD R (TC1=.01 TC2=-.001)
RSEMICOND 2 0 RMOD L=1000u W=1u
.MODEL RMOD R (RSH=1)

Parameters and
Options (+) node
(-) node
Polarity definition for a positive voltage across the resistor. The first node is
defined as positive. Therefore, the voltage across the component is the first
node voltage minus the second node voltage. Positive current flows from the
positive node (first node) to the negative node (second node).
model name
If [model name] is omitted, then [value] is the resistance in Ohms. If
[model name] is given then the resistance is determined from the model
parameters; see the resistance value formula below.
value
Positional specification of device parameter R (resistance). Alternately, this
can be specified as a parameter, R=<value>, or in the (optional) model.
device parameters
Parameters listed in Table 2-42 may be provided as space separated
<parameter>=<value> specifications as needed. Any number of
parameters may be specified.

Comments Resistors can have either positive or negative resistance values (R). A zero resistance
value (R) is also allowed.
The power dissipated in the resistor is calculated with 𝐼 · Δ𝑉 where the voltage drop is
calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− . The power accessors
(P() and W()) are supported for both the level 1 resistor and the level 2 (thermal)
resistor.

200
For compatibility with PSpice, either R or RES can be used in a .MODEL statement for a
resistor.
The Multiplicity Factor (M) can be used to specify multiple, identical resistors in
parallel. The effective resistance becomes R/M. The M value need not be an integer. It
can be any positive real number. M can not be used as a model parameter.

201
Device Parameters

Table 2-42. Resistor Device Instance Parameters

Parameter Description Units Default
Device Temperature – For compatibility only. ◦C
DTEMP 0
Parameter is NOT used
L Length m 0
M Multiplicity Factor – 1
R Resistance ˙ 1000
TC1 Linear Temperature Coefficient ◦ C −1 0
TC2 Quadratic Temperature Coefficient ◦ C −2 0
TCE Exponential Temperature Coefficient %/◦ C 0
◦C Ambient
TEMP Device temperature
Temperature
W Width m 0

In addition to the parameters shown in the table, the resistor supports a vector parameter for the temperature
correction coefficients. TC1=<linear coefficient> and TC2=<quadratic coefficient> may
therefore be specified compactly as TC=<linear coefficient>,<quadratic coefficient>.

Model Parameters

Table 2-43. Resistor Device Model Parameters

Parameter Description Units Default
DEFW Default Instance Width m 1e-05
NARROW Narrowing due to side etching m 0
R Resistance Multiplier – 1
RSH Sheet Resistance ˙ 0
TC1 Linear Temperature Coefficient ◦ C −1 0
TC2 Quadratic Temperature Coefficient ◦ C −2 0
TCE Exponential Temperature Coefficient %/◦ C 0
◦C Ambient
TNOM Parameter Measurement Temperature
Temperature

Note: There is no model parameter for Default Instance Length. The use of the semiconductor resistor
model requires the user to specify a non-zero value for the instance parameter L.

Resistor Equations

202
Resistance Value Formulas If the R parameter is given on the device instance line then that value is
used.
If the R parameter is not given then the semiconductor resistor model will be used if the L instance
parameter and the RSH model parameter are given and both are non-zero. In that case the resistance will be
as follows. (Note: If W is not given on the instance line then the value for the model parameter DEFW will
be used instead.)

[L − NARROW]
RSH
[W − NARROW]

If neither of these two cases apply then the default value for the R parameter will be used.

Temperature Dependence If TCE is specified as either an instance or model parameter for the Level 1
resistor then the resistance at temperature 𝑇 is given by (where the resistance at the nominal temperature
(𝑇0 ) was defined above in the resistance value formulas):

R · 𝑝𝑜𝑤(1.01, TCE · (𝑇 − 𝑇0 ))

otherwise the resistance is given by:

R · (1 + TC1 · (𝑇 − 𝑇0 ) + TC2 · (𝑇 − 𝑇0 ) 2 )

Thermal (level=2) Resistor Xyce supports a thermal resistor model, which is associated with level=2.

Thermal Resistor Instance Parameters

Table 2-44. Resistor Device Instance Parameters

Parameter Description Units Default
A Area of conductor m2 0
DENSITY Resistor material density (unused) kg/m3 0
HEATCAPACITY Resistor material volumetric heat capacity 𝐽/(m3 K) 0
L Length of conductor m 0
M Multiplicity Factor – 1
OUTPUTINTVARS Debug Output switch – false
R Resistance ˙ 1000
RESISTIVITY Resistor material resistivity ˙m 0
◦C Ambient
TEMP Device temperature
Temperature
THERMAL_A Area of material thermally coupled to conductor m2 0
Volumetric heat capacity of material thermally
THERMAL_HEATCAPACITY 𝐽/(m3 K) 0
coupled to conductor

203
 Table 2-44. Resistor Device Instance Parameters
Parameter Description Units Default
THERMAL_L Length of material thermally coupled to conductor m 0
W Width of conductor m 0

Thermal Resistor Model Parameters

Table 2-45. Resistor Device Model Parameters

Parameter Description Units Default
DEFW Default Instance Width m 1e-05
DENSITY Resistor material density (unused) kg/m3 0
HEATCAPACITY Resistor material volumetric heat capacity 𝐽/(m3 K) 0
NARROW Narrowing due to side etching m 0
R Resistance Multiplier – 1
RESISTIVITY Resistor material resistivity ˙m 0
RSH Sheet Resistance ˙ 0
TC1 Linear Temperature Coefficient ◦ C −1 0
TC2 Quadratic Temperature Coefficient ◦ C −2 0
TCE Exponential Temperature Coefficient %/◦ C 0
Volumetric heat capacity of material thermally
THERMAL_HEATCAPACITY 𝐽/(m3 K) 0
coupled to conductor
◦C Ambient
TNOM Nominal device temperature
Temperature

The temperature model for the thermal resistor will be enabled if the A and L instance parameters are given
and the parameters HEATCAPACITY and RESISTIVITY are also given as a pair of either instance
parameters or model parameters. Otherwise, the resistance value and temperature dependence of the Level
2 resistor will follow the equations for the Level 1 resistor given above, with the caveat that TCE is only
allowed as a model parameter for the Level 2 resistor.

If the temperature model for the Level 2 resistor is enabled, then the resistance (𝑅) is given by the
following, where the RESISTIVITY can be a temperature-dependent expression:

RESISTIVITY · L
A

The rate-of-change (𝑑𝑇/𝑑𝑡) of the temperature (𝑇) of the thermal resistor with time is then given by the
following where 𝑖0 is the current through the resistor:

𝑖0 · 𝑖0 · 𝑅
(A · L · HEATCAPACITY) + (THERMAL_A · THERMAL_L · THERMAL_HEATCAPACITY)

204
Solution-Dependent Resistor If the resistance (R) is set equal to an expression then a
“solution-dependent” resistor is used, where the resistor is a function of other simulation variables. The
formulas for temperature-dependence, given above, then use that calculated R value.
The restrictions for this solution dependent resistors are:
• The expression used for R must only use solution variables, which are node voltages and also branch
currents for source devices. It may not use device lead currents, which are post-processed quantities
that are not solution variables.
• The expression must not use time derivatives.
• Resistance (R) is the only instance or model parameters that are allowed to be solution-dependent.

205
2.3.8. Diode

Symbol

Instance Form D<name> <(+) node> <(-) node> <model name> [area value]

Model Form .MODEL <model name> D [model parameters]

Examples DCLAMP 1 0 DMOD

D2 15 17 SWITCH 1.5

Parameters and
Options
(+) node
(-) node
The anode and the cathode.
area value
Scales IS, ISR, IKF, RS, CJO, and IBV, and has a default value of 1. IBV and
BV are both specified as positive values.
PJ value
Used in computing the junction sidewall effects, and has a default value of
zero (no sidewall effects).

Comments The diode is modeled as an ohmic resistance (RS/area) in series with an intrinsic
diode. Positive current is current flowing from the anode through the diode to the
cathode.

The power through the diode is calculated with 𝐼 · Δ𝑉 where the voltage drop is
calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− . This formula may
differ from other simulators, such as HSPICE.

Diode Operating Temperature Model parameters can be assigned unique measurement temperatures
using the TNOM model parameter.

Diode level selection Several distinct implementations of the diode are available. These are selected by
using the LEVEL model parameter. The default implementation is based on SPICE 3F5, and may be
explicitly specified using LEVEL=1 in the model parameters, but is also selected if no LEVEL parameter is
specified. The PSpice implementation [2] is obtained by specifying LEVEL=2. The Xyce LEVEL=200 diode
is the JUNCAP200 model. The Xyce LEVEL=2002 diode is the DIODE_CMC model version 2.0.0.

206
Level 1 and 2 Diode Instance Parameters

Table 2-46. Diode Device Instance Parameters

Parameter Description Units Default
Area scaling value (scales IS, ISR, IKF, RS, CJ0, and
AREA – 1
IBV)
IC – 0
M multiplicity factor – 1
logical
OFF Initial voltage drop across device set to zero 0
(T/F)
PJ Perimeter scaling value – 0
Ambient
TEMP Device temperature –
Temperature

Level 1 and 2 Diode Model Parameters

Table 2-47. Diode Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
BV Reverse breakdown "knee" voltage V 1e+99
CJ Zero-bias p-n depletion capacitance F 0
CJ0 Zero-bias p-n depletion capacitance F 0
CJO Zero-bias p-n depletion capacitance F 0
CJP Sidewall junction capacitance (alias for CJSW) F 0
CJSW Sidewall junction capacitance F 0
EG Bandgap voltage (barrier height) eV 1.11
FC Forward-bias depletion capacitance coefficient – 0.5
Forward-bias sidewall depletion capacitance
FCS – 0.5
coefficient
IBV Reverse breakdown "knee" current A 0.001
IBVL Low-level reverse breakdown "knee" current A 0
IKF High-injection "knee" current A 0
IS Saturation current A 1e-14
ISR Recombination current parameter A 0
JS Saturation current A 1e-14
JSW Sidewall Saturation current A 0
KF Flicker noise coefficient – 0
M Grading parameter for p-n junction – 0.5
MJSW Grading parameter for sidewall junction – 0.33

207
 Table 2-47. Diode Device Model Parameters
Parameter Description Units Default
N Emission coefficient – 1
NBV Reverse breakdown ideality factor – 1
NBVL Low-level reverse breakdown ideality factor – 1
NR Emission coefficient for ISR – 2
NS Sidewall emission coefficient – 1
PHP Potential for sidewall junction V 1
RS Parasitic resistance ˙ 0
TBV1 BV temperature coefficient (linear) ◦ C −1 0
TBV2 BV temperature coefficient (quadratic) ◦ C −2 0
TIKF IKF temperature coefficient (linear) ◦ C −1 0
Ambient
TNOM –
Temperature
TRS RS temperature coefficient (linear) (alias for TRS1) ◦ C −1 0
TRS1 RS temperature coefficient (linear) ◦ C −1 0
TRS2 RS temperature coefficient (quadratic) ◦ C −2 0
TT Transit time s 0
VB Reverse breakdown "knee" voltage V 1e+99
VJ Potential for p-n junction V 1
VJSW Potential for sidewall junction (alias for PHP) V 1
XTI IS temperature exponent – 3

JUNCAP200 (level=200) Parameters The JUNCAP200 model has the instance and model parameters
in the tables below. Complete documentation of JUNCAP200 may be found at
http://www.cea.fr/cea-tech/leti/pspsupport/Documents/juncap200p5_summary.pdf.
The JUNCAP200 device supports output of the internal variables in table 2-50 on the .PRINT line of a
netlist. To access them from a print line, use the syntax N(<instance>:<variable>) where
“<instance>” refers to the name of the specific level 200 D device in your netlist.

Table 2-48. JUNCAP200 Diode Device Instance Parameters

Parameter Description Units Default
AB Junction area m2 1e-12
LG Gate-edge part of junction perimeter m2 1e-06
LS STI-edge part of junction perimeter m2 1e-06
M Alias for MULT — 1
MULT Number of devices in parallel — 1

208
 Table 2-49. JUNCAP200 Diode Device Model Parameters
Parameter Description Units Default
Band-to-band tunneling prefactor of bottom
CBBTBOT A/V3 1e-12
component
Band-to-band tunneling prefactor of gate-edge
CBBTGAT Am/V3 1e-18
component
Band-to-band tunneling prefactor of STI-edge
CBBTSTI Am/V3 1e-18
component
Zero-bias capacitance per unit-of-area of bottom
CJORBOT F/m2 0.001
component
Zero-bias capacitance per unit-of-length of gate-edge
CJORGAT F/m 1e-09
component
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTI F/m 1e-09
component
CSRHBOT Shockley-Read-Hall prefactor of bottom component A/m3 100
CSRHGAT Shockley-Read-Hall prefactor of gate-edge component A/m2 0.0001
CSRHSTI Shockley-Read-Hall prefactor of STI-edge component A/m2 0.0001
Trap-assisted tunneling prefactor of bottom
CTATBOT A/m3 100
component
Trap-assisted tunneling prefactor of gate-edge
CTATGAT A/m2 0.0001
component
Trap-assisted tunneling prefactor of STI-edge
CTATSTI A/m2 0.0001
component
Temperature offset with respect to ambient
DTA K 0
temperature
Normalization field at the reference temperature for
FBBTRBOT Vm−1 1e+09
band-to-band tunneling of bottom component
Normalization field at the reference temperature for
FBBTRGAT Vm−1 1e+09
band-to-band tunneling of gate-edge component
Normalization field at the reference temperature for
FBBTRSTI Vm−1 1e+09
band-to-band tunneling of STI-edge component
Fraction below which junction capacitance
FJUNQ — 0.03
components are considered negligible
FREV Coefficient for reverse breakdown current limitation — 1000
Saturation current density at the reference temperature
IDSATRBOT A/m2 1e-12
of bottom component
Saturation current density at the reference temperature
IDSATRGAT A/m 1e-18
of gate-edge component
Saturation current density at the reference temperature
IDSATRSTI A/m 1e-18
of STI-edge component
Maximum current up to which forward current
IMAX A 1000
behaves exponentially
LEVEL Model level must be 200 — 200

209
 Table 2-49. JUNCAP200 Diode Device Model Parameters
Parameter Description Units Default
Effective mass (in units of m0) for trap-assisted
MEFFTATBOT — 0.25
tunneling of bottom component
Effective mass (in units of m0) for trap-assisted
MEFFTATGAT — 0.25
tunneling of gate-edge component
Effective mass (in units of m0) for trap-assisted
MEFFTATSTI — 0.25
tunneling of STI-edge component
PBOT Grading coefficient of bottom component — 0.5
Breakdown onset tuning parameter of bottom
PBRBOT V 4
component
Breakdown onset tuning parameter of gate-edge
PBRGAT V 4
component
Breakdown onset tuning parameter of STI-edge
PBRSTI V 4
component
PGAT Grading coefficient of gate-edge component — 0.5
Zero-temperature bandgap voltage of bottom
PHIGBOT V 1.16
component
Zero-temperature bandgap voltage of gate-edge
PHIGGAT V 1.16
component
Zero-temperature bandgap voltage of STI-edge
PHIGSTI V 1.16
component
PSTI Grading coefficient of STI-edge component — 0.5
Temperature scaling parameter for band-to-band
STFBBTBOT 1/K -0.001
tunneling of bottom component
Temperature scaling parameter for band-to-band
STFBBTGAT 1/K -0.001
tunneling of gate-edge component
Temperature scaling parameter for band-to-band
STFBBTSTI 1/K -0.001
tunneling of STI-edge component
Flag for JUNCAP-express; 0=full model, 1=express
SWJUNEXP — 0
model
TRJ Reference temperature ◦C 21
Type parameter, in output value 1 reflects n-type, -1
TYPE — 1
reflects p-type
Built-in voltage at the reference temperature of
VBIRBOT V 1
bottom component
Built-in voltage at the reference temperature of
VBIRGAT V 1
gate-edge component
Built-in voltage at the reference temperature of
VBIRSTI V 1
STI-edge component
VBRBOT Breakdown voltage of bottom component V 10
VBRGAT Breakdown voltage of gate-edge component V 10
VBRSTI Breakdown voltage of STI-edge component V 10

210
 Table 2-49. JUNCAP200 Diode Device Model Parameters
Parameter Description Units Default
Typical maximum junction voltage; usually about
VJUNREF V 2.5
2*VSUP
XJUNGAT Junction depth of gate-edge component m 1e-07
XJUNSTI Junction depth of STI-edge component m 1e-07

Table 2-50. Diode level 200 Output Variables

Parameter Description Units Default
vak Voltage between anode and cathode V none
cj Total source junction capacitance F none
cjbot Junction capacitance (bottom component) F none
cjgat Junction capacitance (gate-edge component) F none
cjsti Junction capacitance (STI-edge component) F none
ij Total source junction current A none
ijbot Junction current (bottom component) A none
ijgat Junction current (gate-edge component) A none
ijsti Junction current (STI-edge component) A none
si Total junction current noise spectral density A2 /Hz none
idsatsbot Total bottom saturation current A none
idsatssti Total STI-edge saturation current A none
idsatsgat Total gate-edge saturation current A none
cjosbot Total bottom capacity F none
cjossti Total STI-edge capacity F none
cjosgat Total gate-edge capacity F none
vbisbot built-in voltage of the bottom junction V none
vbissti built-in voltage of the STI-edge junction V none
vbisgat built-in voltage of the gate-edge junction V none

DIODE_CMC (level=2002) Parameters The DIODE_CMC model has the instance and model
parameters in the tables below. Complete documentation of DIODE_CMC may be found at
https://si2.org/standard-models.
The DIODE_CMC device supports output of the internal variables in table 2-53 on the .PRINT line of a
netlist. To access them from a print line, use the syntax N(<instance>:<variable>) where
“<instance>” refers to the name of the specific level 2002 D device in your netlist.

211
 Table 2-51. DIODE_CMC 2.0.0 Device Instance Parameters
Parameter Description Units Default
AB Junction area m2 1e-12
AREA Alias for AB m2 1e-12
LG Gate-edge part of junction perimeter m 0
LS STI-edge part of junction perimeter m 1e-06
MULT Number of devices in parallel — 1
PERIM Alias for LS m 1e-06
PJ Alias for LS m 1e-06

Table 2-52. DIODE_CMC 2.0.0 Device Model Parameters

Parameter Description Units Default
ABMAX maximum allowed junction area m2 1
ABMIN minimum allowed junction area m2 0
AF AF parameter for flicker noise — 1
Band-to-band tunneling prefactor of bottom
CBBTBOT A/V3 1e-12
component
Band-to-band tunneling prefactor of gate-edge
CBBTGAT Am/V3 1e-18
component
Band-to-band tunneling prefactor of STI-edge
CBBTSTI Am/V3 1e-18
component
Zero-bias capacitance per unit-of-area of bottom
CJORBOT F/m2 0.001
component
Zero-bias capacitance per unit-of-length of gate-edge
CJORGAT F/m 1e-09
component
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTI F/m 1e-09
component
CORECOVERY Flag for recovery equations; 0=original, 1=Hiroshima — 0
CSRHBOT Shockley-Read-Hall prefactor of bottom component A/m3 100
CSRHGAT Shockley-Read-Hall prefactor of gate-edge component A/m2 0.0001
CSRHSTI Shockley-Read-Hall prefactor of STI-edge component A/m2 0.0001
Trap-assisted tunneling prefactor of bottom
CTATBOT A/m3 100
component
Trap-assisted tunneling prefactor of gate-edge
CTATGAT A/m2 0.0001
component
Trap-assisted tunneling prefactor of STI-edge
CTATSTI A/m2 0.0001
component
DEPNQS Depletion delay time — 0
Temperature offset with respect to ambient
DTA C 0
temperature

212
 Table 2-52. DIODE_CMC 2.0.0 Device Model Parameters
Parameter Description Units Default
Normalization field at the reference temperature for
FBBTRBOT Vm−1 1e+09
band-to-band tunneling of bottom component
Normalization field at the reference temperature for
FBBTRGAT Vm−1 1e+09
band-to-band tunneling of gate-edge component
Normalization field at the reference temperature for
FBBTRSTI Vm−1 1e+09
band-to-band tunneling of STI-edge component
Fraction below which junction capacitance
FJUNQ — 0.03
components are considered negligible
FREV Additional parameter for current after breakdown — 1000
Saturation current density at the reference temperature
IDSATRBOT A/m2 1e-12
of bottom component
Saturation current density at the reference temperature
IDSATRGAT A/m 1e-18
of gate-edge component
Saturation current density at the reference temperature
IDSATRSTI A/m 1e-18
of STI-edge component
Maximum current up to which forward current
IMAX A 1000
behaves exponentially
INJ1 For carrier density — 1
INJ2 For carrier density in high-injection condition — 10
INJT Temp. co of carrier density in high-injection condition — 0
KF KF parameter for flicker noise — 0
LGMAX maximum allowed junction gate-edge m 1
LGMIN minimum allowed junction gate-edge m 0
LSMAX maximum allowed junction STI-edge m 1
LSMIN minimum allowed junction STI-edge m 0
Effective mass (in units of m0) for trap-assisted
MEFFTATBOT — 0.25
tunneling of bottom component
Effective mass (in units of m0) for trap-assisted
MEFFTATGAT — 0.25
tunneling of gate-edge component
Effective mass (in units of m0) for trap-assisted
MEFFTATSTI — 0.25
tunneling of STI-edge component
NDIBOT Doping concentration of drift region — 1e+16
NDIGAT Doping concentration of drift region — 1e+16
NDISTI Doping concentration of drift region — 1e+16
NFABOT ideality factor bottom component — 1
NFAGAT ideality factor gate-edge component — 1
NFASTI ideality factor STI-edge component — 1
NJDV Transition slope of emission coefficient — 0.1
NJH High-injection emission coefficient — 1

213
 Table 2-52. DIODE_CMC 2.0.0 Device Model Parameters
Parameter Description Units Default
NQS Carrier delay time — 5e-09
PBOT Grading coefficient of bottom component — 0.5
Breakdown onset tuning parameter of bottom
PBRBOT V 4
component
Breakdown onset tuning parameter of gate-edge
PBRGAT V 4
component
Breakdown onset tuning parameter of STI-edge
PBRSTI V 4
component
PGAT Grading coefficient of gate-edge component — 0.5
Zero-temperature bandgap voltage of bottom
PHIGBOT V 1.16
component
Zero-temperature bandgap voltage of gate-edge
PHIGGAT V 1.16
component
Zero-temperature bandgap voltage of STI-edge
PHIGSTI V 1.16
component
PSTI Grading coefficient of STI-edge component — 0.5
PT Alias for XTI — 3
REVISION Model revision — 0
Series resistance per unit-of-area of bottom
RSBOT — 0
component
RSCOM Common series resistance, no scaling — 0
Series resistance per unit-of-length of gate-edge
RSGAT — 0
component
Series resistance per unit-of-length of STI-edge
RSSTI — 0
component
SCALE Scale parameter — 1
SHRINK Scale parameter — 0
Temperature scaling parameter for band-to-band
STFBBTBOT 1/K -0.001
tunneling of bottom component
Temperature scaling parameter for band-to-band
STFBBTGAT 1/K -0.001
tunneling of gate-edge component
Temperature scaling parameter for band-to-band
STFBBTSTI 1/K -0.001
tunneling of STI-edge component
STRS Temperature scaling parameter for series resistance — 0
STVBRBOT1 Temp. co of breakdown voltage bottom component 1/K 0
STVBRBOT2 Temp. co of breakdown voltage bottom component — 0
STVBRGAT1 Temp. co of breakdown voltage gate-edge component 1/K 0
STVBRGAT2 Temp. co of breakdown voltage gate-edge component — 0
STVBRSTI1 Temp. co of breakdown voltage STI-edge component 1/K 0

214
 Table 2-52. DIODE_CMC 2.0.0 Device Model Parameters
Parameter Description Units Default
STVBRSTI2 Temp. co of breakdown voltage STI-edge component — 0
SUBVERSION Model subversion — 0
Flag for JUNCAP-express; 0=full model, 1=express
SWJUNEXP — 0
model
TAU Carrier lifetime — 2e-07
TAUT Temp. co of carrier lifetime — 0
TEMPMAX maximum allowed junction temp C 155
TEMPMIN minimum allowed junction temp C -55
TNOM Alias reference temperature C 21
TRJ Reference temperature C 21
TT Transit time s 0
Type parameter, in output value 1 reflects n-type, -1
TYPE — 1
reflects p-type
Built-in voltage at the reference temperature of
VBIRBOT V 1
bottom component
Built-in voltage at the reference temperature of
VBIRGAT V 1
gate-edge component
Built-in voltage at the reference temperature of
VBIRSTI V 1
STI-edge component
VBRBOT Breakdown voltage of bottom component V 10
VBRGAT Breakdown voltage of gate-edge component V 10
VBRSTI Breakdown voltage of STI-edge component V 10
VERSION Model version — 2
VFMAX maximum allowed forward junction bias V 0
Typical maximum junction voltage; usually about
VJUNREF — 2.5
2*VSUP
VRMAX maximum allowed reverse junction bias V 0
WI Length of drift region m 5e-06
XJUNGAT Junction depth of gate-edge component m 1e-07
XJUNSTI Junction depth of STI-edge component m 1e-07
XTI Temp. co of saturation current — 3

Table 2-53. Diode level 2002 Output Variables

Parameter Description Units Default
Voltage between anode and cathode excluding the
vak V none
series resistor
cj Total source junction capacitance F none

215
 Table 2-53. Diode level 2002 Output Variables
Parameter Description Units Default
cjbot Junction capacitance (bottom component) F none
cjgat Junction capacitance (gate-edge component) F none
cjsti Junction capacitance (STI-edge component) F none
ij Total source junction current A none
ijbot Junction current (bottom component) A none
ijgat Junction current (gate-edge component) A none
ijsti Junction current (STI-edge component) A none
si Total junction current noise spectral density A2 /Hz none
vrs Voltage across series resistor V none
sf Total junction flicker noise spectral density A2 /Hz none
sr Total series resistor thermal noise spectral density A2 /Hz none
rseries Series resistor V/A none
qrr Recovery charge C none

Level 1 Diode Equations The equations in this section use the following variables:

𝑉𝑑𝑖 = voltage across the intrinsic diode only

𝑉𝑡 ℎ = 𝑘 · 𝑇/𝑞 (thermal voltage)
𝑘 = Boltzmann’s constant
𝑞 = electron charge
𝑇 = analysis temperature (Kelvin)
𝑇0 = nominal temperature (set using TNOM option)
𝜔 = Frequency (Hz)

Other variables are listed above in the diode model parameters.

Level=1 The level 1 diode is based on the Spice3f5 level 1 model.

DC Current (Level=1) The intrinsic diode current consists of forward and reverse bias regions where
h   i
𝑉𝑑𝑖
IS · exp − 1 𝑉𝑑𝑖 > −3.0 · N𝑉𝑡 ℎ




 N𝑉𝑡 ℎ ,
𝐼𝐷 =
  3
3.0·N𝑉𝑡 ℎ
 −IS · 1.0 + 𝑉𝑑𝑖 ·𝑒 , 𝑉𝑑𝑖 < −3.0 · N𝑉𝑡 ℎ




When BV and an optional parameter IBV are explicitly given in the model statement, an exponential model
is used to model reverse breakdown (with a “knee” current of IBV at a “knee-on” voltage of BV). The
equation for 𝐼 𝐷 implemented by Xyce is given by

216
  
BVeff + 𝑉𝑑𝑖
𝐼 𝐷 = −IBVeff · exp − , 𝑉𝑑𝑖 ≤ BVeff ,
N𝑉𝑡 ℎ

where BVeff and IBVeff are chosen to satisfy the following constraints:

1. Continuity of 𝐼 𝐷 between reverse bias and reverse breakdown regions (i.e., continuity of 𝐼 𝐷 at
𝑉𝑑𝑖 = −BVeff ):
 3!
3.0 · N𝑉𝑡 ℎ
IBVeff = IS 1 −
𝑒 · BVeff

2. “Knee-on” voltage/current matching:

 
BVeff − BV
IBVeff · exp − = IBV
N𝑉𝑡 ℎ

Substituting the first expression into the second yields a single constraint on BVeff which cannot be solved
for directly. By performing some basic algebraic manipulation and rearranging terms, the problem of
finding BVeff which satisfies the above two constraints can be cast as finding the (unique) solution of the
equation
BVeff = 𝑓 (BVeff ), (2.22)

where 𝑓 (·) is the function that is obtained by solving for the BVeff term which appears in the exponential in
terms of BVeff and the other parameters. Xyce solves Eqn. 2.22 by performing the so-called Picard Iteration
procedure [11], i.e. by producing successive estimates of BVeff (which we will denote as BVeff 𝑘 ) according
to
BVeff 𝑘+1 = 𝑓 (BVeff 𝑘 )

starting with an initial guess of BVeff 0 = BV. The current iteration procedure implemented in Xyce can be
shown to guarantee at least six significant digits of accuracy between the numerical estimate of BVeff and
the true value.

In addition to the above, Xyce also requires that BVeff lie in the range BV ≥ BVeff ≥ 3.0N𝑉𝑡 ℎ . In terms of
IBV, this is equivalent to enforcing the following two constraints:

 3!
3.0 · N𝑉𝑡 ℎ
IS 1 − ≤ IBV (2.23)
𝑒 · BV
 

−3
 −3.0 · N𝑉𝑡 ℎ + BV
IS 1 − 𝑒 exp ≥ IBV (2.24)
N𝑉𝑡 ℎ

Xyce first checks the value of IBV to ensure that the above two constraints are satisfied. If Eqn. 2.23 is
violated, Xyce sets IBVeff to be equal to the left-hand side of Eqn. 2.23 and, correspondingly, sets BVeff to
−3.0 · N𝑉𝑡 ℎ . If Eqn. 2.24 is violated, Xyce sets IBVeff to be equal to the left-hand side of Eqn. 2.24 and,
correspondingly, sets BVeff to BV.

217
Capacitance (Level=1) The p-n diode capacitance consists of a depletion layer capacitance 𝐶𝑑 and a
diffusion capacitance 𝐶𝑑𝑖 𝑓 . The first is given by

  −M
 CJ · AREA 1 − 𝑉VJ


 𝑑𝑖
, 𝑉𝑑𝑖 ≤ FC · VJ
𝐶𝑑 =  
 CJ·AREA

F2 F3 + M 𝑉VJ
𝑑𝑖
, 𝑉𝑑𝑖 > FC · VJ


The diffusion capacitance (sometimes referred to as the transit time capacitance) is

𝑑𝐼 𝐷
𝐶𝑑𝑖 𝑓 = TT𝐺 𝑑 = TT
𝑑𝑉𝑑𝑖

where 𝐺 𝑑 is the junction conductance.

Sidewall currents and capacitances When the instance parameter PJ (perimeter scaling value) is
specified, the diode currents become the sum of the currents above (the “bottom” of the junction) and those
of the periphery (sidewall).

In normal forward and reverse bias regions, the sidewall currents are given by:
h   i
𝑉𝑑𝑖
ISatSW · exp − 1 𝑉𝑑𝑖 > −3.0 · NS𝑉𝑡 ℎ




 NS𝑉𝑡 ℎ ,
=
 3
𝐼 𝐷,𝑆𝑊 
3.0·NS𝑉𝑡 ℎ
 −ISatSW · 1.0 + , 𝑉𝑑𝑖 < −3.0 · NS𝑉𝑡 ℎ

 𝑉𝑑𝑖 ·𝑒


where ISatSW is the temperature-adjusted value of JSW multiplied by the perimeter PJ.

When the breakdown voltage BV has been given and the diode voltage is below −BV, the sidewall current
is:
 
BVeff + 𝑉𝑑𝑖
𝐼 𝐷,𝑠𝑤 = −ISatSW · exp − , 𝑉𝑑𝑖 ≤ BVeff,
NS𝑉𝑡 ℎ

The sidewall capacitances are computed as:

  −M
 𝑉𝑑𝑖
 CJSW · PJ 1 − PHP

 , 𝑉𝑑𝑖 ≤ FCS · PHP
𝐶𝑑,𝑠𝑤 =  
CJSW·PJ 𝑉
 F3SW + MJSW PHP 𝑑𝑖
, 𝑉𝑑𝑖 > FCS · PHP
 F2SW


Temperature Effects (Level=1) The diode model contains explicit temperature dependencies in the
ideal diode current, the generation/recombination current and the breakdown current. Further temperature
dependencies are present in the diode model via the saturation current 𝐼𝑆 , the depletion layer junction

218
capacitance 𝐶𝐽, the junction potential 𝑉𝐽 .
𝑘𝑇
𝑉𝑡 (𝑇) =
𝑞
𝑘TNOM
𝑉𝑡 𝑛𝑜𝑚 (𝑇) =
𝑞
𝛼𝑇 2
𝐸 𝑔 (𝑇) = 𝐸 𝑔0 −
𝛽+𝑇
𝛼TNOM2
𝐸 𝑔𝑁 𝑂𝑀 (𝑇) = 𝐸 𝑔0 −
TNOM + 𝛽
𝐸 𝑔 (𝑇) 𝐸 𝑔300
𝑎𝑟𝑔1(𝑇) = − +
2𝑘𝑇 2𝑘𝑇0
𝐸 𝑔𝑁 𝑂𝑀 (𝑇) 𝐸 𝑔300
𝑎𝑟𝑔2(𝑇) = − +
2𝑘TNOM 2𝑘𝑇0
   
𝑇
𝑝𝑏 𝑓 𝑎𝑐𝑡1(𝑇) = −2.0 · 𝑉𝑡 (𝑇) 1.5 · ln + 𝑞 · 𝑎𝑟𝑔1(𝑇)
𝑇0
   
TNOM
𝑝𝑏 𝑓 𝑎𝑐𝑡2(𝑇) = −2.0 · 𝑉𝑡 𝑛𝑜𝑚 (𝑇) 1.5 · ln + 𝑞 · 𝑎𝑟𝑔2(𝑇)
𝑇0
𝑇0
𝑝𝑏𝑜(𝑇) = (VJ − 𝑝𝑏 𝑓 𝑎𝑐𝑡2(𝑇))
TNOM
𝑇
𝑉𝐽 (𝑇) = 𝑝𝑏 𝑓 𝑎𝑐𝑡1(𝑇) + 𝑝𝑏𝑜(𝑇)
𝑇0
VJ − 𝑝𝑏𝑜(𝑇)
𝑔𝑚𝑎 𝑜𝑙𝑑 (𝑇) =
𝑝𝑏𝑜(𝑇)
𝑉𝐽 (𝑇) − 𝑝𝑏𝑜(𝑇)
𝑔𝑚𝑎 𝑛𝑒𝑤 (𝑇) =
𝑝𝑏𝑜(𝑇)
1.0 + M 4.0 × 10−4 (𝑇 − 𝑇0 ) − 𝑔𝑚𝑎 𝑛𝑒𝑤 (𝑇)


𝐶𝐽 (𝑇) = CJ0

1.0 + M 4.0 × 10−4 (TNOM − 𝑇0 ) − 𝑔𝑚𝑎 𝑜𝑙𝑑 (𝑇)
   
𝑇 EG XTI 𝑇
𝐼𝑆 (𝑇) = IS · exp − 1.0 · + · ln
TNOM N𝑉𝑡 (𝑇) N TNOM

where, for silicon, 𝛼 = 7.02 × 10−4 𝑒𝑉/𝐾, 𝛽 = 1108 𝐾 and 𝐸 𝑔0 = 1.16 𝑒𝑉.
For a more thorough description of p-n junction physics, see [9]. For a thorough description of the U.C.
Berkeley SPICE models see Reference [11].

219
2.3.9. Independent Current Source

Symbol

Instance Form I<name> <(+) node> <(-) node> [ [DC] <value> ]

+ [AC [magnitude value [phase value] ] ] [transient specification]

Examples ISLOW 1 22 SIN(0.5 1.0ma 1KHz 1ms)

IPULSE 1 3 PULSE(-1 1 2ns 2ns 2ns 50ns 100ns)
IPAT 2 4 PAT(5 0 0 1n 2n 5n b0101)

Parameters and
Options transient specification
There are five predefined time-varying functions for sources:
PULSE <parameters> Pulse waveform
SIN <parameters> Sinusoidal waveform
EXP <parameters> Exponential waveform
PAT <parameters> Pattern waveform
PWL <parameters> Piecewise linear waveform
SFFM <parameters> Frequency-modulated waveform

Comments Positive current flows from the positive node through the source to the negative node.
The power supplied or dissipated by the current source is calculated with 𝐼 · Δ𝑉 where
the voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
Dissipated power has a positive sign, while supplied power has a negative sign.
The default value is zero for the DC, AC, and transient values. None, any, or all of the
DC, AC, and transient values can be specified. The AC phase value is in degrees.

Transient Specifications This section outlines the available transient specifications. Δ𝑡 and 𝑇𝐹 are the
time step size and simulation end-time, respectively. Parameters marked as – must have a value specified for
them; otherwise a netlist parsing error will occur.

Pulse

PULSE(V1 V2 TD TR TF PW PER)

220
 Table 2-54. Pulse Parameters
Parameter Description Units Default
V1 Initial Value amp –
V2 Pulse Value amp 0.0
TD Delay Time s 0.0
TR Rise Time s Δ𝑡
TF Fall Time s Δ𝑡
PW Pulse Width s 𝑇𝐹
PER Period s 𝑇𝐹

Sine

SIN(V0 VA FREQ TD THETA PHASE)

Table 2-55. Sine Parameters

Parameter Description Units Default
V0 Offset amp –
VA Amplitude amp –
FREQ Frequency s−1 –
TD Delay s Δ𝑡
THETA Attenuation Factor s Δ𝑡
PHASE Phase degrees 0.0

The waveform is shaped according to the following equations, where 𝜙 = 𝜋 ∗ PHASE/180 :


𝑉0 , 0 < 𝑡 < 𝑇𝐷
𝐼=
𝑉0 + 𝑉 𝐴 sin[2𝜋 · FREQ · (𝑡 − 𝑇𝐷 ) + 𝜙] exp[−(𝑡 − 𝑇𝐷 ) · THETA], 𝑇𝐷 < 𝑡 < 𝑇𝐹

Exponent

EXP(V1 V2 TD1 TAU1 TD2 TAU2)

Table 2-56. Exponent Parameters

Parameter Description Units Default
V1 Initial Amplitude amp –
V2 Amplitude amp –
TD1 Rise Delay Time s 0.0
TAU1 Rise Time Constant s Δ𝑡

221
 Table 2-56. Exponent Parameters
Parameter Description Units Default
TD2 Delay Fall Time s TD1 +Δ𝑡
TAU2 Fall Time Constant s Δ𝑡

The waveform is shaped according to the following equations:



 𝑉1 , 0 < 𝑡 < TD1
 𝑉1 + (𝑉2 − 𝑉1 ){1 − exp[−(𝑡 − TD1)/TAU1]}, TD1 < 𝑡 < TD2


𝐼=

 𝑉1 + (𝑉2 − 𝑉1 ){1 − exp[−(𝑡 − TD1)/TAU1]}
+(𝑉1 − 𝑉2 ){1 − exp[−(𝑡 − TD2)/TAU2]}, TD2 < 𝑡 < 𝑇2




Pattern

PAT(VHI VLO TD TR TF TSAMPLE DATA R)

Table 2-57. Pattern Parameters

Parameter Description Units Default
VHI High Value amp –
VLO Low Value amp –
TD Delay Time s –
TR Rise Time s –
TF Fall Time s –
TSAMPLE Bit period s –
DATA Bit pattern – –
R Repeat – 0

The VHI, VLO, TD, TF, TF TSAMPLE and DATA parameters are all required, and hence have no default values.
Negative values for TD are supported. The R parameter is optional. For its default value of 0, the requested
bit pattern will occur once.
The DATA parameter is the requested bit-pattern. Only the 0’ and ‘1’ states are supported. The ‘M’ and ‘Z’
states are not supported. The DATA field should have a leading ‘b’ (or ‘B’) character (e.g., be specified as
‘b0101’ ).
For times earlier than TD, the waveform value is set by the first bit in DATA. For times after the end of the
(possibly repeated) pattern, the waveform value is set by the last bit in DATA. Piecewise linear interpolation
is used to generate the output value when transitioning between states.
The VHI, VLO, TD, TF, TF and TSAMPLE parameters are compatible with .STEP. The DATA and R parameters
are not.
The HSPICE parameters RB, ENCODE and RD_INIT, for the pattern source, are not supported.

222
Piecewise Linear

PWL T0 V0 [Tn Vn]*

PWL FILE "<name>" [TD=<timeDelay>] [R=<repeatTime>]

Table 2-58. Piecewise Linear Parameters

Parameter Description Units Default
T𝑛 Time at Corner s none
V𝑛 Current at Corner amp none
TD Time Delay s 0
R Repeat Time s none

When the FILE option is given, Xyce will read the corner points from the file specified in the <name> field.
This file should be a plain ASCII text file (or a .CSV file) with the time/current pairs. There should be one
pair per line, and the time and current values should be separated by whitespace or commas. As an
example, the file specified (e.g., ipwl.csv) could have these five lines:

0.00, 0.00
2.00, 3.00
3.00, 2.00
4.00, 2.00
4.01, 5.00

The corresponding example instance lines would be:

IPWL1 1 0 PWL 0S 0A 2S 3A 3S 2A 4S 2A 4.01S 5A

IPWL2 2 0 PWL FILE "ipwl.txt"
IPWL3 3 0 PWL file "ipwl.csv"
IPWL4 4 0 PWL FILE ipwl.csv

The double quotes around the file name are optional, as shown above.

It is a best practice to specify all of the time-current pairs in the PWL specification. However, for
compatibility with HSPICE and PSpice, if the user-specified list of time/current pairs omits the pair at
time=0 as the first pair in the list then Xyce will insert a pair at time=0 with the current value at the first
user-specified time value. As an example, this user-specified list:

2S 3A 3S 2A 4S 2A 4.01S 5A

would be implemented in Xyce as follows:

0S 3A 2S 3A 3S 2A 4S 2A 4.01S 5A

223
TD has units of seconds, and specifies the length of time to delay the start of PWL waveform. The default is
to have no delay, and TD is an optional parameter.
The Repeat Time (R) is an optional parameter. If R is omitted then the waveform will not repeat. If R is
included then the waveform will repeat until the end of the simulation. As examples, R=0 means repeat the
PWL waveform from time=0 to the last time (T 𝑁 ) specified in the waveform specification. (This would use
the time points 0s, 2s, 3s, 4s and 4.01s for the example waveform given above.). In general,
R=<repeatTime> means repeat the waveform from time equal to <repeatTime> seconds in the waveform
specification to the last time (T 𝑁 ) specified in the waveform specification. So, the <repeatTime> must be
greater than or equal to 0 and less than the last time point (T 𝑁 ). If the R parameter is used then it must have
a value.
The specification PWL FILE "<name>" R is illegal in Xyce as a shorthand for R=0. Also, the Xyce syntax
for PWL sources is not compatible with the PSpice REPEAT syntax for PWL sources. See section 6.1.12 for
more details.
The repeat time (R) does enable the specification of discontinuous piecewise linear waveforms. For
example, this waveform is a legal Xyce syntax.

IPWL1 1 0 PWL 0S 0A 2S 3A 3S 2A 4S 2A 4.01S 5A R=2

However, in general, discontinuous source waveforms may cause convergence problems.

Frequency Modulated

SFFM (V0 VA FC MDI FS)

Table 2-59. Frequency Modulated Parameters

Parameter Description Units Default
V0 Offset amp –
VA Amplitude amp –
FC Carrier Frequency hertz 1/TSTOP
MDI Modulation Index - 0
FS Signal Frequency hertz 1/TSTOP

TSTOP is the final time, as entered into the transient (.TRANS) command. The waveform is shaped
according to the following equation:

𝐼 = 𝑉0 + 𝑉 𝐴 · sin(2𝜋 · FC · TIME + MDI · sin(2𝜋 · FS · TIME))

where TIME is the current simulation time.

224
2.3.10. Independent Voltage Source

Symbol

Instance Form V<name> <(+) node> <(-) node> [ [DC] <value> ]

+ [AC [magnitude value [phase value] ] ] [transient specification]

Examples VSLOW 1 22 SIN(0.5 1.0mV 1KHz 1ms)

VPULSE 1 3 PULSE(-1 1 2ns 2ns 2ns 50ns 100ns)
VPAT 2 4 PAT(5 0 0 1n 2n 5n b0101)

Parameters and
Options transient specification
There are five predefined time-varying functions for sources:
PULSE <parameters> Pulse waveform
SIN <parameters> Sinusoidal waveform
EXP <parameters> Exponential waveform
PAT <parameters> Pattern waveform
PWL <parameters> Piecewise linear waveform
SFFM <parameters> Frequency-modulated waveform

Comments Positive current flows from the positive node through the source to the negative node.
The power supplied or dissipated by the voltage source is calculated with 𝐼 · Δ𝑉 where
the voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
Dissipated power has a positive sign, while supplied power has a negative sign.
None, any, or all of the DC, AC, and transient values can be specified. The AC phase
value is in degrees.

Transient Specifications This section outlines the available transient specifications. Δ𝑡 and 𝑇𝐹 are the
time step size and simulation end-time, respectively. Parameters marked as – must have a value specified for
them; otherwise a netlist parsing error will occur.

Pulse

PULSE(V1 V2 TD TR TF PW PER)

225
 Table 2-60. Pulse Parameters
Parameter Description Units Default
V1 Initial Value Volt –
V2 Pulse Value Volt 0.0
TD Delay Time s 0.0
TR Rise Time s Δ𝑡
TF Fall Time s Δ𝑡
PW Pulse Width s 𝑇𝐹
PER Period s 𝑇𝐹

Sine

SIN(V0 VA FREQ TD THETA PHASE)

Table 2-61. Sine Parameters

Parameter Description Units Default
V0 Offset Volt –
VA Amplitude Volt –
FREQ Frequency s−1 –
TD Delay s Δ𝑡
THETA Attenuation Factor s Δ𝑡
PHASE Phase degrees 0.0

The waveform is shaped according to the following equations, where 𝜙 = 𝜋 ∗ PHASE/180 :


𝑉0 , 0 < 𝑡 < 𝑇𝐷
𝑉=
𝑉0 + 𝑉 𝐴 sin[2𝜋 · FREQ · (𝑡 − 𝑇𝐷 ) + 𝜙] exp[−(𝑡 − 𝑇𝐷 ) · THETA], 𝑇𝐷 < 𝑡 < 𝑇𝐹

Exponent

EXP(V1 V2 TD1 TAU1 TD2 TAU2)

Table 2-62. Exponent Parameters

Parameter Description Units Default
V1 Initial Amplitude Volt –
V2 Amplitude Volt –
TD1 Rise Delay Time s 0.0
TAU1 Rise Time Constant s Δ𝑡

226
 Table 2-62. Exponent Parameters
Parameter Description Units Default
TD2 Delay Fall Time s TD1 +Δ𝑡
TAU2 Fall Time Constant s Δ𝑡

The waveform is shaped according to the following equations:



 𝑉1 , 0 < 𝑡 < TD1
 𝑉1 + (𝑉2 − 𝑉1 ){1 − exp[−(𝑡 − TD1)/TAU1]}, TD1 < 𝑡 < TD2


𝑉=
 𝑉1 + (𝑉2 − 𝑉1 ){1 − exp[−(𝑡 − TD1)/TAU1]}

+(𝑉1 − 𝑉2 ){1 − exp[−(𝑡 − TD2)/TAU2]}, TD2 < 𝑡 < 𝑇2




Pattern

PAT(VHI VLO TD TR TF TSAMPLE DATA R)

Table 2-63. Pattern Parameters

Parameter Description Units Default
VHI High Value Volt –
VLO Low Value Volt –
TD Delay Time s –
TR Rise Time s –
TF Fall Time s –
TSAMPLE Bit period s –
DATA Bit pattern – –
R Repeat – 0

The VHI, VLO, TD, TF, TF TSAMPLE and DATA parameters are all required, and hence have no default values.
Negative values for TD are supported. The R parameter is optional. For its default value of 0, the requested
bit pattern will occur once.
The DATA parameter is the requested bit-pattern. Only the 0’ and ‘1’ states are supported. The ‘M’ and ‘Z’
states are not supported. The DATA field should have a leading ‘b’ (or ‘B’) character (e.g., be specified as
‘b0101’ ).
For times earlier than TD, the waveform value is set by the first bit in DATA. For times after the end of the
(possibly repeated) pattern, the waveform value is set by the last bit in DATA. Piecewise linear interpolation
is used to generate the output value when transitioning between states.
The relationship between the various source parameters can be illustrated with the following example:
V1 1 0 PAT(5 0 0 1n 1n 5n b010)
That V1 source definition would produce time-voltages pairs at (0 0) (4.5ns 0) (5ns 2.5) (5.5ns 5.0) (9.5ns
5.0) (10ns 2.5) (10.5ns 0). So, the bit period is 5ns and the voltage value at the start/end of each “sample” is

227
equal to 0.5*(VHI + VLO). The first rise is centered around t=5ns, and hence starts at t=4.5ns and ends at
t=5.5 ns.
The VHI, VLO, TD, TF, TF and TSAMPLE parameters are compatible with .STEP. The DATA and R parameters
are not.
The HSPICE parameters RB, ENCODE and RD_INIT, for the pattern source, are not supported.

Piecewise Linear

PWL T0 V0 [Tn Vn]*

PWL FILE "<name>" [TD=<timeDelay>] [R=<repeatTime>]

Table 2-64. Piecewise Linear Parameters

Parameter Description Units Default
T𝑛 Time at Corner s none
V𝑛 Voltage at Corner Volt none
TD Time Delay s 0
R Repeat Time s none

When the FILE option is given, Xyce will read the corner points from the file specified in the <name> field.
This file should be a plain ASCII text file (or a .CSV file) with time/voltage pairs. There should be one pair
per line, and the time and voltage values should be separated by whitespace or commas. As an example, the
file specified (e.g., vpwl.csv) could have these five lines:

0.00, 0.00
2.00, 3.00
3.00, 2.00
4.00, 2.00
4.01, 5.00

The corresponding example instance lines would be:

VPWL1 1 0 PWL 0S 0V 2S 3V 3S 2V 4S 2V 4.01S 5V

VPWL2 2 0 PWL FILE "vpwl.txt"
VPWL3 3 0 PWL file "vpwl.csv"
VPWL4 4 0 PWL FILE vpwl.csv

The double quotes around the file name are optional, as shown above.
It is a best practice to specify all of the time-voltage pairs in the PWL specification. However, for
compatibility with HSPICE and PSpice, if the user-specified list of time/voltage pairs omits the pair at
time=0 as the first pair in the list then Xyce will insert a pair at time=0 with the voltage value at the first
user-specified time value. As an example, this user-specified list:

228
2S 3V 3S 2V 4S 2V 4.01S 5V

would be implemented in Xyce as follows:

0S 3V 2S 3V 3S 2V 4S 2V 4.01S 5V

TD has units of seconds, and specifies the length of time to delay the start of PWL waveform. The default is
to have no delay, and TD is an optional parameter.
The Repeat Time (R) is an optional parameter. If R is omitted then the waveform will not repeat. If R is
included then the waveform will repeat until the end of the simulation. As examples, R=0 means repeat the
PWL waveform from time=0 to the last time (T 𝑁 ) specified in the waveform specification. (This would use
the time points 0s, 2s, 3s, 4s and 4.01s for the example waveform given above.) In general,
R=<repeatTime> means repeat the waveform from time equal to <repeatTime> seconds in the waveform
specification to the last time (T 𝑁 ) specified in the waveform specification. So, the <repeatTime> must be
greater than or equal to 0 and less than the last time point (T 𝑁 ). If the R parameter is used then it must have
a value.
The specification PWL FILE "<name>" R is illegal in Xyce as a shorthand for R=0. Also, the Xyce syntax
for PWL sources is not compatible with the PSpice REPEAT syntax for PWL sources. See section 6.1.12 for
more details.
The repeat time (R) does enable the specification of discontinuous piecewise linear waveforms. For
example, this waveform is a legal Xyce syntax.

VPWL1 1 0 PWL 0S 0V 2S 3V 3S 2V 4S 2V 4.01V 5V R=2

However, in general, discontinuous source waveforms may cause convergence problems.

Frequency Modulated

SFFM (V0 VA FC MDI FS)

Table 2-65. Frequency Modulated Parameters

Parameter Description Units Default
V0 Offset Volt –
VA Amplitude Volt –
FC Carrier Frequency hertz 1/TSTOP
MDI Modulation Index - 0
FS Signal Frequency hertz 1/TSTOP

TSTOP is the final time, as entered into the transient (.TRANS) command. The waveform is shaped
according to the following equation:

𝑉 = 𝑉0 + 𝑉 𝐴 · sin(2𝜋 · FC · TIME + MDI · sin(2𝜋 · FS · TIME))

where TIME is the current simulation time.

229
2.3.11. Port Device

Instance Form P<name> <(+) node> <(-) node> [[DC] <value> ] port=port number
+ [Z0 = value] [AC [magnitude value [phase value] ] ]
+ [transient specification]

Examples P1 1 0 port = 1
P2 12 0 port=1 z0=100
P1 1 0 port=2 sin 0 1 1e5
P2 2 0 port=2 z0=100 AC 1

Parameters and
Options port
The port number. Numbered sequentially beginning with 1
Z0 System impedance. Currently, it only supports a real-valued impedance.
transient specification
There are six predefined time-varying functions for sources:
PULSE <parameters> Pulse waveform
SIN <parameters> Sinusoidal waveform
EXP <parameters> Exponential waveform
PAT <parameters> Pattern waveform
PWL <parameters> Piecewise linear waveform
SFFM <parameters> Frequency-modulated waveform

Comments The port device identifies the ports used in .LIN analysis. Each port requires a unique
port number. For example, if the netlist has N port devices, it must contain the
sequential set of port numbers, from 1 to N. Each port has an associated impedance
Z0. The default is 50 ohms.
The port device behaves as a voltage source in series with an impedance for all other
analyses, such as DC, AC and transient.
None, any, or all of the DC, AC, and transient values can be specified. The AC phase
value is in degrees. The port device accepts the same transient specifications as the
voltage (V) sources.
Positive current flows from the positive node through the port device to the negative
node.
The power supplied or dissipated by the port device is calculated with 𝐼 · Δ𝑉 where the
voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
Dissipated power has a positive sign, while supplied power has a negative sign.

230
2.3.12. Voltage Controlled Voltage Source

Symbol

Instance Form E<name> <(+) node> <(-) node> <(+) controlling node>
+ <(-) controlling node> <gain>
E<name> <(+) node> <(-) node> VALUE = { <expression> }
+ [device parameters]
E<name> <(+) node> <(-) node> TABLE { <expression> } =
+ < <input value>,<output value> >*
E<name> <(+) node> <(-) node> POLY(<value>)
+ [<+ control node> <- control node>]*
+ [<polynomial coefficient value>]*

Examples EBUFFER 1 2 10 11 5.0

ESQROOT 5 0 VALUE = {5V*SQRT(V(3,2))}
ET2 2 0 TABLE {V(ANODE,CATHODE)} = (0,0) (30,1)
EP1 5 1 POLY(2) 3 0 4 0 0 .5 .5

Parameters and
Options (+) node
(-) node
Output nodes. Positive current flows from the (+) node through the source
to the (-) node.
(+) controlling node
(-) controlling node
Node pairs that define a set of controlling voltages. A given node may appear
multiple times and the output and controlling nodes may be the same.
device parameters
The second form supports two instance parameters smoothbsrc and
rcconst. Parameters may be provided as space separated
<parameter>=<value> specifications as needed. The default value for
smoothbsrc is 0 and the default for rcconst is 1e-9.

Comments In the first form, a specified voltage drop between controlling nodes is multiplied by
the gain to determine the voltage drop across the output nodes.
The second through fourth forms allow nonlinear controlled sources using the VALUE,
TABLE, or POLY keywords, respectively, and are used in analog behavioral modeling.
They are provided primarily for netlist compatibility with other simulators. These
three forms are automatically converted within Xyce to its principal ABM device, the
B nonlinear dependent source device. See the B-source section (2.3.16) and the Xyce
User’s Guide for more guidance on analog behavioral modeling. For details
concerning the use of the POLY format, see section 2.2.5.

231
For HSPICE compatibility, VOL is an allowed synonym for VALUE for the E-source.
The power supplied or dissipated by this source device is calculated with 𝐼 · Δ𝑉 where
the voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
Dissipated power has a positive sign, while supplied power has a negative sign.
NOTE: The expression given on the left hand side of the equals sign in E source
TABLE expressions may be enclosed in braces, but is not required to be. Further, if
braces are present there must be exactly one pair of braces and it must enclose the
entire expression. It is not legal to use additional pairs of braces as parentheses inside
these expressions. So
ET2 2 0 TABLE {V(ANODE,CATHODE)+5} = (0,0) (30,1)
ET3 2 0 TABLE V(ANODE,CATHODE)+5 = (0,0) (30,1)
are legal, but
ET2 2 0 TABLE {V(ANODE,CATHODE)+{5}} = (0,0) (30,1)
is not. This last will result in a parsing error about missing braces.
E-sources were originally developed primarily to support DC and transient analysis.
As such, their support for frequency domain analysis (AC and HB) has some
limitations. The main limitation to be aware of is that time-dependent sources will not
work with AC or HB analysis. These are sources in which the variable TIME is used in
the VALUE= expression. However, this time-dependent usage is not common. The most
common use case is one in which the E-source is purely dependent (depends only on
other solution variables), and this use case will work with AC and HB.

232
2.3.13. Current Controlled Current Source

Symbol

Instance Form F<name> <(+) node> <(-) node>

+ <controlling V device name> <gain>
F<name> <(+) node> <(-) node> POLY(<value>)
+ <controlling V device name>*
+ < <polynomial coefficient value> >*

Examples FSENSE 1 2 VSENSE 10.0

FAMP 13 0 POLY(1) VIN 0 500
FNONLIN 100 101 POLY(2) VCNTRL1 VCINTRL2 0.0 13.6 0.2 0.005

Parameters and
Options (+) node
(-) node
Output nodes. Positive current flows from the (+) node through the source
to the (-) node.
controlling V device
The controlling voltage source which must be an independent voltage source
(V device).

Comments In the first form, a specified current through a controlling device is multiplied by the
gain to determine this device’s output current. The gain may be expressed either as a
number, a parameter, or an arbitrary brace-delimited ABM expression.
The second form using the POLY keyword is used in analog behavioral modeling.
Both forms are automatically converted within Xyce to its principal ABM device, the
B nonlinear dependent source device. See the B-source section (2.3.16) and the Xyce
User’s Guide for more guidance on analog behavioral modeling. For details
concerning the use of the POLY format, see section 2.2.5.
The power supplied or dissipated by this source device is calculated with 𝐼 · Δ𝑉 where
the voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
Dissipated power has a positive sign, while supplied power has a negative sign.
F-sources were originally developed primarily to support DC and transient analysis.
As such, their support for frequency domain analysis (AC and HB) has some
limitations. The main limitation to be aware of is that time-dependent sources will not
work with AC or HB analysis. These are sources in which the variable TIME is used in
the VALUE= expression. However, this time-dependent usage is not common. The most
common use case is one in which the F-source is purely dependent (depends only on
other solution variables), and this use case will work with AC and HB.

233
2.3.14. Voltage Controlled Current Source

Symbol

Instance Form G<name> <(+) node> <(-) node> <(+) controlling node>
+ <(-) controlling node> <transconductance> [M=<value>]
G<name> <(+) <node> <(-) node> VALUE = { <expression> }
G<name> <(+) <node> <(-) node> TABLE { <expression> } =
+ < <input value>,<output value> >*
G<name> <(+) <node> <(-) node> POLY(<value>)
+ [<+ controlling node> <- controlling node>]*
+ [<polynomial coefficient>]*

Examples GBUFFER 1 2 10 11 5.0

GPSK 11 6 VALUE = {5MA*SIN(6.28*10kHz*TIME+V(3))}
GA2 2 0 TABLE {V(5)} = (0,0) (1,5) (10,5) (11,0)
GMULT 1 2 10 11 3.0 M=5

Parameters and
Options (+) node
(-) node
Output nodes. Positive current flows from the (+) node through the source
to the (-) node.
(+) controlling node
(-) controlling node
Node pairs that define a set of controlling voltages. A given node may appear
multiple times and the output and controlling nodes may be the same.

Comments In the first form, the voltage drop between the controlling nodes is multiplied by the
transconductance to obtain the current-source output of the G device.
The second through fourth forms using the VALUE, TABLE, and POLY keywords,
respectively, are used in analog behavioral modeling. They are provided primarily for
netlist compatibility with other simulators. These two forms are automatically
converted within Xyce to its principal ABM device, the B nonlinear dependent source
device. See the B-source section (2.3.16) and the Xyce User’s Guide for more
guidance on analog behavioral modeling. For details concerning the use of the POLY
format, see section 2.2.5.
For HSPICE compatibility, CUR is an allowed synonym for VALUE for the G-source.
Also, this device supports the M multiplier parameter.
The power supplied or dissipated by this source device is calculated with 𝐼 · Δ𝑉 where
the voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
Dissipated power has a positive sign, while supplied power has a negative sign.

234
G-sources were originally developed primarily to support DC and transient analysis.
As such, their support for frequency domain analysis (AC and HB) has some
limitations. The main limitation to be aware of is that time-dependent sources will not
work with AC or HB analysis. These are sources in which the variable TIME is used in
the VALUE= expression. However, this time-dependent usage is not common. The most
common use case is one in which the G-source is purely dependent (depends only on
other solution variables), and this use case will work with AC and HB.

235
2.3.15. Current Controlled Voltage Source

The syntax of this device is exactly the same as for a Current-Controlled Current Source. For a
Current-Controlled Voltage Source just substitute an H for the F. The H device generates a voltage, whereas
the F device generates a current.

Symbol

Instance Form H<name> <(+) node> <(-) node>

+ <controlling V device name> <transresistance>
H<name> <(+) node> <(-) node> POLY(<value>)
+ <controlling V device name>*
+ < <polynomial coefficient value> >*

Examples HSENSE 1 2 VSENSE 10.0

HAMP 13 0 POLY(1) VIN 0 500
HNONLIN 100 101 POLY(2) VCNTRL1 VCINTRL2 0.0 13.6 0.2 0.005

Comments In the first form, the current through a specified controlling voltage source is
multiplied by the transresistance to obtain the voltage-source output. The
transresistance may be expressed either as a number, a parameter, or an arbitrary
brace-delimited ABM expression.
The second form using the POLY keyword is used in analog behavioral modeling. It is
provided primarily for netlist compatibility with other simulators.
H sources in any form are automatically converted within Xyce to its principal ABM
device, the B nonlinear dependent source device. See the B-source section (2.3.16)
and the Xyce User’s Guide for more guidance on analog behavioral modeling. For
details concerning the use of the POLY format, see section 2.2.5.
The power supplied or dissipated by this source device is calculated with 𝐼 · Δ𝑉 where
the voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .
Dissipated power has a positive sign, while supplied power has a negative sign.
H-sources were originally developed primarily to support DC and transient analysis.
As such, their support for frequency domain analysis (AC and HB) has some
limitations. The main limitation to be aware of is that time-dependent sources will not
work with AC or HB analysis. These are sources in which the variable TIME is used in
the VALUE= expression. However, this time-dependent usage is not common. The most
common use case is one in which the H-source is purely dependent (depends only on
other solution variables), and this use case will work with AC and HB.

236
2.3.16. Nonlinear Dependent Source

Instance Form B<name> <(+) node> <(-) node> V=ABM expression [device parameters]
B<name> <(+) node> <(-) node> I=ABM expression

Examples B1 2 0 V={sqrt(V(1))}
B2 4 0 V={V(1)*TIME}
B3 4 2 I={I(V1) + V(4,2)/100}
B4 5 0 V={Table {V(5)}=(0,0) (1.0,2.0) (2.0,3.0) (3.0,10.0)}
B5 6 0 V=tablefile("file.dat")
B6 7 0 I=tablefile("file.dat")
B5 6 0 V=table("file.dat")
B6 7 0 I=table("file.dat")
B5 6 0 V={table("file.dat")}
B5 6 0 V={spline("file.dat")}
B5 6 0 V={BLI("file.dat")}
B5 6 0 V={fasttable("file.dat")}

Comments The nonlinear dependent source device, also known as the B-source device, is used in
analog behavioral modeling (ABM). The (+) and (-) nodes are the output nodes.
Positive current flows from the (+) node through the source to the (-) node.
The power supplied or dissipated by the nonlinear dependent source is calculated with
𝐼 · Δ𝑉 where the voltage drop is calculated as (𝑉+ − 𝑉− ) and positive current flows
from 𝑉+ to 𝑉− . Dissipated power has a positive sign, while supplied power has a
negative sign.
The syntax involving the tablefile keyword internally attempts to load the data in
"file.dat" into a TABLE expression. The data file must be in plain-text and contain
just two pairs of data per line. For an example see the “Analog Behavioral Modeling”
chapter of the Xyce User’s Guide. Either table or tablefile can be used to read a
table in from a file. They are synonyms.
Other related table-based features include fasttable, which is the same as table
but without many breakpoints, and bli for Barycentric Lagrange Interpolation [10].
Various splines are also supported, including spline, cubic, akima [8] and
wodicka [9]. spline and akima are synonymous. All of these methods use the same
syntax as table, and all of them support reading tables in from files.
It is important to note that the B-source allows the user to specify expressions that
could have infinite-slope transitions, such as the following. (Note: the braces
surrounding all expressions are required in this definition.)
Bcrtl OUTA 0 V={ IF( (V(IN) > 3.5), 5, 0 ) }
This can lead to “timestep too small” errors when Xyce reaches the transition point.
Infinite-slope transitions in expressions dependent only on the time variable are a
special case, because Xyce can detect that they are going to happen in the future and

237
set a “breakpoint” to capture them. Infinite-slope transitions depending on other
solution variables cannot be predicted in advance, and cause the time integrator to
scale back the timestep repeatedly in an attempt to capture the feature until the
timestep is too small to continue.
One solution to the problem is to modify the expression to allow a continuous
transition. However, this can become complicated with multiple inputs. The other
solution is to specify device options or instance parameters to allow smooth
transitions. The parameter smoothbsrc enables the smooth transitions. This is done
by adding a RC network to the output of B sources. For example,

Bcrtl OUTA 0 V={ IF( (V(IN) > 3.5), 5, 0 ) } smoothbsrc=1

.options device smoothbsrc=1

The smoothness of the transition can be controlled by specifying the rc constant of the
RC network. For example,

Bcrtl OUTA 0 V={ IF( (V(IN) > 3.5), 5, 0 ) } smoothbsrc=1

+ rcconst = 1e-10

Note that this smoothed B-source only applies to voltage sources. The voltage
behavioral source supports two instance parameters smoothbsrc and rcconst.
Parameters may be provided as space separated <parameter>=<value>
specifications as needed. The default value for smoothbsrc is 0 and the default for
rcconst is 1e-9.
See the “Analog Behavioral Modeling” chapter of the Xyce User’s Guide [1] for
guidance on using the B-source device and ABM expressions, and the Expressions
Section (2.2) for complete documentation of expressions and expression operators.
One important note is that time-dependent expressions are supported for the current
and voltage parameters of a B source, but frequency-dependent expressions are not.
B-sources were originally developed primarily to support DC and transient analysis.
As such, their support for frequency domain analysis (AC and HB) has some
limitations. The main limitation to be aware of is that time-dependent sources will not
work with AC or HB analysis. These are sources in which the variable TIME is used in
the VALUE= expression. The use case of a purely depedent B-source (depends only on
other solution variables) will work with AC and HB.

238
2.3.17. Bipolar Junction Transistor (BJT)

Symbol

Instance Form Q<name> <collector node> <base node> <emitter node>

+ [substrate node] <model name> [area value]

Q<name> <collector node> <base node> <emitter node>

+ [thermal node] <VBIC 1.3 3-terminal model name>

Q<name> <collector node> <base node> <emitter node>

+ <substrate> [thermal node] <VBIC 1.3 4-terminal model name>

Q<name> <collector node> <base node> <emitter node>

+ <substrate> <thermal node> <HICUM model name>

Model Form .MODEL <model name> NPN [model parameters]

.MODEL <model name> PNP [model parameters]

Examples Q2 10 2 9 PNP1
Q12 14 2 0 1 NPN2 2.0
Q6 VC 4 11 [SUB] LAXPNP
Q7 Coll Base Emit DT VBIC13MODEL2
Q8 Coll Base Emit VBIC13MODEL3 SW_ET=0
Q9 Coll Base Emit Subst DT VBIC13MODEL4
Q10 Coll Base Emit Subst DT HICUMMMODEL1

Parameters and
Options substrate node
Optional and defaults to ground. Since Xyce permits alphanumeric node
names and because there is no easy way to make a distinction between these
and the model names, the name (not a number) used for the substrate node
must be enclosed in square brackets [ ]. Otherwise, nodes would be
interpreted as model names. See the fourth example above.
area value
The relative device area with a default value of 1.

Comments The BJT is modeled as an intrinsic transistor using ohmic resistances in series with the
collector (RC/area), with the base (value varies with current, see BJT equations) and
with the emitter (RE/area).For model parameters with optional names, such as VAF
and VA (the optional name is in parentheses), either may be used.For model types
NPN and PNP, the isolation junction capacitance is connected between the

239
 intrinsic-collector and substrate nodes. This is the same as in SPICE and works well
for vertical IC transistor structures.

Only the VBIC 1.3 model is available in Xyce 6.11 and later. The VBIC 1.3 model
is provided in both 3-terminal (Q level 11) and 4-terminal (Q level 12) variants, both
supporting electrothermal and excess-phase effects. These variants of the Q line are
shown in the fourth through sixth examples above. VBIC 1.3 instance lines have three
or four required nodes, depending on model level, and an optional “dt” node. The first
three are the normal collector, base,and emitter. In the level 12 (4-terminal) the fourth
node is the substrate, just as for the level 1 BJT. If the optional “dt” node is specified
for either variant, it can be used to print the local temperature rise due to self-heating,
and could possibly be used to model coupled heating effects of several VBIC devices.
It is, however, unnecessary to specify a “dt” node just to print the local temperature
rise, because when this node is omitted from the instance line it simply becomes and
internal node, and may still be printed using the syntax N(instancename:dt). For
the “Q8” example above, one could print N(Q8:dt).

As of release 6.10 of Xyce, the VBIC 1.3 3-terminal device (Q level 11) has been the
subject of extensive optimization, and runs much faster than in previous releases.

The HICUM models require both a substrate and thermal node.

Figure 2-3. BJT model schematic. Adapted from reference [2].

BJT Level selection Xyce supports the level 1 BJT model, which is based on the documented standard
SPICE 3F5 BJT model, but was coded independently at Sandia. It is mostly based on the classic
Gummel-Poon BJT model [12].

Two variants of the VBIC model are provided as BJT levels 11 and 12. Levels 11 and 12 are the 3-terminal
and 4-terminal variants of the VBIC 1.3.

An experimental release of the FBH HBT_X model version 2.1[13] is provided as BJT level 23.

Both the HICUM/L0 (level 230) and HICUM/L2 (level 234) models are also provided
(https://www.iee.et.tu-dresden.de/iee/eb/hic_new/hic_start.html).

240
The MEXTRAM[14] BJT model version 504.12.1 model is provided. Two variants of this model are
available: the level 504 model without self-heating and without external substrate node, and the level 505
model with self heating but without external substrate node. The level 505 instance line requires a fourth
node for the ’dt’ node, similar to the usage in all of the VBIC models (levels 11-12), but is otherwise
identical to the level 504 model.

BJT Power Calculations Power dissipated in the transistor is calculated with |𝐼 𝐵 ∗ 𝑉𝐵𝐸 | + |𝐼𝐶 ∗ 𝑉𝐶 𝐸 |,
where 𝐼 𝐵 is the base current, 𝐼𝐶 is the collector current, 𝑉𝐵𝐸 is the voltage drop between the base and the
emitter and 𝑉𝐶𝐸 is the voltage drop between the collector and the emitter. This formula may differ from
other simulators.

2.3.17.1. The Level 1 Model

BJT Equations The Level 1 BJT implementation within Xyce is based on [15]. The equations in this
section describe an NPN transistor. For the PNP device, reverse the signs of all voltages and currents. The
equations use the following variables:

𝑉𝑏𝑒 = intrinsic base-intrinsic emitter voltage

𝑉𝑏𝑐 = intrinsic base-intrinsic collector voltage
𝑉𝑏𝑠 = intrinsic base-substrate voltage
𝑉𝑏𝑤 = intrinsic base-extrinsic collector voltage (quasi-saturation only)
𝑉𝑏𝑥 = extrinsic base-intrinsic collector voltage
𝑉𝑐𝑒 = intrinsic collector-intrinsic emitter voltage
𝑉 𝑗𝑠 = (NPN) intrinsic collector-substrate voltage
= (PNP) intrinsic substrate-collector voltage
𝑉𝑡 = 𝑘𝑇/𝑞 (thermal voltage)
𝑉𝑡 ℎ = threshold voltage
𝑘 = Boltzmann’s constant
𝑞 = electron charge
𝑇 = analysis temperature (K)
𝑇0 = nominal temperature (set using TNOM option)

Other variables are listed above in BJT Model Parameters.

DC Current The BJT model is based on the Gummel and Poon model [16] where the different terminal
currents are written
𝑑𝑉𝑏𝑒
𝐼𝑒 = −𝐼𝑐𝑐 − 𝐼𝑏𝑒 + 𝐼𝑟 𝑒 + (𝐶𝑑𝑖 𝑓 𝑒 + 𝐶𝑑𝑒 )
𝑑𝑡
𝑑𝑉𝑏𝑐
𝐼𝑐 = −𝐼𝑐𝑐 + 𝐼𝑏𝑐 − 𝐼𝑟 𝑐 − (𝐶𝑑𝑖 𝑓 𝑐 + 𝐶𝑑𝑐 )
𝑑𝑡
𝐼𝑏 = 𝐼𝑒 − 𝐼 𝑐

Here, 𝐶𝑑𝑖 𝑓 𝑒 and 𝐶𝑑𝑖 𝑓 𝑐 are the capacitances related to the hole charges per unit area in the base, 𝑄 𝑑𝑖 𝑓 𝑒 and
𝑄 𝑑𝑖 𝑓 𝑐 , affiliated with the electrons introduced across the emitter-base and collector-base junctions,

241
respectively. Also, 𝐶𝑏𝑒 and 𝐶𝑏𝑐 are the capacitances related to donations to the hole charge of the base, 𝑄 𝑏𝑒
and 𝑄 𝑏𝑐 , affiliated with the differences in the depletion regions of the emitter-base and collector-base
junctions, respectively. The intermediate currents used are defined as
   
IS 𝑉𝑏𝑒
−𝐼𝑏𝑒 = exp −1
BF NF𝑉𝑡 ℎ
    
𝑄 𝑏𝑜 𝑉𝑏𝑒 𝑉𝑏𝑐
−𝐼𝑐𝑐 = IS exp − exp
𝑄𝑏 NF𝑉𝑡 ℎ NF𝑉𝑡 ℎ
   
IS 𝑉𝑏𝑐
−𝐼𝑏𝑐 = exp −1
BR NR𝑉𝑡 ℎ
   
𝑉𝑏𝑒
𝐼𝑟 𝑒 = ISE exp −1
NE𝑉𝑡 ℎ
   
𝑉𝑏𝑐
𝐼𝑟 𝑐 = ISC exp −1
NC𝑉𝑡 ℎ

where the last two terms are the generation/recombination currents related to the emitter and collector
junctions, respectively. The charge 𝑄 𝑏 is the majority carrier charge in the base at large injection levels and
is a key difference in the Gummel-Poon model over the earlier Ebers-Moll model. The ratio 𝑄 𝑏 /𝑄 𝑏𝑜
(where 𝑄 𝑏𝑜 represents the zero-bias base charge, i.e. the value of 𝑄 𝑏 when 𝑉𝑏𝑒 = 𝑉𝑏𝑐 = 0) as computed by
Xyce is given by
𝑄𝑏 𝑞1  √︁ 
= 1 + 1 + 4𝑞 2
𝑄 𝑏𝑜 2
where
  −1
𝑉𝑏𝑒 𝑉𝑏𝑐
𝑞1 = 1− −
VAR VAF
       
IS 𝑉𝑏𝑒 IS 𝑉𝑏𝑐
𝑞2 = exp −1 + exp −1
IKF NF𝑉𝑡 ℎ IKR NR𝑉𝑡 ℎ

Capacitance Terms The capacitances listed in the above DC 𝐼 − 𝑉 equations each consist of a depletion
layer capacitance 𝐶𝑑 and a diffusion capacitance 𝐶𝑑𝑖 𝑓 . The first is given by

  −M
 CJ 1 − 𝑉VJ


 𝑑𝑖
𝑉𝑑𝑖 ≤ FC · VJ
𝐶𝑑 = h i
 CJ (1 − FC) − (1+M) 1 − FC(1 + M) + M 𝑉VJ
 𝑑𝑖
𝑉𝑑𝑖 > FC · VJ


where CJ = CJE for 𝐶𝑑𝑒 , and where CJ = CJC for 𝐶𝑑𝑐 . The diffusion capacitance (sometimes referred to
as the transit time capacitance) is
𝑑𝐼
𝐶𝑑𝑖 𝑓 = TT𝐺 𝑑 = TT
𝑑𝑉𝑑𝑖
where 𝐼 is the diode DC current given, 𝐺 𝑑 is the corresponding junction conductance, and where TT = TF
for 𝐶𝑑𝑖 𝑓 𝑒 and TT = TR for 𝐶𝑑𝑖 𝑓 𝑐 .

242
 Figure 2-4. VBIC thermal network schematic.

Temperature Effects SPICE temperature effects are default, but all levels of the BJT have a more
advanced temperature compensation available. By specifying TEMPMODEL=QUADRATIC in the netlist,
parameters can be interpolated quadratically between measured values extracted from data. In the BJT, IS
and ISE are interpolated logarithmically because they can change over an order of magnitude or more for
temperature ranges of interest. See the Section ?? for more details on how to include quadratic temperature
effects.

For further information on BJT models, see [16]. For a thorough description of the U.C. Berkeley SPICE
models see Reference [17].

2.3.17.2. VBIC Temperature Considerations

The VBIC (Q levels 11 and 12) model both support a self-heating model. The model works by computing
the power dissipated by all branches of the device, applying this power as a flow through a small thermal
network consisting of a power flow (“current”) source through a thermal resistance and thermal
capacitance, as shown in Figure 2-4. The circuit node DT will therefore be the “thermal potential”
(temperature) across the parallel thermal resistance and capacitance. This temperature is the temperature
rise due to self heating of the device, which is added to the ambient temperature and TRISE parameter to
obtain the device operating temperature.

In VBIC 1.3, the dt node is optional on the netlist line. If not given, the dt node is used internally for thermal
effects calculations, but not accessible from the rest of the netlist. The VBIC 1.3 provides an instance
parameter SW_ET that may be set to zero to turn off electrothermal self-heating effects. When set to zero, no
thermal power is sourced into the dt node. This parameter defaults to 1, meaning that thermal power is
computed and flows into dt even when dt is unspecified on the netlist and remains an internal node.

In VBIC 1.3, setting RTH to zero does NOT disable the self-heating model, and does not short the dt node
to ground, even though one might expect that to be the behavior. Rather, it simply removes the RTH resistor
from the equivalent circuit of figure 2-4 and leaves the dt node floating. This is an important point to
recognize when using the VBIC.

243
If a node name is given as the fourth node of a VBIC Xyce will emit warnings about the node not having a
DC path to ground and being connected to only one device. These warnings may safely be ignored, and are
a harmless artifact of Xyce’s connectivity checker. It is possible to silence this warning by adding a very
large resistance between the dt node and ground — 1GOhm or 1TOhm are effectively the same as leaving
the node floating, and will satisfy the connectivity checker’s tests. This used to be the recommended means
of silencing the connectivity checker for the VBIC 1.2 where dt was a required node, but it is safe if and
only if a nonzero RTH value is specified for the device. If, however, RTH is zero, then dt would otherwise be
floating and your external resistance now becomes the primary path for thermal power flow; rather than
turning off self-heating effects, it will be as if you had set RTH to a very large value. We therefore
recommend that you not tie the dt node to ground via a resistor, and if you are not using it to connect VBIC
devices together via a thermal network, simply leave off the dt node to silence the connectivity checker
warning. Turn off self-heating effects ONLY by setting the SW_ET instance parameter to zero.

Users of earlier versions of Xyce may have been using the VBIC 1.2 model that was removed in release
6.11. All netlists containing the old level=10 VBIC 1.2 model must be modified to run in Xyce 6.11 and
later. The following points should be observed when converting an old VBIC 1.2 netlist and model card to
VBIC 1.3.

• Generally speaking, most VBIC 1.2 model cards can be converted to VBIC 1.3 model cards by the
simple substitution of level=11 for level=10, with the following provisos.

• VBIC 1.2 in Xyce 6.10 and earlier did not support excess phase effects, and so the TD parameter
governing excess phase was ignored.

The Xyce team has observed that some users’ VBIC 1.2 parameter extractions have a non-zero value
for the TD parameter. The impact of this is twofold:

– Circuits that use such model cards with only the level number changed will likely not produce
identical results when compared to simulation results of older versions of Xyce using VBIC 1.2
due to the excess phase effects. If strict comparison between VBIC 1.3 runs with Xyce 6.11 or
later against older runs with VBIC 1.2 is desired, change the TD parameter to zero. This will
disable the excess phase effects and make VBIC 1.3 equivalent to the VBIC 1.2 that was
previously provided.

– The Xyce team has seen some instances where the previously ignored TD parameter value is
such that Xyce will fail to converge when the equivalent VBIC 1.3 model is substituted. The
VBIC 1.2 behavior can be recovered by setting the model parameter TD to zero, which will
disable the excess phase effect in VBIC 1.3. We can only suggest that the model card be
re-extracted using VBIC 1.3 to determine the correct value for TD.

• VBIC 1.2 had a model parameter called DTEMP, which Xyce also recognized on the instance line. In
VBIC 1.3 this parameter has been replaced by another called TRISE, which is only an instance
parameter, and is unrecognized in model cards. VBIC 1.3 also recognizes DTEMP on the instance line
as an alias for TRISE. If you had been specifying DTEMP in your VBIC 1.2 model cards, you will need
to move it to the instance line instead in order for the parameter to be properly recognized by both
VBIC 1.2 and VBIC 1.3.

• Turning off self-heating effects in VBIC 1.2 was done by grounding the mandatory dt node. This is
not the recommended way of disabling self-heating in VBIC 1.3. To disable self-heating, set the
SW_ET parameter to zero on the instance line (as is done in the “Q8” example above).

244
 • If not using the dt node as a way of thermally coupling devices to each other, leave it off of VBIC 1.3
instance lines, allowing it to be an internal variable irrespective of whether self-heating is enabled or
not. This will silence any connectivity warnings from Xyce. Since the dt node may be printed using
the N() syntax even when internal, it is unnecessary to put a dt node on the instance line just to print
the local temperature rise due to self-heating. The only reasons to include it on the instance line
would be for backward compatibility to VBIC 1.2 netlists, or to implement a thermal coupling
network between devices.
• Finally, VBIC 1.3 introduced a number of constraints on model parameters that the previous version
did not. Xyce will emit warnings if any parameter on a VBIC 1.3 model card is out of the range
specified by the VBIC 1.3 authors. These warnings should not be ignored lightly, as they indicate that
the model is being used in a manner not intended by its authors. They are generally a sign that the
model may not be well-behaved, and may indicate an improperly extracted model card.

2.3.17.3. Level 1 BJT Tables

Table 2-66. Bipolar Junction Transistor Device Instance Parameters

Parameter Description Units Default
AREA Relative device area – 1
IC1 Vector of initial values: Vbe,Vce. Vbe=IC1 V 0
IC2 Vector of initial values: Vbe,Vce. Vce=IC2 V 0
M multiplicity factor – 1
logical
OFF Initial condition of no voltage drops accross device false
(T/F)
◦C Ambient
TEMP Device temperature
Temperature

Table 2-67. Bipolar Junction Transistor Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent — 1
BF Ideal maximum foward beta – 100
BFM Ideal maximum foward beta – 100
BR Ideal maximum reverse beta – 1
BRM Ideal maximum reverse beta – 1
BV Reverse early voltage V 0
C2 Coefficient for base-emitter leak current. — 0
C4 Coefficient for base-collector leak current. — 0
CCS Substrate zero-bias p-n capacitance F 0
CDIS Fraction of CJC connected internally to RB – 1
CJC Base-collector zero-bias p-n capacitance F 0

245
 Table 2-67. Bipolar Junction Transistor Device Model Parameters
Parameter Description Units Default
CJE Base-emitter zero-bias p-n capacitance F 0
CJS Substrate zero-bias p-n capacitance F 0
CSUB Substrate zero-bias p-n capacitance F 0
EG Bandgap voltage (barrier highth) eV 1.11
ESUB Substrate p-n grading factor – 0
FC Foward-bias depletion capacitor coefficient – 0.5
IK Corner for foward-beta high-current roll-off A 0
IKF Corner for foward-beta high-current roll-off A 0
IKR Corner for reverse-beta high-current roll-off A 0
IOB Current at which RB falls off by half A 0
IRB Current at which RB falls off by half A 0
IS Transport saturation current A 1e-16
ISC Base-collector leakage saturation current A 0
ISE Base-emitter leakage saturation current A 0
ITF Transit time dependancy on IC — 0
JBF Corner for foward-beta high-current roll-off A 0
JBR Corner for reverse-beta high-current roll-off A 0
JLC Base-collector leakage saturation current A 0
JLE Base-emitter leakage saturation current A 0
JRB Current at which RB falls off by half A 0
JTF Transit time dependancy on IC — 0
KF Flicker noise coefficient – 0
MC Base-collector p-n grading factor – 0.33
ME Base-emitter p-n grading factor – 0.33
MJC Base-collector p-n grading factor – 0.33
MJE Base-emitter p-n grading factor – 0.33
MJS Substrate p-n grading factor – 0
MS Substrate p-n grading factor – 0
NC Base-collector leakage emission coefficient – 2
NE Base-emitter leakage emission coefficient – 1.5
NF Foward current emission coefficient – 1
NK High current rolloff coefficient – 0.5
NKF High current rolloff coefficient – 0.5
NLE Base-emitter leakage emission coefficient – 1.5
NR Reverse current emission coefficient – 1
PC Base-collector built-in potential V 0.75

246
 Table 2-67. Bipolar Junction Transistor Device Model Parameters
Parameter Description Units Default
PE Base-emitter built-in potential V 0.75
PS Substrate built-in potential V 0.75
PSUB Substrate built-in potential V 0.75
PT Temperature exponent for IS. (synonymous with XTI) — 3
PTF Excess Phase at 1/(2pi*TF) Hz degree 0
RB Zero-bias (maximum) base resistance ˙ 0
RBM Maximum base resistance ˙ 0
RC Collector ohmic resistance ˙ 0
RE Emitter ohmic resistance ˙ 0
TB Foward and reverse beta temperature coefficient – 0
TCB Foward and reverse beta temperature coefficient – 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
TF Ideal foward transit time s 0
◦C Ambient
TNOM Parameter measurement temperature
Temperature
TR Ideal reverse transit time s 0
VA Foward early voltage V 0
VAF Foward early voltage V 0
VAR Reverse early voltage V 0
VB Reverse early voltage V 0
VBF Foward early voltage V 0
VJC Base-collector built-in potential V 0.75
VJE Base-emitter built-in potential V 0.75
VJS Substrate built-in potential V 0.75
VRB Reverse early voltage V 0
VTF Transit time dependancy on Vbc V 0
XCJC Fraction of CJC connected internally to RB – 1
XTB Foward and reverse beta temperature coefficient – 0
XTF Transit time bias dependence coefficient – 0
XTI Temperature exponent for IS. (synonymous with PT) — 3

2.3.17.4. Level 11 and 12 BJT Tables (VBIC 1.3)

The VBIC 1.3 (level 11 transistor for 3-terminal, level 12 for 4-terminal) supports a number of instance
parameters that are not available in the VBIC 1.2. The level 11 and level 12 differ only by the number of
required nodes. The level 11 is the 3-terminal device, having only collector, base, and emitter as required

247
nodes. The level 12 is the 4-terminal device, requiring collector, base, emitter and substrate nodes. Both
models support an optional ’dt’ node as their last node on the instance line.
Model cards extracted for the VBIC 1.2 will mostly work with the VBIC 1.3, with one notable
exception: in VBIC 1.2 the DTEMP parameter was a model parameter, and Xyce allowed it also to be
specified on the instance line, overriding whatever was specified in the model. This parameter was replaced
in VBIC 1.3 with the TRISE parameter, which is only an instance parameter. DTEMP and DTA are both
supported as aliases for the TRISE instance parameter.

Table 2-68. VBIC 1.3 3T Device Instance Parameters

Parameter Description Units Default
DTA Alias for trise ◦C 0
DTEMP Alias for trise ◦C 0
M multiplicity factor — 1
OFF Set to 1 to initialize device to OFF instead of normally — 0
SW_ET switch for self-heating: 0=no and 1=yes — 1
SW_NOISE switch for including noise: 0=no and 1=yes — 1
local temperature delta to ambient (before ◦C
TRISE 0
self-heating)

Table 2-69. VBIC 1.3 3T Device Model Parameters

Parameter Description Units Default
ABK SiGe base current kink exponent — 1
AFN b-e flicker noise current exponent — 1
AJC b-c capacitance smoothing factor — -0.5
AJE b-e capacitance smoothing factor — -0.5
AJS c-s capacitance smoothing factor — -0.5
ART smoothing parameter for reach-through — 0.1
AVC1 b-c weak avalanche parameter 1 V−1 0
AVC2 b-c weak avalanche parameter 2 — 0
AVCX1 bx-cx weak avalanche parameter 1 V−1 0
AVCX2 bx-cx weak avalanche parameter 2 — 0
BBK SiGe base current kink current factor A 0
BFN b-e flicker noise 1/f exponent — 1
CBCO extrinsic b-c overlap capacitance F 0
CBEO extrinsic b-e overlap capacitance F 0
CCSO extrinsic c-s overlap capacitance F 0
CJC zero-bias b-c depletion capacitance F 0
CJCP zero-bias extrinsic c-s depletion capacitance F 0

248
 Table 2-69. VBIC 1.3 3T Device Model Parameters
Parameter Description Units Default
CJE zero-bias b-e depletion capacitance F 0
CJEP zero-bias extrinsic b-c depletion capacitance F 0
CTH thermal capacitance — 0
DEAR delta activation energy for isrr V 0
EA activation energy for is V 1.12
EAIC activation energy for ibci and ibeip V 1.12
EAIE activation energy for ibei V 1.12
EAIS activation energy for ibcip V 1.12
EANC activation energy for ibcn and ibenp V 1.12
EANE activation energy for iben V 1.12
EANS activation energy for ibcnp V 1.12
EAP activation energy for isp V 1.12
FC forward bias depletion capacitance limit — 0.9
GAMM epi doping parameter — 0
GMIN minimum conductance ˙ −1 1e-12
HRCF high current collector resistance factor — 0
IBBE b-e breakdown current A 1e-06
IBCI ideal b-c saturation current A 1e-16
IBCIP ideal parasitic b-c saturation current A 0
IBCN non-ideal b-c saturation current A 0
IBCNP non-ideal parasitic b-c saturation current A 0
IBEI ideal b-e saturation current A 1e-18
IBEIP ideal parasitic b-e saturation current A 0
IBEN non-ideal b-e saturation current A 0
IBENP non-ideal parasitic b-e saturation current A 0
IBK0 SiGe base current kink current reference A 0
IKF forward knee current (zero=infinite) A 0
IKP parasitic knee current (zero=infinite) A 0
IKR reverse knee current (zero=infinite) A 0
IS transport saturation current A 1e-16
ISP parasitic transport saturation current A 0
ISRR ratio of is(reverse) to is(forward) — 1
ITF tf coefficient of Ic dependence A 0
KFN b-e flicker noise constant — 0
MAXEXP argument at which to linearize general exponentials — 1e+22
MC b-c grading coefficient — 0.33

249
 Table 2-69. VBIC 1.3 3T Device Model Parameters
Parameter Description Units Default
MCX bx-cx grading coefficient for avalanche — 0.33
ME b-e grading coefficient — 0.33
MS c-s grading coefficient — 0.33
NBBE b-e breakdown emission coefficient — 1
NCI ideal b-c emission coefficient — 1
NCIP ideal parasitic b-c emission coefficient — 1
NCN non-ideal b-c emission coefficient — 2
NCNP non-ideal parasitic b-c emission coefficient — 2
NEI ideal b-e emission coefficient — 1
NEN non-ideal b-e emission coefficient — 2
NF fwd emission coefficient (ideality factor) — 1
NFP parasitic emission coeff (ideality factor) — 1
NKF high current beta roll-off parameter — 0.5
NPN npn transistor type — 0
NR rev emission coefficient (ideality factor) — 1
OFF Set to 1 to initialize device to OFF instead of normally — 0
PC b-c built-in potential V 0.75
PE b-e built-in potential V 0.75
PNJMAXI current at which to linearize diode currents A 1
PNP pnp transistor type — 0
PS c-s built-in potential V 0.75
QBM base charge model selection switch: 0=GP and 1=SGP — 0
QCO epi charge parameter C 0
QNIBEIR ideal b-e quasi-neutral base recombination parameter — 0
QTF variation of tf with base-width modulation — 0
RBI intrinsic base resistance ˙ 0
RBP parasitic transistor base resistance ˙ 0
RBX extrinsic base resistance ˙ 0
RCI intrinsic collector resistance ˙ 0
RCX extrinsic collector resistance ˙ 0
RE extrinsic emitter resistance ˙ 0
RS extrinsic substrate resistance ˙ 0
RTH thermal resistance — 0
SCALE scale factor for instance geometries — 1
SHRINK shrink percentage for instance geometries — 0
TAVC temperature exponent of avc2 ◦ C −1 0

250
 Table 2-69. VBIC 1.3 3T Device Model Parameters
Parameter Description Units Default
TAVCX temperature exponent of avcx2 ◦ C −1 0
TCRTH temperature exponent of rth ◦ C −1 0
TCVEF temperature exponent of vef ◦ C −1 0
TCVER temperature exponent of ver ◦ C −1 0
TD forward excess-phase delay time s 0
TF forward transit time s 0
TMAX maximum ambient temperature ◦C 500
TMAXCLIP clip maximum temperature ◦C 500
TMIN minimum ambient temperature ◦C -100
TMINCLIP clip minimum temperature ◦C -100
TNBBE temperature coefficient of nbbe ◦ C −1 0
TNF temperature exponent of nf and nr ◦ C −1 0
TNOM nominal (reference) temperature ◦C 27
TR reverse transit time s 0
TVBBE1 linear temperature coefficient of vbbe ◦ C −1 0
TVBBE2 quadratic temperature coefficient of vbbe — 0
transistor type: -1=npn and +1=pnp (overriden by npn
TYPE — -1
or pnp)
VBBE b-e breakdown voltage V 0
VEF forward Early voltage (zero=infinite) V 0
VER reverse Early voltage (zero=infinite) V 0
VO epi drift saturation voltage V 0
VPTE SiGe base current kink voltage V 0
VRT reach-through voltage for Cbc limiting V 0
VTF tf coefficient of Vbci dependence V 0
WBE partitioning of Ibe/Ibex and Qbe/Qbex — 1
WSP partitioning of Iccp between Vbep and Vbci — 1
XII temperature exponent of ibei, ibci, ibeip, ibcip — 3
XIKF temperature exponent of ikf — 0
XIN temperature exponent of iben, ibcn, ibenp, ibcnp — 3
XIS temperature exponent of is — 3
XISR temperature exponent for isrr — 0
XRB temperature exponent of rbx and rbi — 0
XRBI temperature exponent of rbi (overrides xrb) — 0
XRBP temperature exponent of rbp (overrides xrc) — 0
XRBX temperature exponent of rbx (overrides xrb) — 0

251
 Table 2-69. VBIC 1.3 3T Device Model Parameters
Parameter Description Units Default
XRC temperature exponent of rci and rcx and rbp — 0
XRCI temperature exponent of rci (overrides xrc) — 0
XRCX temperature exponent of rcx (overrides xrc) — 0
XRE temperature exponent of re — 0
XRS temperature exponent of rs — 0
XTF tf bias dependence coefficient — 0
XVO temperature exponent of vo — 0

252
 Table 2-70. VBIC 1.3 4T Device Instance Parameters
Parameter Description Units Default
DTA Alias for trise ◦C 0
DTEMP Alias for trise ◦C 0
M multiplicity factor — 1
OFF Set to 1 to initialize device to OFF instead of normally — 0
SW_ET switch for self-heating: 0=no and 1=yes — 1
SW_NOISE switch for including noise: 0=no and 1=yes — 1
local temperature delta to ambient (before ◦C
TRISE 0
self-heating)

Table 2-71. VBIC 1.3 4T Device Model Parameters

Parameter Description Units Default
ABK SiGe base current kink exponent — 1
AFN b-e flicker noise current exponent — 1
AJC b-c capacitance smoothing factor — -0.5
AJE b-e capacitance smoothing factor — -0.5
AJS c-s capacitance smoothing factor — -0.5
ART smoothing parameter for reach-through — 0.1
AVC1 b-c weak avalanche parameter 1 V−1 0
AVC2 b-c weak avalanche parameter 2 — 0
AVCX1 bx-cx weak avalanche parameter 1 V−1 0
AVCX2 bx-cx weak avalanche parameter 2 — 0
BBK SiGe base current kink current factor A 0
BFN b-e flicker noise 1/f exponent — 1
CBCO extrinsic b-c overlap capacitance F 0
CBEO extrinsic b-e overlap capacitance F 0
CCSO extrinsic c-s overlap capacitance F 0
CJC zero-bias b-c depletion capacitance F 0
CJCP zero-bias extrinsic c-s depletion capacitance F 0
CJE zero-bias b-e depletion capacitance F 0
CJEP zero-bias extrinsic b-c depletion capacitance F 0
CTH thermal capacitance — 0
DEAR delta activation energy for isrr V 0
EA activation energy for is V 1.12
EAIC activation energy for ibci and ibeip V 1.12
EAIE activation energy for ibei V 1.12

253
 Table 2-71. VBIC 1.3 4T Device Model Parameters
Parameter Description Units Default
EAIS activation energy for ibcip V 1.12
EANC activation energy for ibcn and ibenp V 1.12
EANE activation energy for iben V 1.12
EANS activation energy for ibcnp V 1.12
EAP activation energy for isp V 1.12
FC forward bias depletion capacitance limit — 0.9
GAMM epi doping parameter — 0
GMIN minimum conductance ˙ −1 1e-12
HRCF high current collector resistance factor — 0
IBBE b-e breakdown current A 1e-06
IBCI ideal b-c saturation current A 1e-16
IBCIP ideal parasitic b-c saturation current A 0
IBCN non-ideal b-c saturation current A 0
IBCNP non-ideal parasitic b-c saturation current A 0
IBEI ideal b-e saturation current A 1e-18
IBEIP ideal parasitic b-e saturation current A 0
IBEN non-ideal b-e saturation current A 0
IBENP non-ideal parasitic b-e saturation current A 0
IBK0 SiGe base current kink current reference A 0
IKF forward knee current (zero=infinite) A 0
IKP parasitic knee current (zero=infinite) A 0
IKR reverse knee current (zero=infinite) A 0
IS transport saturation current A 1e-16
ISP parasitic transport saturation current A 0
ISRR ratio of is(reverse) to is(forward) — 1
ITF tf coefficient of Ic dependence A 0
KFN b-e flicker noise constant — 0
MAXEXP argument at which to linearize general exponentials — 1e+22
MC b-c grading coefficient — 0.33
MCX bx-cx grading coefficient for avalanche — 0.33
ME b-e grading coefficient — 0.33
MS c-s grading coefficient — 0.33
NBBE b-e breakdown emission coefficient — 1
NCI ideal b-c emission coefficient — 1
NCIP ideal parasitic b-c emission coefficient — 1
NCN non-ideal b-c emission coefficient — 2

254
 Table 2-71. VBIC 1.3 4T Device Model Parameters
Parameter Description Units Default
NCNP non-ideal parasitic b-c emission coefficient — 2
NEI ideal b-e emission coefficient — 1
NEN non-ideal b-e emission coefficient — 2
NF fwd emission coefficient (ideality factor) — 1
NFP parasitic emission coeff (ideality factor) — 1
NKF high current beta roll-off parameter — 0.5
NPN npn transistor type — 0
NR rev emission coefficient (ideality factor) — 1
OFF Set to 1 to initialize device to OFF instead of normally — 0
PC b-c built-in potential V 0.75
PE b-e built-in potential V 0.75
PNJMAXI current at which to linearize diode currents A 1
PNP pnp transistor type — 0
PS c-s built-in potential V 0.75
QBM base charge model selection switch: 0=GP and 1=SGP — 0
QCO epi charge parameter C 0
QNIBEIR ideal b-e quasi-neutral base recombination parameter — 0
QTF variation of tf with base-width modulation — 0
RBI intrinsic base resistance ˙ 0
RBP parasitic transistor base resistance ˙ 0
RBX extrinsic base resistance ˙ 0
RCI intrinsic collector resistance ˙ 0
RCX extrinsic collector resistance ˙ 0
RE extrinsic emitter resistance ˙ 0
RS extrinsic substrate resistance ˙ 0
RTH thermal resistance — 0
SCALE scale factor for instance geometries — 1
SHRINK shrink percentage for instance geometries — 0
TAVC temperature exponent of avc2 ◦ C −1 0
TAVCX temperature exponent of avcx2 ◦ C −1 0
TCRTH temperature exponent of rth ◦ C −1 0
TCVEF temperature exponent of vef ◦ C −1 0
TCVER temperature exponent of ver ◦ C −1 0
TD forward excess-phase delay time s 0
TF forward transit time s 0
TMAX maximum ambient temperature ◦C 500

255
 Table 2-71. VBIC 1.3 4T Device Model Parameters
Parameter Description Units Default
TMAXCLIP clip maximum temperature ◦C 500
TMIN minimum ambient temperature ◦C -100
TMINCLIP clip minimum temperature ◦C -100
TNBBE temperature coefficient of nbbe ◦ C −1 0
TNF temperature exponent of nf and nr ◦ C −1 0
TNOM nominal (reference) temperature ◦C 27
TR reverse transit time s 0
TVBBE1 linear temperature coefficient of vbbe ◦ C −1 0
TVBBE2 quadratic temperature coefficient of vbbe — 0
transistor type: -1=npn and +1=pnp (overriden by npn
TYPE — -1
or pnp)
VBBE b-e breakdown voltage V 0
VEF forward Early voltage (zero=infinite) V 0
VER reverse Early voltage (zero=infinite) V 0
VO epi drift saturation voltage V 0
VPTE SiGe base current kink voltage V 0
VRT reach-through voltage for Cbc limiting V 0
VTF tf coefficient of Vbci dependence V 0
WBE partitioning of Ibe/Ibex and Qbe/Qbex — 1
WSP partitioning of Iccp between Vbep and Vbci — 1
XII temperature exponent of ibei, ibci, ibeip, ibcip — 3
XIKF temperature exponent of ikf — 0
XIN temperature exponent of iben, ibcn, ibenp, ibcnp — 3
XIS temperature exponent of is — 3
XISR temperature exponent for isrr — 0
XRB temperature exponent of rbx and rbi — 0
XRBI temperature exponent of rbi (overrides xrb) — 0
XRBP temperature exponent of rbp (overrides xrc) — 0
XRBX temperature exponent of rbx (overrides xrb) — 0
XRC temperature exponent of rci and rcx and rbp — 0
XRCI temperature exponent of rci (overrides xrc) — 0
XRCX temperature exponent of rcx (overrides xrc) — 0
XRE temperature exponent of re — 0
XRS temperature exponent of rs — 0
XTF tf bias dependence coefficient — 0
XVO temperature exponent of vo — 0

256
2.3.17.5. Level 23 BJT Tables (FBH HBT_X)

Table 2-72. FBH HBT_X v2.1 Device Instance Parameters

Parameter Description Units Default
L Length of emitter fingers m 3e-05
N Number of emitter fingers – 1
TEMP Device operating temperature ◦C 25
W Width of emitter fingers m 3e-06

Table 2-73. FBH HBT_X v2.1 Device Model Parameters

Parameter Description Units Default
AHC – 0
BF – 100
BR – 1
BVCEO – 0
BVEBO – 0
CJC – 1e-15
CJE – 1e-15
CMIN – 1e-16
CPB – 0
CPC – 0
CQ – 0
CTH – 7e-07
DEBUG – 0
DEBUGPLUS – 0
IKF – 0
IKR – 0
J0 – 0.001
JK – 0.0004
JSC – 0
JSE – 0
JSEE – 0
JSF – 2e-23
JSR – 2e-17
KBETA – 0
KC – 0
KJC – 1

257
 Table 2-73. FBH HBT_X v2.1 Device Model Parameters
Parameter Description Units Default
L Length of emitter fingers m 3e-05
LB – 0
LC – 0
LE – 0
MC – 0
MJC – 0.5
MJE – 0.5
MODE – 1
N Number of emitter fingers – 1
NC – 0
NE – 0
NEE – 0
NF – 1
NOISE – 1
NR – 1
RB – 1
RB2 – 1
RBBXX – 1e+06
RBXX – 1e+06
RC – 1
RCI0 – 0.001
RCXX – 1e+06
RE – 1
RJK – 0.001
RTH – 0.1
TEMP Device operating temperature ◦C 25
TF – 1e-12
TFT – 0
THCS – 0
TNOM – 20
TR – 1e-15
TRX – 1e-15
VAF – 0
VAR – 0
VCES – 0.001
VG – 1.3

258
 Table 2-73. FBH HBT_X v2.1 Device Model Parameters
Parameter Description Units Default
VGB – 0
VGBB – 0
VGC – 0
VGR – 0
VJC – 1.3
VJE – 1.3
W Width of emitter fingers m 3e-06
XCJC – 0.5
XJ0 – 1

259
2.3.17.6. Level 230 BJT Tables (HICUM/L0)

The HICUM/L0 device supports output of the internal variables in table 2-76 on the .PRINT line of a
netlist. To access them from a print line, use the syntax N(<instance>:<variable>) where
“<instance>” refers to the name of the specific HICUM/L0 Q device in your netlist.

Table 2-74. HICUM L0 v1.32 Device Instance Parameters

Parameter Description Units Default
DT Temperature change for particular transistor – 0

Table 2-75. HICUM L0 v1.32 Device Model Parameters

Parameter Description Units Default
AF flicker noise exponent factor – 2
AHC Smoothing facor for current dependence – 0.1
AHQ Smoothing factor for the d.c. injection width – 0
AJE Ratio of maximum to zero-bias value – 2.5
BE capacitance ratio Ratio maximum to zero-bias
AJEDC – 2.5
value for d.c. transfer current
ALCES Relative TC of vces – 0
ALEAV TC of avalanche exponential factor – 0
ALIQFH Frist-order TC of iqfh – 0
ALIT Factor for additional delay time of transfer current – 0.333
ALKAV TC of avalanche prefactor – 0
ALQF Factor for additional delay time of minority charge – 0.167
ALT0 Frist-order TC of tf0 – 0
ALVS Relative TC of satur.drift velocity – 0
AVER bias dependence for reverse Early voltage – 0
CBCPAR Collector-base isolation (overlap) capacitance – 0
CBEPAR Emitter-base oxide capacitance – 0
CJCI0 Total zero-bias BC depletion capacitance – 1e-20
CJCX0 Zero-bias external BC depletion capacitance – 1e-20
CJE0 Zero-bias BE depletion capacitance – 1e-20
CJS0 Zero-bias SC depletion capacitance – 1e-20
CTH Thermal capacitance – 0
DT0H – 0
DVGBE Bandgap difference between base and BE-junction – 0
EAVL Exponent factor – 0
-
F1VG Coefficient K1 in T-dependent bandgap equation –
0.000102377

260
 Table 2-75. HICUM L0 v1.32 Device Model Parameters
Parameter Description Units Default
F2VG Coefficient K2 in T-dependent bandgap equation – 0.00043215
FBC Split factor = Cjci0/Cjc0 – 1
FGEO Geometry factor – 0.656
FIQF flag for turning on base related critical current – 0
FLNQS Flag for turning on and off of vertical NQS effect – 0
FLSH Flag for self-heating calculation – 0
GTE Exponent factor for emmiter transit time – 1
IBCS BC saturation current – 0
IBES BE saturation current – 1e-18
IQF forward d.c. high-injection toll-off current – 1e+06
IQFH high-injection correction current – 1e+06
IQR inverse d.c. high-injection roll-off current – 1e+06
IRES BE recombination saturation current – 0
IS (Modified) saturation current – 1e-16
ISCS SC saturation current – 0
IT_MOD Flag for using third order solution for transfer current – 0
ITSS Substrate transistor transfer saturation current – 0
KAVL Prefactor – 0
KF flicker noise coefficient – 0
KIQFH Second-order TC of iqfh – 0
KT0 Second-order TC of tf0 – 0
MBC BC non-ideality factor – 1
MBE BE non-ideality factor – 1
MCF Non-ideality coefficient of forward collector current – 1
MCR Non-ideality coefficient of reverse collector current – 1
MRE BE recombination non-ideality factor – 2
MSC SC non-ideality factor – 1
MSF Substrate transistor transfer current non-ideality factor – 1
RBI0 Internal base resistance at zero-bias – 0
RBX External base series resistance – 0
RCI0 Low-field collector resistance under emitter – 150
RCX Emitter series resistance – 0
RE External collector series resistance – 0
RTH Thermal resistance – 0
T0 low current transit time at Vbici=0 – 0
TBVL SCR width modulation contribution – 0

261
 Table 2-75. HICUM L0 v1.32 Device Model Parameters
Parameter Description Units Default
TEF0 Storage time in neutral emitter – 0
Flag for turning temperature dependence of tef0 on
TEF_TEMP – 1
and off
TFH high-injection correction factor – 0
THCS Saturation time at high current densities – 0
TNOM Temperature for which parameters are valid – 27
TR Storage time at inverse operation – 0
TYPE For transistor type NPN(+1) or PNP (-1) – 1
VCES Saturation voltage – 0.1
VDCI BC built-in voltage – 0.7
VDCX External BC built-in voltage – 0.7
VDE BE built-in voltage – 0.9
VDEDC BE charge built-in voltage for d.c. transfer current – 0.9
VDS SC built-in voltage – 0.3
VEF forward Early voltage (normalization volt.) – 1e+06
VER reverse Early voltage (normalization volt.) – 1e+06
VGB Bandgap-voltage – 1.2
VGC Effective collector bandgap-voltage – 1.17
VGE Effective emitter bandgap-voltage – 1.17
VGS Effective substrate bandgap-voltage – 1.17
VLIM Voltage dividing ohmic and satur.region – 0.5
VPT Punch-through voltage – 100
VPTCI Punch-through voltage of BC junction – 100
VPTCX Punch-through voltage – 100
VPTS SC punch-through voltage – 100
VR0C forward Early voltage (normalization volt.) – 1e+06
VR0E forward Early voltage (normalization volt.) – 2.5
ZCI BC exponent factor – 0.333
ZCX External BC exponent factor – 0.333
ZE BE exponent factor – 0.5
ZEDC charge BE exponent factor for d.c. transfer current – 0.5
Exponent coefficient in BE junction current
ZETABET – 3.5
temperature dependence
ZETACI TC of epi-collector diffusivity – 0
Exponent coefficient in transfer current temperature
ZETACT – 3
dependence
ZETAIQF TC of iqf – 0

262
 Table 2-75. HICUM L0 v1.32 Device Model Parameters
Parameter Description Units Default
ZETARBI TC of internal base resistance – 0
ZETARBX TC of external base resistance – 0
ZETARCX TC of external collector resistance – 0
ZETARE TC of emitter resistances – 0
Exponent factor for temperature dependent thermal
ZETARTH – 0
resistance
ZETAVER TC of Reverse Early voltage – -1
ZETAVGBE TC of AVER – 1
ZS External SC exponent factor – 0.3

Table 2-76. BJT level 230 Output Variables

Parameter Description Units Default
qjci B-C internal junction charge C none
qjei B-E internal junction charge C none
it Transfer Current A none
ijbc Base-collector diode current A none
iavl Avalanche current A none
ijsc Substrate-collector diode current A none
Base-collector diode current minus the avalanche
Ibici A none
current
ijbe Base-emitter diode current A none
IAVL Avalanche current A none
VBE External BE voltage V none
VBC External BC voltage V none
VCE External CE voltage V none
VSC External SC voltage V none
GMi Internal transconductance ˙ −1 none
RPIi Internal input resistance Ohm none
RMUi Internal feedback resistance Ohm none
ROi Internal Output resistance Ohm none
CPIi Total BE capacitance F none
CMUi Total internal BC capacitance F none
CBCX Total external BC capacitance F none
CCS CS junction capacitance F none
RBi Internal base resistance Ohm none

263
 Table 2-76. BJT level 230 Output Variables
Parameter Description Units Default
RB Total base resistance Ohm none
RCX External (saturated) collector series resistance Ohm none
RE Emitter series resistance Ohm none
BETAAC Small signal current gain – none
TF Total forward transit time s none
FT Transit frequency Hz none

264
2.3.17.7. Level 234 BJT Table (HICUM/L2)

NOTE: The HICUM/L2 model has no instance parameters. The HICUM/L2 device supports output of the
internal variables in table 2-78 on the .PRINT line of a netlist. To access them from a print line, use the
syntax N(<instance>:<variable>) where “<instance>” refers to the name of the specific HICUM/L2
Q device in your netlist.

Table 2-77. HICUM v2.4.0 Device Model Parameters

Parameter Description Units Default
ABET Exponent factor for tunneling current – 40
ACBAR Smoothing parameter for barrier voltage – 0.01
AF Flicker noise exponent factor – 2
AFRE Emitter resistance flicker noise exponent factor – 2
Smoothing factor for current dependence of base and
AHC – 0.1
collector transit time
AHJEI Parameter describing the slope of hjEi(VBE) – 0
AICK Smoothing term for ICK – 0.001
Ratio of maximum to zero-bias value of internal B-E
AJEI – 2.5
capacitance
Ratio of maximum to zero-bias value of peripheral
AJEP – 2.5
B-E capacitance
ALB Relative TC of forward current gain for V2.1 model – 0
ALCES Relative TC of VCES – 0
ALFAV Relative TC for FAVL – 0
ALIT Factor for additional delay time of transfer current – 0.333
ALKAV Relative TC for KAVL – 0
ALQAV Relative TC for QAVL – 0
ALQF Factor for additional delay time of minority charge – 0.167
ALRTH First order relative TC of parameter Rth – 0
ALT0 First order relative TC of parameter T0 – 0
ALVS Relative TC of saturation drift velocity – 0
C10 GICCR constant – 2e-30
CBCPAR Total parasitic B-C capacitance – 0
CBEPAR Total parasitic B-E capacitance – 0
Flag for determining where to tag the flicker noise
CFBE – -1
source
CJCI0 Internal B-C zero-bias depletion capacitance – 1e-20
CJCX0 External B-C zero-bias depletion capacitance – 1e-20
CJEI0 Internal B-E zero-bias depletion capacitance – 1e-20
CJEP0 Peripheral B-E zero-bias depletion capacitance – 1e-20

265
 Table 2-77. HICUM v2.4.0 Device Model Parameters
Parameter Description Units Default
CJS0 C-S zero-bias depletion capacitance – 0
CSCP0 Perimeter S-C zero-bias depletion capacitance – 0
CSU Substrate shunt capacitance – 0
CTH Thermal capacitance – 0
DELCK Fitting factor for critical current – 2
Temperature change w.r.t. chip temperature for
DT – 0
particular transistor
Time constant for base and B-C space charge layer
DT0H – 0
width modulation
Bandgap difference between B and B-E junction used
DVGBE – 0
for hjEi0 and hf0
-
F1VG Coefficient K1 in T-dependent band-gap equation –
0.000102377
F2VG Coefficient K2 in T-dependent band-gap equation – 0.00043215
FAVL Avalanche current factor – 0
FBCPAR Partitioning factor of parasitic B-C cap – 0
FBEPAR Partitioning factor of parasitic B-E cap – 1
Ratio of HF shunt to total internal capacitance (lateral
FCRBI – 0
NQS effect)
Correction factor for modulation by B-E and B-C
FDQR0 – 0
space charge layer
Factor for geometry dependence of emitter current
FGEO – 0.6557
crowding
FLCOMP Flag for compatibility with v2.1 model (0=v2.1) – 0
Flag for turning on and off of correlated noise
FLCONO – 0
implementation
FLNQS Flag for turning on and off of vertical NQS effect – 0
FLSH Flag for turning on and off self-heating effect – 0
FQI Ration of internal to total minority charge – 1
FTHC Partitioning factor for base and collector portion – 0
Exponent factor for current dependence of neutral
GTFE – 1
emitter storage time
HF0 Weight factor for the low current minority charge – 1
HFC Collector minority charge weighting factor in HBTs – 1
HFE Emitter minority charge weighting factor in HBTs – 1
HJCI B-C depletion charge weighting factor in HBTs – 1
HJEI B-E depletion charge weighting factor in HBTs – 1
IBCIS Internal B-C saturation current – 1e-16

266
 Table 2-77. HICUM v2.4.0 Device Model Parameters
Parameter Description Units Default
IBCXS External B-C saturation current – 0
IBEIS Internal B-E saturation current – 1e-18
IBEPS Peripheral B-E saturation current – 0
IBETS B-E tunneling saturation current – 0
ICBAR Normalization parameter – 0
ICH High-current correction for 2D and 3D effects – 0
IREIS Internal B-E recombination saturation current – 0
IREPS Peripheral B-E recombination saturation current – 0
ISCS C-S diode saturation current – 0
ITSS Substrate transistor transfer saturation current – 0
KAVL Flag/factor for turning strong avalanche on – 0
KF Flicker noise coefficient – 0
KFRE Emitter resistance flicker noise coefficient – 0
KT0 Second order relative TC of parameter T0 – 0
Scaling factor for collector minority charge in
LATB – 0
direction of emitter width
Scaling factor for collector minority charge in
LATL – 0
direction of emitter length
MBCI Internal B-C current ideality factor – 1
MBCX External B-C current ideality factor – 1
MBEI Internal B-E current ideality factor – 1
MBEP Peripheral B-E current ideality factor – 1
MCF Non-ideality factor for III-V HBTs – 1
MREI Internal B-E recombination current ideality factor – 2
MREP Peripheral B-E recombination current ideality factor – 2
MSC Ideality factor of C-S diode current – 1
MSF Forward ideality factor of substrate transfer current – 1
QAVL Exponent factor for avalanche current – 0
QP0 Zero-bias hole charge – 2e-14
RBI0 Zero bias internal base resistance – 0
RBX External base series resistance – 0
RCI0 Internal collector resistance at low electric field – 150
RCX External collector series resistance – 0
RE Emitter series resistance – 0
RHJEI Smoothing parameter for hjEi(VBE) at high voltage – 1
RSU Substrate series resistance – 0

267
 Table 2-77. HICUM v2.4.0 Device Model Parameters
Parameter Description Units Default
RTH Thermal resistance – 0
T0 Low current forward transit time at VBC=0V – 0
Base current recombination time constant at B-C
TBHREC – 0
barrier for high forward injection
TBVL Time constant for modeling carrier jam at low VCE – 0
TEF0 Neutral emitter storage time – 0
THCS Saturation time constant at high current densities – 0
TNOM Temperature at which parameters are specified – 27
TR Storage time for inverse operation – 0
Transit time for forward operation of substrate
TSF – 0
transistor
Specifies the base node connection for the tunneling
TUNODE – 1
current
TYPE For transistor type NPN(+1) or PNP (-1) – 1
VCBAR Barrier voltage – 0
VCES Internal C-E saturation voltage – 0.1
VDCI Internal B-C built-in potential – 0.7
VDCX External B-C built-in potential – 0.7
VDEI Internal B-E built-in potential – 0.9
VDEP Peripheral B-E built-in potential – 0.9
VDS C-S built-in potential – 0.6
VDSP Perimeter S-C built-in potential – 0.6
VGB Bandgap voltage extrapolated to 0 K – 1.17
VGC Effective collector bandgap voltage – 1.17
VGE Effective emitter bandgap voltage – 1.17
VGS Effective substrate bandgap voltage – 1.17
Voltage separating ohmic and saturation velocity
VLIM – 0.5
regime
VPT Collector punch-through voltage – 100
VPTCI Internal B-C punch-through voltage – 100
VPTCX External B-C punch-through voltage – 100
VPTS C-S punch-through voltage – 100
VPTSP Perimeter S-C punch-through voltage – 100
ZCI Internal B-C grading coefficient – 0.4
ZCX External B-C grading coefficient – 0.4
ZEI Internal B-E grading coefficient – 0.5
ZEP Peripheral B-E grading coefficient – 0.5

268
 Table 2-77. HICUM v2.4.0 Device Model Parameters
Parameter Description Units Default
Exponent coefficient in B-E junction current
ZETABET – 3.5
temperature dependence
ZETACI Temperature exponent for RCI0 – 0
Exponent coefficient in transfer current temperature
ZETACT – 3
dependence
Temperature exponent of mobility in substrate
ZETACX – 1
transistor transit time
ZETAHJEI Temperature coefficient for ahjEi – 1
ZETARBI Temperature exponent of internal base resistance – 0
ZETARBX Temperature exponent of external base resistance – 0
ZETARCX Temperature exponent of external collector resistance – 0
ZETARE Temperature exponent of emitter resistance – 0
ZETARTH Temperature coefficient for Rth – 0
ZETAVGBE Temperature coefficient for hjEi0 – 1
ZS C-S grading coefficient – 0.5
ZSP Perimeter S-C grading coefficient – 0.5

Table 2-78. BJT level 234 Output Variables

Parameter Description Units Default
rcx_t External (saturated) collector series resistance Ohm none
re_t Emitter series resistance Ohm none
rbi Internal base resistance as calculated in the model Ohm none
rb Total base resistance as calculated in the model Ohm none
IAVL Avalanche current A none
VBE External BE voltage V none
VBC External BC voltage V none
VCE External CE voltage V none
VSC External SC voltage V none
GMi Internal transconductance A/V none
GMS Transconductance of the parasitic substrate PNP A/V none
RPIi Internal base-emitter (input) resistance Ohm none
RPIx External base-emitter (input) resistance Ohm none
RMUi Internal feedback resistance Ohm none
RMUx External feedback resistance Ohm none
ROi Output resistance Ohm none
CPIi Total internal BE capacitance F none

269
 Table 2-78. BJT level 234 Output Variables
Parameter Description Units Default
CPIx Total external BE capacitance F none
CMUi Total internal BC capacitance F none
CMUx Total external BC capacitance F none
CCS CS junction capacitance F none
BETAAC Small signal current gain – none
Shunt capacitance across RBI as calculated in the
CRBI F none
model
TF Forward transit time s none
FT Transit frequency Hz none
TK Actual device temperature K none
DTSH Temperature increase due to self-heating K none

270
2.3.17.8. Level 504 and 505 BJT Tables (MEXTRAM)

The MEXTRAM device supports output of the internal variables in tables 2-81 and 2-81 on the .PRINT
line of a netlist. To access them from a print line, use the syntax N(<instance>:<variable>) where
“<instance>” refers to the name of the specific MEXTRAM Q device in your netlist.

Table 2-79. MEXTRAM 504.12.1 Device Instance Parameters

Parameter Description Units Default
M Alias for MULT — 1
MULT Multiplication factor — 1

Table 2-80. MEXTRAM 504.12.1 Device Model Parameters

Parameter Description Units Default
AB Temperature coefficient of the resistivity of the base — 1
Temperature coefficient of the resistivity of the
AC — 2
collector contact
Temperature coefficient of the resistivity of the
ACBL — 2
collector buried layer
AE Temperature coefficient of the resistivity of the emitter — 0
Temperature coefficient of the resistivity of the
AEPI — 2.5
epilayer
Temperature coefficient of the resistivity of the
AEX — 0.62
extrinsic base
AF Exponent of the Flicker-noise — 2
AQBO Temperature coefficient of the zero-bias base charge — 0.3
AS Substrate temperature coefficient — 1.58
Temperature coefficient for mobility of minorities in
ASUB — 2
the substrate
Temperature coefficient band-gap voltage for Zener
AVGEB — 0.000473
effect emitter-base junction
Smoothness parameter for the onset of
AXI — 0.3
quasi-saturation
BF Ideal forward current gain — 215
BRI Ideal reverse current gain — 7
CBCO Collector-base overlap capacitance — 0
CBEO Emitter-base overlap capacitance — 0
CJC Zero-bias collector-base depletion capacitance — 7.8e-14
CJE Zero-bias emitter-base depletion capacitance — 7.3e-14
CJS Zero-bias collector-substrate depletion capacitance — 3.15e-13

271
 Table 2-80. MEXTRAM 504.12.1 Device Model Parameters
Parameter Description Units Default
Fine tuning of temperature dependence of C-E
DAIS — 0
saturation current
DEG Bandgap difference over the base — 0
Difference between the local and global ambient
DTA — 0
temperatures
Band-gap voltage difference of the forward current
DVGBF — 0.05
gain
DVGBR Band-gap voltage difference of the reverse current gain — 0.045
DVGTE Band-gap voltage difference of emitter stored charge — 0.05
EXAVL Flag for extended modeling of avalanche currents — 0
EXMOD Flag for extended modeling of the reverse current gain — 1
Flag for the distributed high-frequency effects in
EXPHI — 1
transient
EXSUB Flag for extended modelling of substrate currents — 0
FTAUN Fraction of noise transit time to total transit time — 0
GMIN Minimum conductance — 1e-13
Saturation current of the non-ideal forward base
IBF — 2.7e-15
current
Saturation current of the non-ideal reverse base
IBR — 1e-15
current
ICSS Collector-substrate ideal saturation current — -1
IHC Critical current for velocity saturation in the epilayer — 0.004
IK Collector-emitter high injection knee current — 0.1
IKS Base-substrate high injection knee current — 0.00025
IS Collector-emitter saturation current — 2.2e-17
ISS Base-substrate saturation current — 4.8e-17
IZEB Pre-factor of emitter-base Zener tunneling current — 0
KAVL Switch for white noise contribution due to avalanche — 0
KC Switch for RF correlation noise model selection — 0
KE Fraction of QE in excess phase shift — 0
KF Flicker-noise coefficient of the ideal base current — 2e-11
KFN Flicker-noise coefficient of the non-ideal base current — 2e-11
LEVEL Model level — 504
M Alias for MULT — 1
Coefficient for current modulation of CB depletion
MC — 0.5
capacitance
Non-ideality factor of the non-ideal forward base
MLF — 2
current

272
 Table 2-80. MEXTRAM 504.12.1 Device Model Parameters
Parameter Description Units Default
MTAU Non-ideality factor of the emitter stored charge — 1
MULT Multiplication factor — 1
NZEB Coefficient of emitter-base Zener tunneling current — 22
PC Collector-base grading coefficient — 0.5
PE Emitter-base grading coefficient — 0.4
PS Collector-substrate grading coefficient — 0.34
RBC Constant part of the base resistance — 23
Zero-bias value of the variable part of the base
RBV — 18
resistance
RCBLI Resistance Collector Buried Layer Intrinsic — 0
RCBLX Resistance Collector Buried Layer eXtrinsic — 0
RCC Constant part of the collector resistance — 12
RCV Resistance of the un-modulated epilayer — 150
RE Emitter resistance — 5
SCRCV Space charge resistance of the epilayer — 1250
Current spreading factor of avalanche model when
SFH — 0.3
EXAVL=1
TAUB Transit time of stored base charge — 4.2e-12
TAUE Minimum transit time of stored emitter charge — 2e-12
TAUR Transit time of reverse extrinsic stored base charge — 5.2e-10
TEPI Transit time of stored epilayer charge — 4.1e-11
TREF Reference temperature — 25
Temperature coefficient band-gap voltage for Zener
TVGEB — 636
effect emitter-base junction
TYPE Flag for NPN (1) or PNP (-1) transistor type — 1
VAVL Voltage determining curvature of avalanche current — 3
VDC Collector-base diffusion voltage — 0.68
VDE Emitter-base diffusion voltage — 0.95
VDS Collector-substrate diffusion voltage — 0.62
VEF Forward Early voltage — 44
VER Reverse Early voltage — 2.5
VGB Band-gap voltage of the base — 1.17
VGC Band-gap voltage of the collector — 1.18
VGJ Band-gap voltage recombination emitter-base junction — 1.15
VGS Band-gap voltage of the substrate — 1.2
Band-gap voltage at Tref of Zener effect emitter-base
VGZEB — 1.15
junction

273
 Table 2-80. MEXTRAM 504.12.1 Device Model Parameters
Parameter Description Units Default
Cross-over voltage of the non-ideal reverse base
VLR — 0.2
current
WAVL Epilayer thickness used in weak-avalanche model — 1.1e-06
Fraction of CB depletion capacitance under the
XCJC — 0.032
emitter
Sidewall fraction of the emitter-base depletion
XCJE — 0.4
capacitance
Part of currents and charges that belong to extrinsic
XEXT — 0.63
region
XIBI Part of ideal base current that belongs to the sidewall — 0
XP Constant part of Cjc — 0.35
XQB Emitter-fraction of base diffusion charge — 0.333333
XREC Pre-factor of the recombination part of Ib1 — 0

Table 2-81. BJT level 504 Output Variables

Parameter Description Units Default
OP_ic External DC collector current A none
OP_ib External DC base current A none
OP_betadc External DC current gain Ic/Ib none
OP_ie External DC emitter current A none
OP_vbe External base-emitter bias V none
OP_vce External collector-emitter bias V none
OP_vbc External base-collector bias V none
OP_is External DC substrate current A none
OP_vse External substrate-emitter bias V none
OP_vbs External base-substrate bias V none
OP_vsc External substrate-collector bias V none
OP_vb2e1 Internal base-emitter bias V none
OP_vb2c2 Internal base-collector bias V none
OP_vb2c1 Internal base-collector bias including epilayer V none
External base-collector bias without contact
OP_vb1c1 V none
resistances
OP_vc4c1 Bias over intrinsic buried layer V none
OP_vc3c4 Bias over extrinsic buried layer V none
OP_ve1e Bias over emitter resistance V none
OP_in Main current A none

274
 Table 2-81. BJT level 504 Output Variables
Parameter Description Units Default
OP_ic1c2 Epilayer current A none
OP_ib1b2 Pinched-base current A none
OP_ib1 Ideal forward base current A none
OP_sib1 Ideal side-wall base current A none
OP_izteb Zener tunneling current in the emitter base junction A none
OP_ib2 Non-ideal forward base current A none
OP_ib3 Non-ideal reverse base current A none
OP_iavl Avalanche current A none
OP_iex Extrinsic reverse base current A none
OP_xiex Extrinsic reverse base current A none
OP_isub Substrate current A none
OP_xisub Substrate current A none
OP_isf Substrate failure current A none
OP_ire Current through emitter resistance A none
OP_irbc Current through constant base resistance A none
OP_ircblx Current through extrinsic buried layer resistance A none
OP_ircbli Current through intrinsic buried layer resistance A none
OP_ircc Current through collector contact resistance A none
OP_qe Emitter charge or emitter neutral charge C none
OP_qte Base-emitter depletion charge C none
OP_sqte Sidewall base-emitter depletion charge C none
OP_qbe Base-emitter diffusion charge C none
OP_qbc Base-collector diffusion charge C none
OP_qtc Base-collector depletion charge C none
OP_qepi Epilayer diffusion charge C none
OP_qb1b2 AC current crowding charge C none
OP_qtex Extrinsic base-collector depletion charge C none
OP_xqtex Extrinsic base-collector depletion charge C none
OP_qex Extrinsic base-collector diffusion charge C none
OP_xqex Extrinsic base-collector diffusion charge C none
OP_qts Collector-substrate depletion charge C none
OP_gx Forward transconductance ˙ −1 none
OP_gy Reverse transconductance ˙ −1 none
OP_gz Reverse transconductance ˙ −1 none
OP_sgpi Conductance sidewall b-e junction ˙ −1 none
OP_gpix Conductance floor b-e junction ˙ −1 none

275
 Table 2-81. BJT level 504 Output Variables
Parameter Description Units Default
OP_gpiy Early effect on recombination base current ˙ −1 none
OP_gpiz Early effect on recombination base current ˙ −1 none
OP_gmux Early effect on avalanche current limiting ˙ −1 none
OP_gmuy Conductance of avalanche current ˙ −1 none
OP_gmuz Conductance of avalanche current ˙ −1 none
OP_gmuex Conductance of extrinsic b-c junction ˙ −1 none
OP_xgmuex Conductance of extrinsic b-c junction ˙ −1 none
OP_grcvy Conductance of epilayer current ˙ −1 none
OP_grcvz Conductance of epilayer current ˙ −1 none
OP_rbv Base resistance Ohm none
OP_grbvx Early effect on base resistance ˙ −1 none
OP_grbvy Early effect on base resistance ˙ −1 none
OP_grbvz Early effect on base resistance ˙ −1 none
OP_re Emitter resistance Ohm none
OP_rbc Constant base resistance Ohm none
OP_rcc Collector contact resistance Ohm none
OP_rcblx Extrinsic buried layer resistance Ohm none
OP_rcbli Intrinsic buried layer resistance Ohm none
OP_gs Conductance parasitic PNP transistor ˙ −1 none
OP_xgs Conductance parasitic PNP transistor ˙ −1 none
OP_gsf Conductance substrate failure current ˙ −1 none
OP_scbe Capacitance sidewall b-e junction F none
OP_cbex Capacitance floor b-e junction F none
OP_cbey Early effect on b-e diffusion charge F none
OP_cbez Early effect on b-e diffusion charge F none
OP_cbcx Early effect on b-c diffusion charge F none
OP_cbcy Capacitance floor b-c junction F none
OP_cbcz Capacitance floor b-c junction F none
OP_cbcex Capacitance extrinsic b-c junction F none
OP_xcbcex Capacitance extrinsic b-c junction F none
OP_cb1b2 Capacitance AC current crowding F none
OP_cb1b2x Cross-capacitance AC current crowding F none
OP_cb1b2y Cross-capacitance AC current crowding F none
OP_cb1b2z Cross-capacitance AC current crowding F none
OP_cts Capacitance s-c junction F none
OP_gm transconductance ˙ −1 none

276
 Table 2-81. BJT level 504 Output Variables
Parameter Description Units Default
OP_beta Current amplification none
OP_gout Output conductance ˙ −1 none
OP_gmu Feedback transconductance ˙ −1 none
OP_rb Base resistance Ohm none
OP_rc Collector resistance Ohm none
OP_cbe Base-emitter capacitance C none
OP_cbc Base-collector capacitance C none
OP_ft Good approximation for cut-off frequency none
OP_iqs Current at onset of quasi-saturation A none
OP_xiwepi Thickness of injection layer m none
OP_vb2c2star Physical value of internal base-collector bias V none
OP_tk Actual temperature K none

277
 Table 2-82. MEXTRAM 504.12.1 with self heating Device Instance Parameters
Parameter Description Units Default
M Alias for MULT — 1
MULT Multiplication factor — 1

Table 2-83. MEXTRAM 504.12.1 with self heating Device Model Parameters
Parameter Description Units Default
AB Temperature coefficient of the resistivity of the base — 1
Temperature coefficient of the resistivity of the
AC — 2
collector contact
Temperature coefficient of the resistivity of the
ACBL — 2
collector buried layer
AE Temperature coefficient of the resistivity of the emitter — 0
Temperature coefficient of the resistivity of the
AEPI — 2.5
epilayer
Temperature coefficient of the resistivity of the
AEX — 0.62
extrinsic base
AF Exponent of the Flicker-noise — 2
AQBO Temperature coefficient of the zero-bias base charge — 0.3
AS Substrate temperature coefficient — 1.58
Temperature coefficient for mobility of minorities in
ASUB — 2
the substrate
ATH Temperature coefficient of the thermal resistance — 0
Temperature coefficient band-gap voltage for Zener
AVGEB — 0.000473
effect emitter-base junction
Smoothness parameter for the onset of
AXI — 0.3
quasi-saturation
BF Ideal forward current gain — 215
BRI Ideal reverse current gain — 7
CBCO Collector-base overlap capacitance — 0
CBEO Emitter-base overlap capacitance — 0
CJC Zero-bias collector-base depletion capacitance — 7.8e-14
CJE Zero-bias emitter-base depletion capacitance — 7.3e-14
CJS Zero-bias collector-substrate depletion capacitance — 3.15e-13
CTH Thermal capacitance — 3e-09
Fine tuning of temperature dependence of C-E
DAIS — 0
saturation current
DEG Bandgap difference over the base — 0

278
 Table 2-83. MEXTRAM 504.12.1 with self heating Device Model Parameters
Parameter Description Units Default
Difference between the local and global ambient
DTA — 0
temperatures
Band-gap voltage difference of the forward current
DVGBF — 0.05
gain
DVGBR Band-gap voltage difference of the reverse current gain — 0.045
DVGTE Band-gap voltage difference of emitter stored charge — 0.05
EXAVL Flag for extended modeling of avalanche currents — 0
EXMOD Flag for extended modeling of the reverse current gain — 1
Flag for the distributed high-frequency effects in
EXPHI — 1
transient
EXSUB Flag for extended modelling of substrate currents — 0
FTAUN Fraction of noise transit time to total transit time — 0
GMIN Minimum conductance — 1e-13
Saturation current of the non-ideal forward base
IBF — 2.7e-15
current
Saturation current of the non-ideal reverse base
IBR — 1e-15
current
ICSS Collector-substrate ideal saturation current — -1
IHC Critical current for velocity saturation in the epilayer — 0.004
IK Collector-emitter high injection knee current — 0.1
IKS Base-substrate high injection knee current — 0.00025
IS Collector-emitter saturation current — 2.2e-17
ISS Base-substrate saturation current — 4.8e-17
IZEB Pre-factor of emitter-base Zener tunneling current — 0
KAVL Switch for white noise contribution due to avalanche — 0
KC Switch for RF correlation noise model selection — 0
KE Fraction of QE in excess phase shift — 0
KF Flicker-noise coefficient of the ideal base current — 2e-11
KFN Flicker-noise coefficient of the non-ideal base current — 2e-11
LEVEL Model level — 504
M Alias for MULT — 1
Coefficient for current modulation of CB depletion
MC — 0.5
capacitance
Non-ideality factor of the non-ideal forward base
MLF — 2
current
MTAU Non-ideality factor of the emitter stored charge — 1
MULT Multiplication factor — 1
NZEB Coefficient of emitter-base Zener tunneling current — 22

279
 Table 2-83. MEXTRAM 504.12.1 with self heating Device Model Parameters
Parameter Description Units Default
PC Collector-base grading coefficient — 0.5
PE Emitter-base grading coefficient — 0.4
PS Collector-substrate grading coefficient — 0.34
RBC Constant part of the base resistance — 23
Zero-bias value of the variable part of the base
RBV — 18
resistance
RCBLI Resistance Collector Buried Layer Intrinsic — 0
RCBLX Resistance Collector Buried Layer eXtrinsic — 0
RCC Constant part of the collector resistance — 12
RCV Resistance of the un-modulated epilayer — 150
RE Emitter resistance — 5
RTH Thermal resistance — 300
SCRCV Space charge resistance of the epilayer — 1250
Current spreading factor of avalanche model when
SFH — 0.3
EXAVL=1
TAUB Transit time of stored base charge — 4.2e-12
TAUE Minimum transit time of stored emitter charge — 2e-12
TAUR Transit time of reverse extrinsic stored base charge — 5.2e-10
TEPI Transit time of stored epilayer charge — 4.1e-11
TREF Reference temperature — 25
Temperature coefficient band-gap voltage for Zener
TVGEB — 636
effect emitter-base junction
TYPE Flag for NPN (1) or PNP (-1) transistor type — 1
VAVL Voltage determining curvature of avalanche current — 3
VDC Collector-base diffusion voltage — 0.68
VDE Emitter-base diffusion voltage — 0.95
VDS Collector-substrate diffusion voltage — 0.62
VEF Forward Early voltage — 44
VER Reverse Early voltage — 2.5
VGB Band-gap voltage of the base — 1.17
VGC Band-gap voltage of the collector — 1.18
VGJ Band-gap voltage recombination emitter-base junction — 1.15
VGS Band-gap voltage of the substrate — 1.2
Band-gap voltage at Tref of Zener effect emitter-base
VGZEB — 1.15
junction
Cross-over voltage of the non-ideal reverse base
VLR — 0.2
current

280
 Table 2-83. MEXTRAM 504.12.1 with self heating Device Model Parameters
Parameter Description Units Default
WAVL Epilayer thickness used in weak-avalanche model — 1.1e-06
Fraction of CB depletion capacitance under the
XCJC — 0.032
emitter
Sidewall fraction of the emitter-base depletion
XCJE — 0.4
capacitance
Part of currents and charges that belong to extrinsic
XEXT — 0.63
region
XIBI Part of ideal base current that belongs to the sidewall — 0
XP Constant part of Cjc — 0.35
XQB Emitter-fraction of base diffusion charge — 0.333333
XREC Pre-factor of the recombination part of Ib1 — 0

Table 2-84. BJT level 505 Output Variables

Parameter Description Units Default
OP_ic External DC collector current A none
OP_ib External DC base current A none
OP_betadc External DC current gain Ic/Ib none
OP_ie External DC emitter current A none
OP_vbe External base-emitter bias V none
OP_vce External collector-emitter bias V none
OP_vbc External base-collector bias V none
OP_is External DC substrate current A none
OP_vse External substrate-emitter bias V none
OP_vbs External base-substrate bias V none
OP_vsc External substrate-collector bias V none
OP_vb2e1 Internal base-emitter bias V none
OP_vb2c2 Internal base-collector bias V none
OP_vb2c1 Internal base-collector bias including epilayer V none
External base-collector bias without contact
OP_vb1c1 V none
resistances
OP_vc4c1 Bias over intrinsic buried layer V none
OP_vc3c4 Bias over extrinsic buried layer V none
OP_ve1e Bias over emitter resistance V none
OP_in Main current A none
OP_ic1c2 Epilayer current A none
OP_ib1b2 Pinched-base current A none

281
 Table 2-84. BJT level 505 Output Variables
Parameter Description Units Default
OP_ib1 Ideal forward base current A none
OP_sib1 Ideal side-wall base current A none
OP_izteb Zener tunneling current in the emitter base junction A none
OP_ib2 Non-ideal forward base current A none
OP_ib3 Non-ideal reverse base current A none
OP_iavl Avalanche current A none
OP_iex Extrinsic reverse base current A none
OP_xiex Extrinsic reverse base current A none
OP_isub Substrate current A none
OP_xisub Substrate current A none
OP_isf Substrate failure current A none
OP_ire Current through emitter resistance A none
OP_irbc Current through constant base resistance A none
OP_ircblx Current through extrinsic buried layer resistance A none
OP_ircbli Current through intrinsic buried layer resistance A none
OP_ircc Current through collector contact resistance A none
OP_qe Emitter charge or emitter neutral charge C none
OP_qte Base-emitter depletion charge C none
OP_sqte Sidewall base-emitter depletion charge C none
OP_qbe Base-emitter diffusion charge C none
OP_qbc Base-collector diffusion charge C none
OP_qtc Base-collector depletion charge C none
OP_qepi Epilayer diffusion charge C none
OP_qb1b2 AC current crowding charge C none
OP_qtex Extrinsic base-collector depletion charge C none
OP_xqtex Extrinsic base-collector depletion charge C none
OP_qex Extrinsic base-collector diffusion charge C none
OP_xqex Extrinsic base-collector diffusion charge C none
OP_qts Collector-substrate depletion charge C none
OP_gx Forward transconductance ˙ −1 none
OP_gy Reverse transconductance ˙ −1 none
OP_gz Reverse transconductance ˙ −1 none
OP_sgpi Conductance sidewall b-e junction ˙ −1 none
OP_gpix Conductance floor b-e junction ˙ −1 none
OP_gpiy Early effect on recombination base current ˙ −1 none
OP_gpiz Early effect on recombination base current ˙ −1 none

282
 Table 2-84. BJT level 505 Output Variables
Parameter Description Units Default
OP_gmux Early effect on avalanche current limiting ˙ −1 none
OP_gmuy Conductance of avalanche current ˙ −1 none
OP_gmuz Conductance of avalanche current ˙ −1 none
OP_gmuex Conductance of extrinsic b-c junction ˙ −1 none
OP_xgmuex Conductance of extrinsic b-c junction ˙ −1 none
OP_grcvy Conductance of epilayer current ˙ −1 none
OP_grcvz Conductance of epilayer current ˙ −1 none
OP_rbv Base resistance Ohm none
OP_grbvx Early effect on base resistance ˙ −1 none
OP_grbvy Early effect on base resistance ˙ −1 none
OP_grbvz Early effect on base resistance ˙ −1 none
OP_re Emitter resistance Ohm none
OP_rbc Constant base resistance Ohm none
OP_rcc Collector contact resistance Ohm none
OP_rcblx Extrinsic buried layer resistance Ohm none
OP_rcbli Intrinsic buried layer resistance Ohm none
OP_gs Conductance parasitic PNP transistor ˙ −1 none
OP_xgs Conductance parasitic PNP transistor ˙ −1 none
OP_gsf Conductance substrate failure current ˙ −1 none
OP_scbe Capacitance sidewall b-e junction F none
OP_cbex Capacitance floor b-e junction F none
OP_cbey Early effect on b-e diffusion charge F none
OP_cbez Early effect on b-e diffusion charge F none
OP_cbcx Early effect on b-c diffusion charge F none
OP_cbcy Capacitance floor b-c junction F none
OP_cbcz Capacitance floor b-c junction F none
OP_cbcex Capacitance extrinsic b-c junction F none
OP_xcbcex Capacitance extrinsic b-c junction F none
OP_cb1b2 Capacitance AC current crowding F none
OP_cb1b2x Cross-capacitance AC current crowding F none
OP_cb1b2y Cross-capacitance AC current crowding F none
OP_cb1b2z Cross-capacitance AC current crowding F none
OP_cts Capacitance s-c junction F none
OP_gm transconductance ˙ −1 none
OP_beta Current amplification none
OP_gout Output conductance ˙ −1 none

283
 Table 2-84. BJT level 505 Output Variables
Parameter Description Units Default
OP_gmu Feedback transconductance ˙ −1 none
OP_rb Base resistance Ohm none
OP_rc Collector resistance Ohm none
OP_cbe Base-emitter capacitance C none
OP_cbc Base-collector capacitance C none
OP_ft Good approximation for cut-off frequency none
OP_iqs Current at onset of quasi-saturation A none
OP_xiwepi Thickness of injection layer m none
OP_vb2c2star Physical value of internal base-collector bias V none
OP_pdiss Dissipation W none
OP_tk Actual temperature K none

284
2.3.18. Junction Field-Effect Transistor (JFET)

Symbol

Instance Form J<name> <drain node> <gate node> <source node> <model name> +
[area value] [device parameters]

Examples JIN 100 1 0 JFAST

J13 22 14 23 JNOM 2.0
J1 1 2 0 2N5114

Model Form .MODEL <model name> NJF [model parameters]

.MODEL <model name> PJF [model parameters]

Parameters and
Options
drain node
Node connected to drain.
gate node
Node connected to gate.
source node
Node connected to source.
source node
Name of model defined in .MODEL line.
area value
The JFET is modeled as an intrinsic FET using an ohmic resistance
(RD/area) in series with the drain and another ohmic resistance (RS/area)
in series with the source. area is an area factor with a default of 1.
device parameters
Parameters listed in Table 2-85 may be provided as space separated
<parameter>=<value> specifications as needed. Any number of
parameters may be specified.

Comments The JFET was first proposed and analyzed by Shockley. The SPICE- compatible JFET
model is an approximation to the Shockley analysis that employs an adjustable
parameter B. Both the Shockley formulation and the SPICE approximation are
available in Xyce.

285
Device Parameters

Table 2-85. JFET Device Instance Parameters

Parameter Description Units Default
AREA Device area m2 1
Ambient
TEMP Device temperature –
Temperature

Model Parameters

Table 2-86. JFET Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
B Doping tail parameter (level 1) V−1 1
BETA Transconductance parameter A/V2 0.0001
CGD Zero-bias gate-drain junction capacitance F 0
CGS Zero-bias gate-source junction capacitance F 0
DELTA Saturation voltage parrameter (level 2) V 0
FC Coefficient for forward-bias depletion capacitance F 0.5
IS Gate junction saturation current A 1e-14
KF Flicker noise coefficient – 0.05
LAMBDA Channel length modulation V−1 0
PB Gate junction potential V 1
RD Drain ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
THETA Mobility modulation parameter (level 2) V−1 0
◦C Ambient
TNOM Nominal device temperature
Temperature
VTO Threshold voltage V -2

286
Device Parameters

Table 2-87. JFET Device Instance Parameters

Parameter Description Units Default
AREA Device area m2 1
Ambient
TEMP Device temperature –
Temperature

Model Parameters

Table 2-88. JFET Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
B Doping tail parameter (level 1) V−1 1
BETA Transconductance parameter A/V2 0.0001
CGD Zero-bias gate-drain junction capacitance F 0
CGS Zero-bias gate-source junction capacitance F 0
DELTA Saturation voltage parrameter (level 2) V 0
FC Coefficient for forward-bias depletion capacitance F 0.5
IS Gate junction saturation current A 1e-14
KF Flicker noise coefficient – 0.05
LAMBDA Channel length modulation V−1 0
PB Gate junction potential V 1
RD Drain ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
THETA Mobility modulation parameter (level 2) V−1 0
◦C Ambient
TNOM Nominal device temperature
Temperature
VTO Threshold voltage V -2

JFET Level selection Xyce supports two JFET models. LEVEL=1, the default, is the SPICE 3f5
treatment. This model employs a doping profile parameter B. When B=1, the original SPICE square law is
exactly implemented, and when B=0.6 the model is close to that of Shockley.

When LEVEL=2 is selected, the Shockley model is used with some additional physics effects: channel
length modulation and the effect of gate electric field on mobility. An additional parameter, DELTA, is
added to the LEVEL 2 model that allows the user to adjust the saturation voltage.

287
JFET Power Calculations Power dissipated in the transistor is calculated with 𝐼 𝐷 ∗ 𝑉𝐷𝑆 + 𝐼𝐺 ∗ 𝑉𝐺𝑆
where 𝐼 𝐷 is the drain current, 𝐼𝐺 is the gate current, 𝑉𝐷𝑆 is the voltage drop between the drain and the
source and 𝑉𝐺𝑆 is the voltage drop between the gate and the source. This formula may differ from other
simulators, such as HSPICE and PSpice.

288
2.3.19. Metal-Semiconductor FET (MESFET)

Symbol

Instance Form Z<name> < drain node> <gate node> <source node> <model name>
+ [area value] [device parameters]

Model Form .MODEL <model name> NMF [model parameters]

.MODEL <model name> PMF [model parameters]

Examples Z1 2 3 0 MESMOD AREA=1.4

Z1 7 2 3 ZM1

Parameters and
Options
drain node
Node connected to drain.
gate node
Node connected to gate.
source node
Node connected to source.
source node
Name of model defined in .MODEL line.
area value
The MESFET is modeled as an intrinsic FET using an ohmic resistance
(RD/area) in series with the drain and another ohmic resistance (RS/area)
in series with the source. area value is a scaling factor with a default of 1.
device parameters
Parameters listed in Table 2-89 may be provided as space separated
<parameter>=<value> specifications as needed. Any number of
parameters may be specified.

Comments Although MESFETs can be made of Si, such devices are not as common as GaAs
MESFETS. And since the mobility of electrons is much higher than holes in GaAs,
nearly all commercial devices are n-type MESFETS.

289
Device Parameters

Table 2-89. MESFET Device Instance Parameters

Parameter Description Units Default
AREA device area m2 1
Ambient
TEMP Device temperature –
Temperature

Model Parameters

Table 2-90. MESFET Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
ALPHA Saturation voltage parameter V−1 2
B Doping tail parameter V−1 0.3
BETA Transconductance parameter A/V2 0.0025
CGD Zero-bias gate-drain junction capacitance F 0
CGS Zero-bias gate-source junction capacitance F 0
FC Coefficient for forward-bias depletion capacitance F 0.5
IS Gate junction saturation current A 1e-14
KF Flicker noise coefficient – 0.05
LAMBDA Channel length modulation V−1 0
PB Gate junction potential V 1
RD Drain ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
◦C Ambient
TNOM Nominal device temperature
Temperature
VTO Threshold voltage V 0

MESFET Power Calculations Power dissipated in the transistor is calculated with 𝐼 𝐷 ∗ 𝑉𝐷𝑆 + 𝐼𝐺 ∗ 𝑉𝐺𝑆
where 𝐼 𝐷 is the drain current, 𝐼𝐺 is the gate current, 𝑉𝐷𝑆 is the voltage drop between the drain and the
source and 𝑉𝐺𝑆 is the voltage drop between the gate and the source. This formula may differ from other
simulators, such as HSPICE and PSpice.

290
2.3.20. MOS Field Effect Transistor (MOSFET)

Symbol

Instance Form M<name> <drain node> <gate node> <source node>

+ <bulk/substrate node> <model name>
+ [L=<value>] [W=<value>]
+ [AD=<value>] [AS=<value>]
+ [PD=<value>] [PS=<value>]
+ [NRD=<value>] [NRS=<value>]
+ [M=<value] [IC=<value, ...>]

Special Form M<name> <drain node> <gate node> <source node>

(BSIMSOI) + <substrate node (E)>
+ [<External body contact (P)>]
+ [<internal body contact (B)>]
+ [<temperature node (T)>]
+ <model name>
+ [L=<value>] [W=<value>]
+ [AD=<value>] [AS=<value>]
+ [PD=<value>] [PS=<value>]
+ [NRD=<value>] [NRS=<value>] [NRB=<value>]
+ [BJTOFF=<value>]
+ [IC=<val>,<val>,<val>,<val>,<val>]
+ [RTH0=<val>] [CTH0=<val>]
+ [NBC=<val>] [NSEG=<val>] [PDBCP=<val>] [PSBCP=<val>]
+ [AGBCP=<val>] [AEBCP=<val>] [VBSUSR=<val>] [TNODEOUT]
+ [FRBODY=<val>] [M=<value>]

Special Form M<name> <drain node> <gate node> <source node> <model name>
(MVS)
Special Form M<name> <drain node> <gate node> <source node> <bulk node> <dt node> <model nam
(PSP103 with
self-heating)
Model Form .MODEL <model name> NMOS [model parameters]
.MODEL <model name> PMOS [model parameters]

Examples M5 4 12 3 0 PNOM L=20u W=10u

M3 5 13 10 0 PSTRONG
M6 7 13 10 0 PSTRONG M=2
M8 10 12 100 100 NWEAK L=30u W=20u
+ AD=288p AS=288p PD=60u PS=60u NRD=14 NRS=24

291
Parameters and
Options L
M The MOSFET channel length and width that are decreased to get the actual
channel length and width. They may be given in the device .MODEL or
.OPTIONS statements. The value in the device statement overrides the value
in the model statement, which overrides the value in the .OPTIONS
statement. If L or W values are not given, their default value is 100 𝜇m.
AD
AS The drain and source diffusion areas. Defaults for AD and AS can be set in the
.OPTIONS statement. If AD or AS defaults are not set, their default value is 0.
PD
PS The drain and source diffusion perimeters. Their default value is 0.
NRD
NRS Multipliers (in units of □) that can be multiplied by RSH to yield the parasitic
(ohmic) resistances of the drain (RD) and source (RS), respectively. NRD, NRS
default to 0.
Consider a square sheet of resistive material. Analysis shows that the
resistance between two parallel edges of such a sheet depends upon its
composition and thickness, but is independent of its size as long as it is
square. In other words, the resistance will be the same whether the square’s
edge is 2 mm, 2 cm, or 2 m. For this reason, the sheet resistance of such a
layer, abbreviated RSH, has units of Ohms per square, written ˙/□.

M If specified, the value is used as a number of parallel MOSFETs to be

simulated. For example, if M=2 is specified, Xyce simulates two identical
mosfets connected to the same nodes in parallel.

IC The BSIM3 (model level 9), BSIM4 (model level 14 or 54) and BSIMSOI
(model level 10) allow one to specify the initial voltage difference across
nodes of the device during the DC operating point calculation. For the
BSIM3 and BSIM4 the syntax is IC=𝑉𝑑𝑠 , 𝑉𝑔𝑠 , 𝑉𝑏𝑠 where 𝑉𝑑𝑠 is the voltage
difference between the drain and source, 𝑉𝑔𝑠 is the voltage difference
between the gate and source and 𝑉𝑏𝑠 is the voltage difference between the
body and source. The BSIMSOI device’s initial condition syntax is
IC=𝑉𝑑𝑠 , 𝑉𝑔𝑠 , 𝑉𝑏𝑠 , 𝑉𝑒𝑠 , 𝑉 𝑝𝑠 where the two extra terms are the voltage
difference between the substrate and source, and the external body and
source nodes respectively. Note that for any of these lists of voltage
differences, fewer than the full number of options may be specified. For
example, IC=5.0 specifies an initial condition on 𝑉𝑑𝑠 but does not specifiy
any initial conditions on the other nodes. Therefore, one cannot specify 𝑉𝑔𝑠
without specifying 𝑉𝑑𝑠 , etc.
It is illegal to specify initial conditions on any nodes that are tied together.
Xyce attempts to catch such errors, but complex circuits may stymie this
error trap.

292
BSIM-SOI There are a large number of extra instance parameters and optional nodes available for
Options the BSIM-SOI (level 10 (BSIM-SOI 3.2), level 70 (BSIM-SOI 4.6.1), and level 70450
(BSIM-SOI 4.5.0)) MOSFET. Please consult the BSIM-SOI technical manual,
available at http://bsim.berkeley.edu/models/bsimsoi/, for full details.

substrate node
The fourth node of the BSIM-SOI device is always the substrate node, which
is referred to as the E node.
external body contact node
If given, the fifth node is the external body contact node, P. It is connected to
the internal body node through a body tie resistor. If P is not given, the
internal body node is not accessible from the netlist and floats.
For the BSIM-SOI 3.2 (level=10) only): If there are only five nodes specified
and TNODEOUT is also specified, the fifth node is the temperature node
instead.
internal body contact node
If given, the sixth node is the internal body contact node, B. It is connected to
the external body node through a body tie resistor. If B is not given and P is
given, the internal body node is not accessible from the netlist, but is still
tied to the external body contact through the tie resistance.
For the BSIM-SOI 3.2 (level=10) only): If there are only six nodes specified
and TNODEOUT is also specified, the sixth node is the temperature node
instead.
temperature node
For the BSIM-SOI 3.2 (level=10) only): If the parameter TNODEOUT is
specified, the final node (fifth, sixth, or seventh) is interpreted as a
temperature node. The temperature node is intended for thermal coupling
simulation.
For the BSIM-SOI 4.x (level=70 or 70450) only): The temperature node is
only accessible for thermal coupling if it is the seventh node. It is available
for printing as an internal node in all other configurations.
BJTOFF
Turns off the parasitic BJT currents.

IC The IC parameter allows specification of the five junction initial conditions,

𝑉𝑑𝑠 , 𝑉𝑔𝑠 , 𝑉𝑏𝑠 , 𝑉𝑒𝑠 and 𝑉 𝑝𝑠 . 𝑉 𝑝𝑠 is ignored in a four-terminal device.
RTH0
Thermal resistance per unit width. Taken from model card if not given.
CTH0
Thermal capacitance per unit width. Taken from model card if not given.

NBC Number of body contact isolation edges.

NSEG
Number of segments for channel width partitioning.

293
 PDBCP
Parasitic perimeter length for body contact at drain side.
PSBCP
Parasitic perimeter length for body contact at source side.
AGBCP
Parasitic gate-to-body overlap area for body contact.
AEBCP
Parasitic body-to-substrate overlap area for body contact.
VBSUSR
Optional initial value of VBS specified by user for use in transient analysis.
(unused in Xyce).
FRBODY
Layout-dependent body resistance coefficient.

Comments The simulator provides multiple MOSFET device models, which differ in the
formulation of the I-V characteristic. The LEVEL parameter selects among different
models as shown below.

For HSPICE compatibility, the BSIM4 model can be specified with either level 14 or
level 54.

If a model supports parameter aliases (e.g. “U0” and “UO” or “VT0” and “VTO” in
the levels 1-6 MOSFETS), it would be a mistake to specify both parameters and give
them different values. There is no warning or error message if you do that. Don’t do
that.

MOSFET Operating Temperature Model parameters may be assigned unique measurement

temperatures using the TNOM model parameter. See the MOSFET model parameters for more
information.

MOSFET Power Calculations Power dissipated in the transistor is calculated with 𝐼 𝐷 ∗ 𝑉𝐷𝑆 + 𝐼𝐺 ∗ 𝑉𝐺𝑆
where 𝐼 𝐷 is the drain current, 𝐼𝐺 is the gate current, 𝑉𝐷𝑆 is the voltage drop between the drain and the
source and 𝑉𝐺𝑆 is the voltage drop between the gate and the source. This formula may differ from other
simulators, such as HSPICE and PSpice.

Internal Device Variables Accessible with N() Syntax For the BSIM3, BSIM4, and BSIM-CMG
version 110 models, several internal variables have been made accessible with the N() syntax on a .PRINT
line. They are 𝑔𝑚 (tranconductance), 𝑉𝑡 ℎ , 𝑉𝑑𝑠 , 𝑉𝑔𝑠 , 𝑉𝑏𝑠 , and 𝑉𝑑𝑠𝑎𝑡 . An example .PRINT line command for a
MOSFET device named m1 would be:

.print dc N(m1:gm) N(m1:Vth) N(m1:Vdsat) N(m1:Vds) N(m1:Vgs) N(m1:Vbs)

294
The BSIM-CMG also supports output of 𝐼 𝑑𝑠 (drain-source current) in this manner.

If the user runs Xyce -namesfile <filename> <netlist> then Xyce will output into the first filename
a list of all solution variables generated by that netlist. This can be useful for determining the
“fully-qualified” device name, needed for the N() syntax, if the device is in a subcircuit.

Instance Parameters Tables 2-91, 2-93, 2-95, 2-97, 2-99 and 2-101 give the available instance
parameters for the levels 1,2,3,6,9 and 10 MOSFETs, respectively.

In addition to the parameters shown in the tables, where a list of numbered initial condition parameters are
shown, the MOSFETs support a vector parameter for the initial conditions. IC1 and IC2 may therefore be
specified compactly as IC=<ic1>,<ic2>.

Model Parameters Tables 2-92, 2-94, 2-96, 2-98, 2-100, and 2-102 give the available model
parameters for the levels 1,2,3,6,9 and 10 MOSFETs, respectively.

For a thorough description of MOSFET models see [17, 18, 19, 20, 21, 22, 23, 24, 25, 26].

All MOSFET models The parameters shared by all MOSFET model levels are principally parasitic
element values (e.g., series resistance, overlap capacitance, etc.).

Model levels 1 and 3 The DC behaviors of the level 1 and 3 MOSFET models are defined by the
parameters VTO, KP, LAMBDA, PHI, and GAMMA. The simulator calculates these if the process
parameters (e.g., TOX, and NSUB) are specified, but these are always overridden by any user-defined
values. The VTO value is positive (negative) for modeling the enhancement mode and negative (positive)
for the depletion mode of N-channel (P-channel) devices.

For MOSFETs, the capacitance model enforces charge conservation, influencing just the Level 1 and 3
models.

Effective device parameter lengths and widths are calculated as follows:

𝑃𝑖 = 𝑃0 + 𝑃 𝐿 /𝐿 𝑒 + 𝑃𝑊 /𝑊𝑒

where
𝐿 𝑒 = effective length = L − (2 · LD)
𝑊𝑒 = effective width = W − (2 · WD)

See .MODEL (model definition) for more information.

295
Model level 9 (BSIM3 version 3.2.2) The University of California, Berkeley BSIM3 model is a
physical-based model with a large number of dependencies on essential dimensional and processing
parameters. It incorporates the key effects that are critical in modeling deep-submicrometer MOSFETs.
These include threshold voltage reduction, nonuniform doping, mobility reduction due to the vertical field,
bulk charge effect, carrier velocity saturation, drain-induced barrier lowering (DIBL), channel length
modulation (CLM), hot-carrier-induced output resistance reduction, subthreshold conduction, source/drain
parasitic resistance, substrate current induced body effect (SCBE) and drain voltage reduction in LDD
structure.

The BSIM3 Version 3.2.2 model is a deep submicron MOSFET model with several major enhancements
over earlier versions. These include a single I-V formula used to define the current and output conductance
for operating regions, improved narrow width device modeling, a superior capacitance model with
improved short and narrow geometry models, a new relaxation-time model to better transient modeling and
enhanced model fitting of assorted W/L ratios using a single parameter set. This version preserves the large
number of integrated dependencies on dimensional and processing parameters of the Version 2 model. For
further information, see Reference [18].

Additional notes

1. If any of the following BSIM3 3.2.2 model parameters are not specified, they are computed via the
following:

If VTHO is not specified, then:

√︁
VTHO = VFB + 𝜙 𝑠 K1 𝜙 𝑠

where:
VFB = −1.0

If VTHO is given, then:

√︁
VFB = VTHO − 𝜙 𝑠 + K1 𝑝ℎ𝑖 𝑠
𝑞 · NCH · XT2
VBX = 𝜙 𝑠 −
2𝜀
   𝑠𝑖 
2𝜀 𝑜𝑥 1
CF = ln 1 +
𝜋 4 × 107 · TOX

where:

𝑇2
𝐸 𝑔 (𝑇) = the energy bandgap at temperature 𝑇 = 1.16 −
7.02 × 104 (𝑇 + 1108)

2. If K1 and K2 are not given then they are computed via the following:
√︁
K1 = GAMMA2 − 2 · K2 𝜙 𝑠 − VBM
√ √
(GAMMA1 − GAMMA2) ( 𝜙 𝑠 − VBX − 𝜙 𝑠 )
K2 = √ √ √
2 𝜙 𝑠 ( 𝜙 𝑠 − VBM − 𝜙 𝑠 ) + VBM

296
 where:
 
NCH
𝜙𝑠 = 2𝑉𝑡 ln
𝑛𝑖
𝑉𝑡 = 𝑘𝑇/𝑞
  1.5  
10 𝑇 𝐸 𝑔 (𝑇)
𝑛𝑖 = 1.45 × 10 exp 21.5565981 −
300.15 2𝑉𝑡

3. If NCH is not specified and GAMMA1 is, then:

GAMMA12 × COX2
NCH =
2𝑞𝜀 𝑠𝑖

If GAMMA1 and NCH are not specified, then NCH defaults to 1.7 × 1023 𝑚 −3 and GAMMA1 is
computed using NCH: √︁
2𝑞𝜀 𝑠𝑖 · NCH
GAMMA1 =
COX
If GAMMA2 is not specified, then:
√︁
2𝑞𝜀 𝑠𝑖 · NSUB
GAMMA2 =
COX

4. If CGSO is not specified and DLC > 0, then:


0, ((DLC · COX) − CGSL) < 0
CGSO =
0.6 · XJ · COX, ((DLC · COX) − CGSL) ≥ 0

5. If CGDO is not specified and DLC > 0, then:


0, ((DLC · COX) − CGSL) < 0
CGDO =
0.6 · XJ · COX, ((DLC · COX) − CGSL) ≥ 0

Model level 10 (BSIM-SOI version 3.2) The BSIM-SOI is an international standard model for SOI
(silicon on insulator) circuit design and is formulated on top of the BSIM3v3 framework. A detailed
description can be found in the BSIM-SOI 3.1 User’s Manual [27] and the BSIM-SOI 3.2 release
notes [28].
This version (v3.2) of the BSIM-SOI includes three depletion models; the partially depleted BSIM-SOI PD
(soiMod=0), the fully depleted BSIM-SOI FD (soiMod=2), and the unified SOI model (soiMod=1).
BSIMPD is the Partial-Depletion (PD) mode of the BSIM-SOI. A typical PD SOI MOSFET is formed on a
thin SOI film which is layered on top of a buried oxide. BSIMPD has the following features and
enhancements:
• Real floating body simulation of both I-V and C-V. The body potential is determined by the balance
of all body current components.
• An improved parasitic bipolar current model. This includes enhancements in the various diode
leakage components, second order effects (high-level injection and Early effect), diffusion charge
equation, and temperature dependence of the diode junction capacitance.

297
 • An improved impact-ionization current model. The contribution from BJT current is also modeled by
the parameter Fbjtii.

• A gate-to-body tunneling current model, which is important to thin-oxide SOI technologies.

• Enhancements in the threshold voltage and bulk charge formulation of the high positive body bias
regime.

• Instance parameters (Pdbcp, Psbcp, Agbcp, Aebcp, Nbc) are provided to model the parasitics of
devices with various body-contact and isolation structures.

• An external body node (the 6th node) and other improvements are introduced to facilitate the
modeling of distributed body resistance.

• Self heating. An external temperature node (the 7th node) is supported to facilitate the simulation of
thermal coupling among neighboring devices.

• A unique SOI low frequency noise model, including a new excess noise resulting from the floating
body effect.

• Width dependence of the body effect is modeled by parameters (K1,K1w1,K1w2).

• Improved history dependence of the body charges with two new parameters (Fbody, DLCB).

• An instance parameter Vbsusr is provided for users to set the transient initial condition of the body
potential.

• The new charge-thickness capacitance model introduced in BSIM3v3.2, capMod=3, is included.

Quadratic Temperature Compensation SPICE temperature effects are the default, but MOSFET
levels 18, 19 and 20 have a more advanced temperature compensation available. By specifying
TEMPMODEL=QUADRATIC in the netlist, parameters can be interpolated quadratically between measured
values extracted from data. See Section ?? for more details.

298
MOSFET Equations The following equations define an N-channel MOSFET. The P-channel devices use
a reverse the sign for all voltages and currents. The equations use the following variables:

𝑉𝑏𝑠 = intrinsic substrate-intrinsic source voltage

𝑉𝑏𝑑 = intrinsic substrate-intrinsic drain voltage
𝑉𝑑𝑠 = intrinsic drain-substrate source voltage
𝑉𝑑𝑠𝑎𝑡 = saturation voltage
𝑉𝑔𝑠 = intrinsic gate-intrinsic source voltage
𝑉𝑔𝑑 = intrinsic gate-intrinsic drain voltage
𝑉𝑡 = 𝑘𝑇/𝑞 (thermal voltage)
𝑉𝑡 ℎ = threshold voltage
𝐶𝑜𝑥 = the gate oxide capacitance per unit area
𝑓 = noise frequency
𝑘 = Boltzmann’s constant
𝑞 = electron charge
𝐿𝑒 𝑓 𝑓 = effective channel length
𝑊𝑒 𝑓 𝑓 = effective channel width
𝑇 = analysis temperature (K)
𝑇0 = nominal temperature (set using TNOM option)

Other variables are listed in the BJT Equations section 2.3.17.1.

299
 All Levels

𝐼𝑔 = gate current = 0
𝐼𝑏 = bulk current = 𝐼𝑏𝑠 + 𝐼𝑏𝑑
where  
𝐼𝑏𝑠 = bulk-source leakage current = 𝐼𝑠𝑠 𝑒 𝑉𝑏𝑠 /( 𝑁 𝑉𝑡 ) − 1
 
𝐼 𝑑𝑠 = bulk-drain leakage current = 𝐼 𝑑𝑠 𝑒 𝑉𝑏𝑑 /( 𝑁 𝑉𝑡 ) − 1
where
if
JS = 0, or AS = 0 or AD = 0
then
𝐼𝑠𝑠 = IS
𝐼 𝑑𝑠 = IS
else
𝐼𝑠𝑠 = AS × JS + PS × JSSW
𝐼 𝑑𝑠 = AD × JS + PD × JSSW
𝐼 𝑑 = drain current = 𝐼 𝑑𝑟 𝑎𝑖𝑛 − 𝐼𝑏𝑑
𝐼𝑠 = source current = −𝐼 𝑑𝑟 𝑎𝑖𝑛 − 𝐼𝑏𝑠

Level 1: Idrain

300
 Normal Mode: 𝑉𝑑𝑠 > 0
Case 1
For cutoff region: 𝑉𝑔𝑠 − 𝑉𝑡𝑜 < 0

𝐼 𝑑𝑟 𝑎𝑖𝑛 = 0

Case 2
For linear region: 𝑉𝑑𝑠 < 𝑉𝑔𝑠 − 𝑉𝑡𝑜

𝐼 𝑑𝑟 𝑎𝑖𝑛 = (𝑊/𝐿) (KN/2) (1 + LAMBDA × 𝑉𝑑𝑠 )𝑉𝑑𝑠 (2(𝑉𝑔𝑠 − 𝑉𝑡𝑜 ) − 𝑉𝑑𝑠 )

Case 3
For saturation region: 0 ≤ 𝑉𝑔𝑠 − 𝑉𝑡𝑜 ≤ 𝑉𝑑𝑠

𝐼 𝑑𝑟 𝑎𝑖𝑛 = (𝑊/𝐿) (KN/2) (1 + LAMBDA · 𝑉𝑑𝑠 ) (𝑉𝑔𝑠 − 𝑉𝑡𝑜 ) 2

where
  1/2
𝑉𝑡𝑜 = VTO + GAMMA · (PHI − 𝑉𝑏𝑠 ) 1/2

Inverted Mode: 𝑉𝑑𝑠 < 0

Here, simply switch the source and drain in the normal mode equations given above.

Level 3: Idrain
See Reference [21] below for detailed information.

Capacitance

Level 1 ,2, and 3

𝐶𝑏𝑠 = bulk-source capacitance = area cap. + sidewall cap. + transit time cap.
𝐶𝑏𝑑 = bulk-drain capacitance = area cap. + sidewall cap. + transit time cap.

301
 where

𝑖𝑓
CBS = 0 and CBD = 0
𝑡ℎ𝑒𝑛
𝐶𝑏𝑠 = AS · CJ · 𝐶𝑏𝑠 𝑗 + PS · CJSW · 𝐶𝑏𝑠𝑠 + TT · 𝐺 𝑏𝑠
𝐶𝑏𝑑 = AD · CJ · 𝐶𝑏𝑑 𝑗 + PD · CJSW · 𝐶𝑏𝑑𝑠 + TT · 𝐺 𝑑𝑠
𝑒𝑙𝑠𝑒
𝐶𝑏𝑠 = CBS · 𝐶𝑏𝑠 𝑗 + PS · CJSW · 𝐶𝑏𝑠𝑠 + TT · 𝐺 𝑏𝑠
𝐶𝑏𝑑 = CBD · 𝐶𝑏𝑑 𝑗 + PD · CJSW · 𝐶𝑏𝑑𝑠 + TT · 𝐺 𝑑𝑠
𝑤ℎ𝑒𝑟𝑒
𝐺 𝑏𝑠 = DC bulk-source conductance = 𝑑𝐼𝑏𝑠 /𝑑𝑉𝑏𝑠
𝐺 𝑏𝑑 = DC bulk-drain conductance = 𝑑𝐼𝑏𝑑 /𝑑𝑉𝑏𝑑

𝑖𝑓
𝑉𝑏𝑠 ≤ FC · PB
𝑡ℎ𝑒𝑛
𝐶𝑏𝑠 𝑗 = (1 − 𝑉𝑏𝑠 /PB) −MJ
𝐶𝑏𝑠𝑠 = (1 − 𝑉𝑏𝑠 /PBSW) −MJSW
𝑖𝑓
𝑉𝑏𝑠 > FC · PB
𝑡ℎ𝑒𝑛
𝐶𝑏𝑠 𝑗 = (1 − FC) − (1+MJ) (1 − FC(1 + MJ) + MJ · 𝑉𝑏𝑠 /PB)
𝐶𝑏𝑠𝑠 = (1 − FC) − (1+MJSW) (1 − FC(1 + MJSW) + MJSW · 𝑉𝑏𝑠 /PBSW)
𝑖𝑓
𝑉𝑏𝑑 ≤ FC · PB
𝑡ℎ𝑒𝑛
𝐶𝑏𝑑 𝑗 = (1 − 𝑉𝑏𝑑 /PB) −MJ
𝐶𝑏𝑑𝑠 = (1 − 𝑉𝑏𝑑 /PBSW) −MJSW
𝑖𝑓
𝑉𝑏𝑑 > FC · PB
𝑡ℎ𝑒𝑛
𝐶𝑏𝑑 𝑗 = (1 − FC) − (1+MJ) (1 − FC(1 + MJ) + MJ · 𝑉𝑏𝑑 /PB)
𝐶𝑏𝑑𝑠 = (1 − FC) − (1+MJSW) (1 − FC(1 + MJSW))

𝐶𝑔𝑠 = gate-source overlap capacitance = CGSO · W

𝐶𝑔𝑑 = gate-drain overlap capacitance = CGDO · W
𝐶𝑔𝑏 = gate-bulk overlap capacitance = CGBO · L

302
Temperature Effects

All Levels

IS(𝑇) = IS · exp 𝐸 𝑔 (𝑇0 ) · 𝑇/𝑇0 − 𝐸 𝑔 (𝑇) /𝑉𝑡


JS(𝑇) = JS · exp 𝐸 𝑔 (𝑇0 ) · 𝑇/𝑇0 − 𝐸 𝑔 (𝑇) /𝑉𝑡


JSSW(𝑇) = JSSW · exp 𝐸 𝑔 (𝑇0 ) · 𝑇/𝑇0 − 𝐸 𝑔 (𝑇) /𝑉𝑡
PB(𝑇) = PB · 𝑇/𝑇0 − 3𝑉𝑡 ln(𝑇/𝑇0 ) − 𝐸 𝑔 (𝑇0 ) · 𝑇/𝑇0 + 𝐸 𝑔𝑇
PBSW(𝑇) = PBSW · 𝑇/𝑇0 − 3𝑉𝑡 ln(𝑇/𝑇0 ) − 𝐸 𝑔 (𝑇0 ) · 𝑇/𝑇0 + 𝐸 𝑔𝑇
PHI(𝑇) = PHI · 𝑇/𝑇0 − 3𝑉𝑡 ln(𝑇/𝑇0 ) − 𝐸 𝑔 (𝑇0 ) · 𝑇/𝑇0 + 𝐸 𝑔𝑇

where

𝐸 𝑔 (𝑇) = silicon bandgap energy = 1.16 − 0.000702𝑇 2 /(𝑇 + 1108)

CBD(𝑇) = CBD · (1 + MJ · (0.0004(𝑇 − 𝑇0 ) + (1 − PB(𝑇)/PB)))
CBS(𝑇) = CBS · (1 + MJ · (0.0004(𝑇 − 𝑇0 ) + (1 − PB(𝑇)/PB)))
CJ(𝑇) = CJ · (1 + MJ · (0.0004(𝑇 − 𝑇0 ) + (1 − PB(𝑇)/PB)))
CJSW(𝑇) = CJSW · (1 + MJSW · (0.0004(𝑇 − 𝑇0 ) + (1 − PB(𝑇)/PB)))
KP(𝑇) = KP · (𝑇/𝑇0 ) −3/2
UO(𝑇) = UO · (𝑇/𝑇0 ) −3/2
MUS(𝑇) = MUS · (𝑇/𝑇0 ) −3/2
MUZ(𝑇) = MUZ · (𝑇/𝑇0 ) −3/2
X3MS(𝑇) = X3MS · (𝑇/𝑇0 ) −3/2

303
2.3.20.1. Level 1 MOSFET Tables (SPICE Level 1)

Table 2-91. MOSFET level 1 Device Instance Parameters

Parameter Description Units Default
AD Drain diffusion area m2 0
AS Source diffusion area m2 0
IC1 Initial condition on Drain-Source voltage V 0
IC2 Initial condition on Gate-Source voltage V 0
IC3 Initial condition on Bulk-Source voltage V 0
L Channel length m 0
M Multiplier for M devices connected in parallel – 1
Multiplier for RSH to yield parasitic resistance of
NRD □ 1
drain
Multiplier for RSH to yield parasitic resistance of
NRS □ 1
source
logical
OFF Initial condition of no voltage drops across device false
(T/F)
PD Drain diffusion perimeter m 0
PS Source diffusion perimeter m 0
◦C Ambient
TEMP Device temperature
Temperature
W Channel width m 0

Table 2-92. MOSFET level 1 Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
CBD Zero-bias bulk-drain p-n capacitance F 0
CBS Zero-bias bulk-source p-n capacitance F 0
CGBO Gate-bulk overlap capacitance/channel length F/m 0
CGDO Gate-drain overlap capacitance/channel width F/m 0
CGSO Gate-source overlap capacitance/channel width F/m 0
CJ Bulk p-n zero-bias bottom capacitance/area F/m2 0
CJSW Bulk p-n zero-bias sidewall capacitance/area F/m2 0
FC Bulk p-n forward-bias capacitance coefficient – 0.5
GAMMA Bulk threshold parameter V1/2 0
IS Bulk p-n saturation current A 1e-14
JS Bulk p-n saturation current density A/m2 0
KF Flicker noise coefficient – 0

304
 Table 2-92. MOSFET level 1 Device Model Parameters
Parameter Description Units Default
KP Transconductance coefficient A/V2 2e-05
L Default channel length m 0.0001
LAMBDA Channel-length modulation V−1 0
LD Lateral diffusion length m 0
MJ Bulk p-n bottom grading coefficient – 0.5
MJSW Bulk p-n sidewall grading coefficient – 0.5
NSS Surface state density cm−2 0
NSUB Substrate doping density cm−3 0
PB Bulk p-n bottom potential V 0.8
PHI Surface potential V 0.6
RD Drain ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
RSH Drain,source diffusion sheet resistance ˙ 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
TNOM Nominal device temperature ◦C 27
TOX Gate oxide thickness m 1e-07
Gate material type (-1 = same as substrate) 0 =
TPG – 0
aluminum,1 = opposite of substrate)
U0 Surface mobility (alias for UO) 1/(Vcm2 s) 600
UO Surface mobility 1/(Vcm2 s) 600
VT0 Zero-bias threshold voltage (alias for VTO) V 0
VTO Zero-bias threshold voltage V 0
W Default channel width m 0.0001

305
2.3.20.2. Level 2 MOSFET Tables (SPICE Level 2)

Table 2-93. MOSFET level 2 Device Instance Parameters

Parameter Description Units Default
AD Drain diffusion area m2 0
AS Source diffusion area m2 0
IC1 Initial condition on Drain-Source voltage V 0
IC2 Initial condition on Gate-Source voltage V 0
IC3 Initial condition on Bulk-Source voltage V 0
L Channel length m 0
M Multiplier for M devices connected in parallel – 1
Multiplier for RSH to yield parasitic resistance of
NRD □ 1
drain
Multiplier for RSH to yield parasitic resistance of
NRS □ 1
source
logical
OFF Initial condition of no voltage drops across device false
(T/F)
PD Drain diffusion perimeter m 0
PS Source diffusion perimeter m 0
◦C Ambient
TEMP Device temperature
Temperature
W Channel width m 0

Table 2-94. MOSFET level 2 Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
CBD Zero-bias bulk-drain p-n capacitance F 0
CBS Zero-bias bulk-source p-n capacitance F 0
CGBO Gate-bulk overlap capacitance/channel length F/m 0
CGDO Gate-drain overlap capacitance/channel width F/m 0
CGSO Gate-source overlap capacitance/channel width F/m 0
CJ Bulk p-n zero-bias bottom capacitance/area F/m2 0
CJSW Bulk p-n zero-bias sidewall capacitance/area F/m2 0
DELTA Width effect on threshold – 0
FC Bulk p-n forward-bias capacitance coefficient – 0.5
GAMMA Bulk threshold parameter V1/2 0
IS Bulk p-n saturation current A 1e-14
JS Bulk p-n saturation current density A/m2 0

306
 Table 2-94. MOSFET level 2 Device Model Parameters
Parameter Description Units Default
KF Flicker noise coefficient – 0
KP Transconductance coefficient A/V2 2e-05
L Default channel length m 0.0001
LAMBDA Channel-length modulation V−1 0
LD Lateral diffusion length m 0
MJ Bulk p-n bottom grading coefficient – 0.5
MJSW Bulk p-n sidewall grading coefficient – 0.5
NEFF Total channel charge coeff. – 1
NFS Fast surface state density – 0
NSS Surface state density cm−2 0
NSUB Substrate doping density cm−3 0
PB Bulk p-n bottom potential V 0.8
PHI Surface potential V 0.6
RD Drain ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
RSH Drain,source diffusion sheet resistance ˙ 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
TNOM Nominal device temperature ◦C 27
TOX Gate oxide thickness m 1e-07
Gate material type (-1 = same as substrate, 0 =
TPG – 0
aluminum,1 = opposite of substrate)
U0 Surface mobility (alias for UO) 1/(Vcm2 s) 600
UCRIT Crit. field for mob. degradation – 10000
UEXP Crit. field exp for mob. deg. – 0
UO Surface mobility 1/(Vcm2 s) 600
VMAX Maximum carrier drift velocity – 0
VT0 Zero-bias threshold voltage (alias for VTO) V 0
VTO Zero-bias threshold voltage V 0
W Default channel width m 0.0001
XJ Junction depth – 0

307
2.3.20.3. Level 3 MOSFET Tables (SPICE Level 3)

Table 2-95. MOSFET level 3 Device Instance Parameters

Parameter Description Units Default
AD Drain diffusion area m2 0
AS Source diffusion area m2 0
IC1 Initial condition on Drain-Source voltage V 0
IC2 Initial condition on Gate-Source voltage V 0
IC3 Initial condition on Bulk-Source voltage V 0
L Channel length m 0
M Multiplier for M devices connected in parallel – 1
Multiplier for RSH to yield parasitic resistance of
NRD □ 1
drain
Multiplier for RSH to yield parasitic resistance of
NRS □ 1
source
logical
OFF Initial condition of no voltage drops across device false
(T/F)
PD Drain diffusion perimeter m 0
PS Source diffusion perimeter m 0
◦C Ambient
TEMP Device temperature
Temperature
W Channel width m 0

Table 2-96. MOSFET level 3 Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
CBD Zero-bias bulk-drain p-n capacitance F 0
CBS Zero-bias bulk-source p-n capacitance F 0
CGBO Gate-bulk overlap capacitance/channel length F/m 0
CGDO Gate-drain overlap capacitance/channel width F/m 0
CGSO Gate-source overlap capacitance/channel width F/m 0
CJ Bulk p-n zero-bias bottom capacitance/area F/m2 0
CJSW Bulk p-n zero-bias sidewall capacitance/area F/m2 0
DELTA Width effect on threshold – 0
ETA Static feedback – 0
FC Bulk p-n forward-bias capacitance coefficient – 0.5
GAMMA Bulk threshold parameter V1/2 0
IS Bulk p-n saturation current A 1e-14

308
 Table 2-96. MOSFET level 3 Device Model Parameters
Parameter Description Units Default
JS Bulk p-n saturation current density A/m2 0
KAPPA Saturation field factor – 0.2
KF Flicker noise coefficient – 0
KP Transconductance coefficient A/V2 2e-05
L Default channel length m 0.0001
LD Lateral diffusion length m 0
MJ Bulk p-n bottom grading coefficient – 0.5
MJSW Bulk p-n sidewall grading coefficient – 0.33
NFS Fast surface state density cm−2 0
NSS Surface state density cm−2 0
NSUB Substrate doping density cm−3 0
PB Bulk p-n bottom potential V 0.8
PHI Surface potential V 0.6
RD Drain ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
RSH Drain,source diffusion sheet resistance ˙ 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
THETA Mobility modulation V−1 0
TNOM Nominal device temperature ◦C 27
TOX Gate oxide thickness m 1e-07
Gate material type (-1 = same as substrate,0 =
TPG – 1
aluminum,1 = opposite of substrate)
U0 Surface mobility (alias for UO) 1/(Vcm2 s) 600
UO Surface mobility 1/(Vcm2 s) 600
VMAX Maximum drift velocity m/s 0
VT0 Zero-bias threshold voltage (alias for VTO) V 0
VTO Zero-bias threshold voltage V 0
W Default channel width m 0.0001
XJ Metallurgical junction depth m 0

309
2.3.20.4. Level 6 MOSFET Tables (SPICE Level 6)

Table 2-97. MOSFET level 6 Device Instance Parameters

Parameter Description Units Default
AD Drain diffusion area m2 0
AS Source diffusion area m2 0
IC1 Initial condition on Drain-Source voltage V 0
IC2 Initial condition on Gate-Source voltage V 0
IC3 Initial condition on Bulk-Source voltage V 0
L Channel length m 0
M Multiplier for M devices connected in parallel – 1
Multiplier for RSH to yield parasitic resistance of
NRD □ 1
drain
Multiplier for RSH to yield parasitic resistance of
NRS □ 1
source
logical
OFF Initial condition of no voltage drops across device false
(T/F)
PD Drain diffusion perimeter m 0
PS Source diffusion perimeter m 0
◦C Ambient
TEMP Device temperature
Temperature
W Channel width m 0

Table 2-98. MOSFET level 6 Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
CBD Zero-bias bulk-drain p-n capacitance F 0
CBS Zero-bias bulk-source p-n capacitance F 0
CGBO Gate-bulk overlap capacitance/channel length F/m 0
CGDO Gate-drain overlap capacitance/channel width F/m 0
CGSO Gate-source overlap capacitance/channel width F/m 0
CJ Bulk p-n zero-bias bottom capacitance/area F/m2 0
CJSW Bulk p-n zero-bias sidewall capacitance/area F/m2 0
FC Bulk p-n forward-bias capacitance coefficient – 0.5
GAMMA Bulk threshold parameter – 0
GAMMA1 Bulk threshold parameter 1 – 0
IS Bulk p-n saturation current A 1e-14
JS Bulk p-n saturation current density A/m2 0

310
 Table 2-98. MOSFET level 6 Device Model Parameters
Parameter Description Units Default
KC Saturation current factor – 5e-05
KF Flicker noise coefficient – 0
KV Saturation voltage factor – 2
LAMBDA Channel length modulation param. – 0
LAMBDA0 Channel length modulation param. 0 – 0
LAMBDA1 Channel length modulation param. 1 – 0
LD Lateral diffusion length m 0
MJ Bulk p-n bottom grading coefficient – 0.5
MJSW Bulk p-n sidewall grading coefficient – 0.5
NC Saturation current coeff. – 1
NSS Surface state density cm−2 0
NSUB Substrate doping density cm−3 0
NV Saturation voltage coeff. – 0.5
NVTH Threshold voltage coeff. – 0.5
PB Bulk p-n bottom potential V 0.8
PHI Surface potential V 0.6
PS Sat. current modification par. – 0
RD Drain ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
RSH Drain,source diffusion sheet resistance ˙ 0
SIGMA Static feedback effect par. – 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
TNOM Nominal device temperature ◦C 27
TOX Gate oxide thickness m 1e-07
Gate material type (-1 = same as substrate,0 =
TPG – 1
aluminum,1 = opposite of substrate)
U0 Surface mobility (alias for UO) 1/(Vcm2 s) 600
UO Surface mobility 1/(Vcm2 s) 600
VT0 Zero-bias threshold voltage (alias for VTO) V 0
VTO Zero-bias threshold voltage V 0

311
2.3.20.5. Level 9 MOSFET Tables (BSIM3)

For complete documentation of the BSIM3 model, see the users’ manual for the BSIM3, available for
download at http://bsim.berkeley.edu/models/bsim4/bsim3/. Xyce implements Version 3.2.2 of
the BSIM3.
In addition to the parameters shown in table 2-99, the BSIM3 supports a vector parameter for the initial
conditions. IC1 through IC3 may therefore be specified compactly as IC=<ic1>,<ic2>,<ic3>.
NOTE: Many BSIM3 parameters listed in tables 2-99 and 2-100 as having default values of zero are
actually replaced with internally computed defaults if not given. Specifying zero in your model card
will override this internal computation. It is recommended that you only set model parameters that
you are actually changing from defaults and that you not generate model cards containing default
values from the tables.

Table 2-99. BSIM3 Device Instance Parameters

Parameter Description Units Default
Control Parameters
M Multiplier for M devices connected in parallel – 1
NQSMOD Flag for NQS model – 0
Geometry Parameters
AD Drain diffusion area m2 0
AS Source diffusion area m2 0
L Channel length m 0
Multiplier for RSH to yield parasitic resistance of
NRD □ 1
drain
Multiplier for RSH to yield parasitic resistance of
NRS □ 1
source
PD Drain diffusion perimeter m 0
PS Source diffusion perimeter m 0
W Channel width m 0
Temperature Parameters

◦C Ambient
TEMP Device temperature
Temperature
Voltage Parameters
IC1 Initial condition on Vds V 0
IC2 Initial condition on Vgs V 0
IC3 Initial condition on Vbs V 0
logical
OFF Initial condition of no voltage drops accross device false
(T/F)

312
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
Bin Parameters
LMAX Maximum channel length m 1
LMIN Minimum channel length m 0
WMAX Maximum channel width m 1
WMIN Minimum channel width m 0
Capacitance Parameters
Exponetial coefficient for charge thickness in capmod
ACDE m/V 1
= 3 for accumulation and depletion regions
CF Firing field capacitance F/m 0
CGBO Gate-bulk overlap capacitance per unit channel length F/m 0
CGDL Light-doped drain-gate region overlap capacitance F/m 0
Non-LLD region drain-gate overlap capacitance per
CGDO F/m 0
unit channel length
CGSL Light-doped source-gate region overlap capacitance F/m 0
Non-LLD region source-gate overlap capacitance per
CGSO F/m 0
unit channel length
CJ Bulk p-n zero-bias bottom capacitance/area F/m2 0.0005
CJSW Bulk p-n zero-bias sidewall capacitance/area F/m2 5e-10
Source/grain gate sidewall junction capacitance per
CJSWG F/m 0
unit width
Coefficient for lightly doped region overlap
CKAPPA F/m 0.6
capacitance fireing field capacitance
CLC Constant term for short-channel model m 1e-07
CLE Exponetial term for the short-channel model – 0.6
DLC Length offset fitting parameter from C-V m 0
DWC Width offset fitting parameter from C-V m 0
Source/grain gate sidewall junction capacitance
MJSWG – 0
grading coeficient
Coefficient for the gate-bias dependent surface
MOIN – 15
potential
CV parameter in Vgsteff,CV for weak to strong
NOFF – 1
inversion
PBSW Source/drain side junction built-in potential V 1
PBSWG Source/drain gate sidewall junction built-in potential V 0
VFBCV Flat-band voltage parameter (for CAPMOD = 0 only) V -1
CV parameter in Vgsteff,CV for weak to strong
VOFFCV V 0
inversion
XPART Charge partitioning rate flag – 0

313
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
Control Parameters
BINUNIT Binning unit selector – 1
CAPMOD Flag for capacitance models – 3
MOBMOD Mobility model selector – 1
NOIMOD Flag for noise models – 1
PARAMCHK Parameter value check – 0
VERSION Version number – ’3.2.2’
DC Parameters
A0 Bulk charge effect coefficient for channel length – 1
A1 First non-saturation effect parameter V−1 0
A2 Second non-saturation factor – 1
AGS Gate-bias coefficient of abulk V−1 0
ALPHA0 First parameter of impact-ionization current m/V 0
ALPHA1 Isub parameter for length scaling V−1 0
B0 Bulk charge effect coefficient for channel width m 0
B1 Bulk charge effect offset m 0
BETA0 Second parameter of impact-ionization current V 30
CDSC Drain/source to channel coupling capacitance F/m2 0.00024
CDSCB Body-bias sensitivity of CDSC F/(Vm2 ) 0
CDSCD Drain-bias sensitivity of CDSC F/(Vm2 ) 0
CIT Interface trap capacitance F/m2 0
DELTA Effective Vds parameter V 0.01
L-depedance Coefficient of the DIBL correction
DROUT – 0.56
parameter in Rout
DSUB DIBL coefficient exponent in subthreshhold region – 0
First coefficient of short-channel effect effect on
DVT0 – 2.2
threshold voltage
First coefficient of narrow-width effect effect on
DVT0W m−1 0
threshold voltage for small channel length
Second coefficient of short-channel effect effect on
DVT1 – 0.53
threshold voltage
Second coefficient of narrow-width effect effect on
DVT1W m−1 5.3e+06
threshold voltage for small channel length
Body-bias coefficient of short-channel effect effect on
DVT2 V−1 -0.032
threshold voltage
Body-bias coefficient of narrow-width effect effect on
DVT2W V−1 -0.032
threshold voltage for small channel length
DWB Coefficient of substrate body bias dependence of Weff m/V1/2 0

314
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
DWG Coefficient of gate depedence of Weff m/V1/2 0
ETA0 DIBL coefficient in subthreshold region – 0.08
ETAB Body-bias coefficient for the subthreshold DIBL effect V−1 -0.07
IJTH Diode limiting current A 0.1
JSW Sidewall saturation current per unit length A/m 0
K1 First-order body effect coefficient V1/2 0
K2 second-order body effect coefficient – 0
K3 Narrow width coefficient – 80
K3B Body effect coefficient of K3 V−1 0
KETA Body-bias coefficient of bulk charge effect V−1 -0.047
LINT Length of offset fiting parameter from I-V without bias m 0
LINTNOI lint offset for noise calculation m 0
NFACTOR Subthreshold swing factor – 1
NGATE Poly gate doping concentration cm−3 0
NLX Lateral non-uniform doping parameter m 1.74e-07
PCLM Channel length modulation parameter – 1.3
First output resistance DIBL effect correction
PDIBLC1 – 0.39
parameter
Second output resistance DIBL effect correction
PDIBLC2 – 0.0086
parameter
PDIBLCB Body effect coefficient of DIBL correction parameter V−1 0
PRWB Body effect coefficient of RDSW V−1/2 0
PRWG Gate-bias effect coefficient of RDSW V−1 0
PSCBE1 First substrate current body effect parameter Vm−1 4.24e+08
PSCBE2 second substrate current body effect parameter Vm−1 1e-05
PVAG Gate dependence of early voltage – 0
RDSW Parasitic resistance per unit width ˙ 𝜇m 0
UA First-order mobility degradation coefficient m/V 2.25e-09
UB First-order mobility degradation coefficient m2 /V2 5.87e-19
UC Body effect of mobility degridation coefficient m/V2 0
Maximum applied body-bias in threshold voltage
VBM V -3
calculation
VFB Flat-band voltage V 0
Offset voltage in the subthreshold region at large W
VOFF V -0.08
and L
VSAT Saturation velocity at temp = TNOM m/s 80000
VTH0 Threshold voltage at Vbs = 0 for large L V 0

315
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
W0 Narrow-width paameter m 2.5e-06
WINT Width-offset fitting parameter from I-V without bias m 0
WR Width offset from Weff for Rds Calculation – 1
Dependency Parameters
LA0 Length dependence of A0 m 0
LA1 Length dependence of A1 m/V 0
LA2 Length dependence of A2 m 0
LACDE Length dependence of ACDE m2 /V 0
LAGS Length dependence of AGS m/V 0
LALPHA0 Length dependence of ALPHA0 m2 /V 0
LALPHA1 Length dependence of ALPHA1 m/V 0
LAT Length dependence of AT m2 /s 0
LB0 Length dependence of B0 m2 0
LB1 Length dependence of B1 m2 0
LBETA0 Length dependence of BETA0 Vm 0
LCDSC Length dependence of CDSC F/m 0
LCDSCB Length dependence of CDSCB F/(Vm) 0
LCDSCD Length dependence of CDSCD F/(Vm) 0
LCF Length dependence of CF F 0
LCGDL Length dependence of CGDL F 0
LCGSL Length dependence of CGSL F 0
LCIT Length dependence of CIT F/m 0
LCKAPPA Length dependence of CKAPPA F 0
LCLC Length dependence of CLC m2 0
LCLE Length dependence of CLE m 0
LDELTA Length dependence of DELTA Vm 0
LDROUT Length dependence of DROUT m 0
LDSUB Length dependence of DSUB m 0
LDVT0 Length dependence of DVT0 m 0
LDVT0W Length dependence of DVT0W – 0
LDVT1 Length dependence of DVT1 m 0
LDVT1W Length dependence of DVT1W – 0
LDVT2 Length dependence of DVT2 m/V 0
LDVT2W Length dependence of DVT2W m/V 0
LDWB Length dependence of DWB m2 /V1/2 0
LDWG Length dependence of DWG m2 /V1/2 0

316
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
LELM Length dependence of ELM m 0
LETA0 Length dependence of ETA0 m 0
LETAB Length dependence of ETAB m/V 0
LGAMMA1 Length dependence of GAMMA1 V1/2 m 0
LGAMMA2 Length dependence of GAMMA2 V1/2 m 0
LK1 Length dependence of K1 V1/2 m 0
LK2 Length dependence of K2 m 0
LK3 Length dependence of K3 m 0
LK3B Length dependence of K3B m/V 0
LKETA Length dependence of KETA m/V 0
LKT1 Length dependence of KT1 Vm 0
LKT1L Length dependence of KT1L Vm2 0
LKT2 Length dependence of KT2 m 0
LMOIN Length dependence of MOIN m 0
LNCH Length dependence of NCH m/cm3 0
LNFACTOR Length dependence of NFACTOR m 0
LNGATE Length dependence of NGATE m/cm3 0
LNLX Length dependence of NLX m2 0
LNOFF Length dependence of NOFF m 0
LNSUB Length dependence of NSUB m/cm3 0
LPCLM Length dependence of PCLM m 0
LPDIBLC1 Length dependence of PDIBLC1 m 0
LPDIBLC2 Length dependence of PDIBLC2 m 0
LPDIBLCB Length dependence of PDIBLCB m/V 0
LPRT Length dependence of PRT ˙ 𝜇m m 0
LPRWB Length dependence of PRWB m/V1/2 0
LPRWG Length dependence of PRWG m/V 0
LPSCBE1 Length dependence of PSCBE1 V 0
LPSCBE2 Length dependence of PSCBE2 V 0
LPVAG Length dependence of PVAG m 0
LRDSW Length dependence of RDSW ˙ 𝜇m m 0
LU0 Length dependence of U0 m/(Vcm2 s) 0
LUA Length dependence of UA m2 /V 0
LUA1 Length dependence of UA1 m2 /V 0
LUB Length dependence of UB m3 /V2 0
LUB1 Length dependence of UB1 m3 /V2 0

317
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
LUC Length dependence of UC m2 /V2 0
LUC1 Length dependence of UC1 m2 /(◦ CV2 ) 0
LUTE Length dependence of UTE m 0
LVBM Length dependence of VBM Vm 0
LVBX Length dependence of VBX Vm 0
LVFB Length dependence of VFB Vm 0
LVFBCV Length dependence of VFBCV Vm 0
LVOFF Length dependence of VOFF Vm 0
LVOFFCV Length dependence of VOFFCV Vm 0
LVSAT Length dependence of VSAT m2 /s 0
LVTH0 Length dependence of VTH0 Vm 0
LW0 Length dependence of W0 m2 0
LWR Length dependence of WR m 0
LXJ Length dependence of XJ m2 0
LXT Length dependence of XT m2 0
PA0 Cross-term dependence of A0 m2 0
PA1 Cross-term dependence of A1 m2 /V 0
PA2 Cross-term dependence of A2 m2 0
PACDE Cross-term dependence of ACDE m3 /V 0
PAGS Cross-term dependence of AGS m2 /V 0
PALPHA0 Cross-term dependence of ALPHA0 m3 /V 0
PALPHA1 Cross-term dependence of ALPHA1 m2 /V 0
PAT Cross-term dependence of AT m3 /s 0
PB0 Cross-term dependence of B0 m3 0
PB1 Cross-term dependence of B1 m3 0
PBETA0 Cross-term dependence of BETA0 Vm2 0
PCDSC Cross-term dependence of CDSC F 0
PCDSCB Cross-term dependence of CDSCB F/V 0
PCDSCD Cross-term dependence of CDSCD F/V 0
PCF Cross-term dependence of CF Fm 0
PCGDL Cross-term dependence of CGDL Fm 0
PCGSL Cross-term dependence of CGSL Fm 0
PCIT Cross-term dependence of CIT F 0
PCKAPPA Cross-term dependence of CKAPPA Fm 0
PCLC Cross-term dependence of CLC m3 0
PCLE Cross-term dependence of CLE m2 0

318
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
PDELTA Cross-term dependence of DELTA Vm2 0
PDROUT Cross-term dependence of DROUT m2 0
PDSUB Cross-term dependence of DSUB m2 0
PDVT0 Cross-term dependence of DVT0 m2 0
PDVT0W Cross-term dependence of DVT0W m 0
PDVT1 Cross-term dependence of DVT1 m2 0
PDVT1W Cross-term dependence of DVT1W m 0
PDVT2 Cross-term dependence of DVT2 m2 /V 0
PDVT2W Cross-term dependence of DVT2W m2 /V 0
PDWB Cross-term dependence of DWB m3 /V1/2 0
PDWG Cross-term dependence of DWG m3 /V1/2 0
PELM Cross-term dependence of ELM m2 0
PETA0 Cross-term dependence of ETA0 m2 0
PETAB Cross-term dependence of ETAB m2 /V 0
PGAMMA1 Cross-term dependence of GAMMA1 V1/2 m2 0
PGAMMA2 Cross-term dependence of GAMMA2 V1/2 m2 0
PK1 Cross-term dependence of K1 V1/2 m2 0
PK2 Cross-term dependence of K2 m2 0
PK3 Cross-term dependence of K3 m2 0
PK3B Cross-term dependence of K3B m2 /V 0
PKETA Cross-term dependence of KETA m2 /V 0
PKT1 Cross-term dependence of KT1 Vm2 0
PKT1L Cross-term dependence of KT1L Vm3 0
PKT2 Cross-term dependence of KT2 m2 0
PMOIN Cross-term dependence of MOIN m2 0
PNCH Cross-term dependence of NCH m2 /cm3 0
PNFACTOR Cross-term dependence of NFACTOR m2 0
PNGATE Cross-term dependence of NGATE m2 /cm3 0
PNLX Cross-term dependence of NLX m3 0
PNOFF Cross-term dependence of NOFF m2 0
PNSUB Cross-term dependence of NSUB m2 /cm3 0
PPCLM Cross-term dependence of PCLM m2 0
PPDIBLC1 Cross-term dependence of PDIBLC1 m2 0
PPDIBLC2 Cross-term dependence of PDIBLC2 m2 0
PPDIBLCB Cross-term dependence of PDIBLCB m2 /V 0
PPRT Cross-term dependence of PRT ˙ 𝜇m m2 0

319
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
PPRWB Cross-term dependence of PRWB m2 /V1/2 0
PPRWG Cross-term dependence of PRWG m2 /V 0
PPSCBE1 Cross-term dependence of PSCBE1 Vm 0
PPSCBE2 Cross-term dependence of PSCBE2 Vm 0
PPVAG Cross-term dependence of PVAG m2 0
PRDSW Cross-term dependence of RDSW ˙ 𝜇m m2 0
PU0 Cross-term dependence of U0 m2 /(Vcm2 s)0
PUA Cross-term dependence of UA m3 /V 0
PUA1 Cross-term dependence of UA1 m3 /V 0
PUB Cross-term dependence of UB m4 /V2 0
PUB1 Cross-term dependence of UB1 m4 /V2 0
PUC Cross-term dependence of UC m3 /V2 0
PUC1 Cross-term dependence of UC1 m3 /(◦ CV2 ) 0
PUTE Cross-term dependence of UTE m2 0
PVBM Cross-term dependence of VBM Vm2 0
PVBX Cross-term dependence of VBX Vm2 0
PVFB Cross-term dependence of VFB Vm2 0
PVFBCV Cross-term dependence of VFBCV Vm2 0
PVOFF Cross-term dependence of VOFF Vm2 0
PVOFFCV Cross-term dependence of VOFFCV Vm2 0
PVSAT Cross-term dependence of VSAT m3 /s 0
PVTH0 Cross-term dependence of VTH0 Vm2 0
PW0 Cross-term dependence of W0 m3 0
PWR Cross-term dependence of WR m2 0
PXJ Cross-term dependence of XJ m3 0
PXT Cross-term dependence of XT m3 0
WA0 Width dependence of A0 m 0
WA1 Width dependence of A1 m/V 0
WA2 Width dependence of A2 m 0
WACDE Width dependence of ACDE m2 /V 0
WAGS Width dependence of AGS m/V 0
WALPHA0 Width dependence of ALPHA0 m2 /V 0
WALPHA1 Width dependence of ALPHA1 m/V 0
WAT Width dependence of AT m2 /s 0
WB0 Width dependence of B0 m2 0
WB1 Width dependence of B1 m2 0

320
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
WBETA0 Width dependence of BETA0 Vm 0
WCDSC Width dependence of CDSC F/m 0
WCDSCB Width dependence of CDSCB F/(Vm) 0
WCDSCD Width dependence of CDSCD F/(Vm) 0
WCF Width dependence of CF F 0
WCGDL Width dependence of CGDL F 0
WCGSL Width dependence of CGSL F 0
WCIT Width dependence of CIT F/m 0
WCKAPPA Width dependence of CKAPPA F 0
WCLC Width dependence of CLC m2 0
WCLE Width dependence of CLE m 0
WDELTA Width dependence of DELTA Vm 0
WDROUT Width dependence of DROUT m 0
WDSUB Width dependence of DSUB m 0
WDVT0 Width dependence of DVT0 m 0
WDVT0W Width dependence of DVT0W – 0
WDVT1 Width dependence of DVT1 m 0
WDVT1W Width dependence of DVT1W – 0
WDVT2 Width dependence of DVT2 m/V 0
WDVT2W Width dependence of DVT2W m/V 0
WDWB Width dependence of DWB m2 /V1/2 0
WDWG Width dependence of DWG m2 /V1/2 0
WELM Width dependence of ELM m 0
WETA0 Width dependence of ETA0 m 0
WETAB Width dependence of ETAB m/V 0
WGAMMA1 Width dependence of GAMMA1 V1/2 m 0
WGAMMA2 Width dependence of GAMMA2 V1/2 m 0
WK1 Width dependence of K1 V1/2 m 0
WK2 Width dependence of K2 m 0
WK3 Width dependence of K3 m 0
WK3B Width dependence of K3B m/V 0
WKETA Width dependence of KETA m/V 0
WKT1 Width dependence of KT1 Vm 0
WKT1L Width dependence of KT1L Vm2 0
WKT2 Width dependence of KT2 m 0
WMOIN Width dependence of MOIN m 0

321
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
WNCH Width dependence of NCH m/cm3 0
WNFACTOR Width dependence of NFACTOR m 0
WNGATE Width dependence of NGATE m/cm3 0
WNLX Width dependence of NLX m2 0
WNOFF Width dependence of NOFF m 0
WNSUB Width dependence of NSUB m/cm3 0
WPCLM Width dependence of PCLM m 0
WPDIBLC1 Width dependence of PDIBLC1 m 0
WPDIBLC2 Width dependence of PDIBLC2 m 0
WPDIBLCB Width dependence of PDIBLCB m/V 0
WPRT Width dependence of PRT ˙ 𝜇m m 0
WPRWB Width dependence of PRWB m/V1/2 0
WPRWG Width dependence of PRWG m/V 0
WPSCBE1 Width dependence of PSCBE1 V 0
WPSCBE2 Width dependence of PSCBE2 V 0
WPVAG Width dependence of PVAG m 0
WRDSW Width dependence of RDSW ˙ 𝜇m m 0
WU0 Width dependence of U0 m/(Vcm2 s) 0
WUA Width dependence of UA m2 /V 0
WUA1 Width dependence of UA1 m2 /V 0
WUB Width dependence of UB m3 /V2 0
WUB1 Width dependence of UB1 m3 /V2 0
WUC Width dependence of UC m2 /V2 0
WUC1 Width dependence of UC1 m2 /(◦ CV2 ) 0
WUTE Width dependence of UTE m 0
WVBM Width dependence of VBM Vm 0
WVBX Width dependence of VBX Vm 0
WVFB Width dependence of VFB Vm 0
WVFBCV Width dependence of VFBCV Vm 0
WVOFF Width dependence of VOFF Vm 0
WVOFFCV Width dependence of VOFFCV Vm 0
WVSAT Width dependence of VSAT m2 /s 0
WVTH0 Width dependence of VTH0 Vm 0
WW0 Width dependence of W0 m2 0
WWR Width dependence of WR m 0
WXJ Width dependence of XJ m2 0

322
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
WXT Width dependence of XT m2 0
Doping Parameters
MJ Bulk p-n bottom grading coefficient – 0.5
MJSW Bulk p-n sidewall grading coefficient – 0.33
NSUB Substrate doping density cm−3 6e+16
Flicker and Thermal Noise Parameters
AF Flicker noise exponent – 1
EF Flicker exponent – 1
EM Saturation field Vm−1 4.1e+07
KF Flicker noise coefficient – 0
NOIA Noise parameter a – 0
NOIB Noise parameter b – 0
NOIC Noise parameter c – 0
Geometry Parameters
L Channel length m 5e-06
LL Coefficient of length dependence for length offset m 𝐿𝐿 𝑁 0
Coefficient of length dependence for CV channel
LLC m 𝐿𝐿 𝑁 0
length offset
LLN Power of length dependence for length offset – 0
LW Coefficient of width dependence for length offset m 𝐿𝑊 𝑁 0
Coefficient of width dependence for channel length
LWC m 𝐿𝑊 𝑁 0
offset
Coefficient of length and width cross term for length
LWL m 𝐿𝐿 𝑁 +𝐿𝑊 𝑁0
offset
Coefficient of length and width dependence for CV
LWLC m 𝐿𝐿 𝑁 +𝐿𝑊 𝑁0
channel length offset
LWN Power of width dependence for length offset – 0
TOX Gate oxide thickness m 1.5e-08
W Channel width m 5e-06
WL Coefficient of length dependence for width offset m𝑊 𝐿 𝑁 0
Coefficient of length dependence for CV channel
WLC m𝑊 𝐿 𝑁 0
width offset
WLN Power of length dependece of width offset – 0
WW Coefficient of width dependence for width offset m𝑊𝑊 𝑁 0
Coefficient of width dependence for CV channel
WWC m𝑊𝑊 𝑁 0
width offset

323
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
Coefficient of length and width cross term for width
WWL m𝑊 𝐿 𝑁 +𝑊𝑊 𝑁0
offset
Coefficient of length and width dependence for CV
WWLC m𝑊 𝐿 𝑁 +𝑊𝑊 𝑁0
channel width offset
WWN Power of width dependence of width offset – 0
XJ Junction depth m 1.5e-07
NQS Parameters
ELM Elmore constant of the channel – 5
Resistance Parameters
RSH Drain,source diffusion sheet resistance ˙ 0
Process Parameters
GAMMA1 Body effect coefficient near the surface V1/2 0
GAMMA2 Body effect coefficient in the bulk V1/2 0
JS Bulk p-n saturation current density A/m2 0.0001
NCH Channel doping concentration cm−3 1.7e+17
TOXM Gate oxide thickness used in extraction m 0
U0 Surface mobility 1/(Vcm2 s) 0
VBX Vbs at which the depetion region = XT V 0
XT Doping depth m 1.55e-07
Temperature Parameters
AT Temperature coefficient for saturation velocity m/s 33000
KT1 Temperature coefficient for threshold voltage V -0.11
Channel length dependence of the temerature
KT1L Vm 0
coefficient for the threshold voltage
Body-bias coefficient fo the threshold voltage
KT2 – 0.022
temperature effect
NJ Emission coefficient of junction – 1
PRT Temerature coefficient for RDSW ˙ 𝜇m 0
TCJ Temperature coefficient of Cj K−1 0
TCJSW Temperature coefficient of Cswj K−1 0
TCJSWG Temperature coefficient of Cjswg K−1 0
◦C Ambient
TNOM Nominal device temperature
Temperature
TPB Temperature coefficient of Pb V/K 0
TPBSW Temperature coefficient of Pbsw V/K 0
TPBSWG Temperature coefficient of Pbswg V/K 0
UA1 Temperature coefficient for UA m/V 4.31e-09

324
 Table 2-100. BSIM3 Device Model Parameters
Parameter Description Units Default
UB1 Temperature coefficient for UB m2 /V2 -7.61e-18
UC1 Temperature coefficient for UC m/(◦ CV2 ) 0
UTE Mobility temerature exponent – -1.5
XTI Junction current temperature exponent coefficient – 3
Voltage Parameters
PB Bulk p-n bottom potential V 1

325
2.3.20.6. Level 10 MOSFET Tables (BSIM-SOI)

For complete documentation of the BSIM-SOI model, see the users’ manual for the BSIM-SOI, available
for download at http://bsim.berkeley.edu/models/bsimsoi/. Xyce implements Version 3.2 of the
BSIM-SOI, you will have to get the documentation from the FTP archive on the Berkeley site.
In addition to the parameters shown in table 2-101, the BSIM3SOI supports a vector parameter for the
initial conditions. IC1 through IC5 may therefore be specified compactly as IC=<ic1>,<ic2>,<ic3>,
<ic4>,<ic5>.
NOTE: Many BSIM SOI parameters listed in tables 2-101 and 2-102 as having default values of zero
are actually replaced with internally computed defaults if not given. Specifying zero in your model
card will override this internal computation. It is recommended that you only set model parameters
that you are actually changing from defaults and that you not generate model cards containing
default values from the tables.

Table 2-101. BSIM3 SOI Device Instance Parameters

Parameter Description Units Default
logical
BJTOFF BJT on/off flag 0
(T/F)
logical
DEBUG BJT on/off flag 0
(T/F)
logical
TNODEOUT Flag indicating external temp node 0
(T/F)
logical
VLDEBUG false
(T/F)
Control Parameters
M Multiplier for M devices connected in parallel – 1
SIO model selector,SOIMOD=0:
SOIMOD BSIMPD,SOIMOD=1: undefined model for PD and – 0
FE,SOIMOD=2: ideal FD
DC Parameters
VBSUSR Vbs specified by user V 0
Geometry Parameters
AD Drain diffusion area m2 0
AEBCP Substrate to body overlap area for bc prasitics m2 0
AGBCP Gate to body overlap area for bc parasitics m2 0
AS Source diffusion area m2 0
FRBODY Layout dependent body-resistance coefficient – 1
L Channel length m 5e-06
NBC Number of body contact isolation edge – 0
NRB Number of squares in body – 1

326
 Table 2-101. BSIM3 SOI Device Instance Parameters
Parameter Description Units Default
Multiplier for RSH to yield parasitic resistance of
NRD □ 1
drain
Multiplier for RSH to yield parasitic resistance of
NRS □ 1
source
NSEG Number segments for width partitioning – 1
PD Drain diffusion perimeter m 0
PDBCP Perimeter length for bc parasitics at drain side m 0
PS Source diffusion perimeter m 0
PSBCP Perimeter length for bc parasitics at source side m 0
W Channel width m 5e-06
RF Parameters
RGATEMOD Gate resistance model selector – 0
Temperature Parameters
CTH0 Thermal capacitance F 0
RTH0 normalized thermal resistance ˙ 0
TEMP Device temperature ◦C 27
Voltage Parameters
IC1 Initial condition on Vds V 0
IC2 Initial condition on Vgs V 0
IC3 Initial condition on Vbs V 0
IC4 Initial condition on Ves V 0
IC5 Initial condition on Vps V 0
logical
OFF Initial condition of no voltage drops accross device false
(T/F)

Table 2-102. BSIM3 SOI Device Model Parameters

Parameter Description Units Default
The smoothing parameter in the Vox smoothing
DELTAVOX – 0
function
DTOXCV Delta oxide thickness in meters in CapMod3 m 0
FNOIMOD Flicker noise model selector – 1
IGBMOD Flicker noise model selector – 0
IGCMOD Gate-channel tunneling current model selector – 0
KB1 Scaling factor for backgate charge – 1
NOIF Floating body excess noise ideality factor – 1

327
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
NTNOI Thermal noise parameter – 1
POXEDGE Factor for the gate edge Tox – 1
RNOIA Thermal noise coefficient – 0.577
RNOIB Thermal noise coefficient – 0.37
RSHG Gate sheet resistance – 0.1
TNOIA Thermal noise parameter – 1.5
TNOIB Thermal noise parameter – 3.5
TNOIMOD Thermal noise model selector – 0
Lower bound of built-in potential lowering for FD
VBS0FD V 0.5
operation
Upper bound of built-in potential lowering for FD
VBS0PD – 0
operation
VOXH The limit of Vox in gate current calculation – 0
VTHO Threshold voltage – 0
Bin Parameters
LMAX Maximum channel length m 1
LMIN Minimum channel length m 0
WMAX Maximum channel width m 1
WMIN Minimum channel width m 0
Capacitance Parameters
Exponetial coefficient for charge thickness in capmod
ACDE m/V 1
= 3 for accumulation and depletion regions
ASD Sorce/Drain bottom diffusion smoothing parameter – 0.3
CF Firing field capacitance F/m 0
CGDL Light-doped drain-gate region overlap capacitance F/m 0
Non-LLD region drain-gate overlap capacitance per
CGDO F/m 0
unit channel length
Gate substrate overlap capacitance per unit channel
CGEO F/m 0
length
CGSL Light-doped source-gate region overlap capacitance F/m 0
Non-LLD region source-gate overlap capacitance per
CGSO F/m 0
unit channel length
Source/grain gate sidewall junction capacitance per
CJSWG F/m 1e-10
unit width
Coefficient for lightly doped region overlap
CKAPPA F/m 0.6
capacitance fireing field capacitance
CLC Constant term for short-channel model m 1e-08
CLE Exponetial term for the short-channel model – 0

328
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
Sorce/Drain sidewall fringing capacitance per unit
CSDESW F/m 0
length
CSDMIN Sorce/Drain bottom diffusion minimum capacitance V 0
DELVT Threshold voltage adjust for C-V V 0
DLBG Length offset fitting parameter for backgate charge m 0
DLC Length offset fitting parameter from C-V m 0
DLCB Length offset fitting parameter for body charge m 0
DWC Width offset fitting parameter from C-V m 0
FBODY Scaling factor for body charge – 1
Channel length dependency coefficient of diffusion
LDIF0 – 1
capacitance
Source/grain gate sidewall junction capacitance
MJSWG – 0.5
grading coeficient
Coefficient for the gate-bias dependent surface
MOIN – 15
potential
Power coefficient of channel length dependency for
NDIF – -1
diffusion capacitance
CV parameter in Vgsteff,CV for weak to strong
NOFF – 1
inversion
PBSWG Source/drain gate sidewall junction built-in potential V 0.7
TT Diffusion capacitance transit time coefficient s 1e-12
Sorce/Drain bottom diffusion capacitance flatband
VSDFB V 0
voltage
Sorce/Drain bottom diffusion capacitance threshold
VSDTH V 0
voltage
XPART Charge partitioning rate flag – 0
Control Parameters
BINUNIT Binning unit selector – 1
CAPMOD Flag for capacitance models – 2
MOBMOD Mobility model selector – 1
PARAMCHK Parameter value check – 0
SHMOD Flag for self-heating,0-no self-heating,1-self-heating – 0
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature
VERSION Version number – ’3.2’
Current Parameters
AIGC Parameter for Igc (F/g)1/2 s/mV0
AIGSD Parameter for Igs,d (F/g)1/2 s/mV0

329
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
BIGC Parameter for Igc (F/g)1/2 s/mV0
BIGSD Parameter for Igs,d (F/g)1/2 s/mV0
CIGC Parameter for Igc V−1 0
CIGSD Parameter for Igs,d V−1 0
DLCIG Delta L for Ig model V−1 0
NIGC Parameter for Igc slope – 1
PIGCD Parameter for Igc partition – 1
DC Parameters
A0 Bulk charge effect coefficient for channel length – 1
A1 First non-saturation effect parameter V−1 0
A2 Second non-saturation factor – 1
Channel length dependency of early voltage for
AELY Vm−1 0
bipolar current
AGIDL GIDL constant ˙ −1 0
AGS Gate-bias coefficient of abulk V−1 0
AHLI High level injection parameter for bipolar current – 0
ALPHA0 First parameter of impact-ionization current m/V 0
B0 Bulk charge effect coefficient for channel width m 0
B1 Bulk charge effect offset m 0
BETA0 Second parameter of impact-ionization current V 0
Second Vds dependent parameter of impact ionizatin
BETA1 – 0
current
Third Vds dependent parameter of impact ionizatin
BETA2 V 0.1
current
BGIDL GIDL exponential coefficient Vm−1 0
CDSC Drain/source to channel coupling capacitance F/m2 0.00024
CDSCB Body-bias sensitivity of CDSC F/(Vm2 ) 0
CDSCD Drain-bias sensitivity of CDSC F/(Vm2 ) 0
CIT Interface trap capacitance F/m2 0
DELTA Effective Vds parameter V 0.01
L-depedance Coefficient of the DIBL correction
DROUT – 0.56
parameter in Rout
DSUB DIBL coefficient exponent in subthreshhold region – 0
First coefficient of short-channel effect effect on
DVT0 – 2.2
threshold voltage
First coefficient of narrow-width effect effect on
DVT0W m−1 0
threshold voltage for small channel length

330
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
Second coefficient of short-channel effect effect on
DVT1 – 0.53
threshold voltage
Second coefficient of narrow-width effect effect on
DVT1W m−1 5.3e+06
threshold voltage for small channel length
Body-bias coefficient of short-channel effect effect on
DVT2 V−1 -0.032
threshold voltage
Body-bias coefficient of narrow-width effect effect on
DVT2W V−1 -0.032
threshold voltage for small channel length
DWB Coefficient of substrate body bias dependence of Weff m/V1/2 0
DWBC Width offset for body contact isolation edge m 0
DWG Coefficient of gate depedence of Weff m/V1/2 0
Saturation channel electric field for impact ionization
ESATII Vm−1 1e+07
current
ETA0 DIBL coefficient in subthreshold region – 0.08
ETAB Body-bias coefficient for the subthreshold DIBL effect V−1 -0.07
Fraction of bipolar current affecting the impact
FBJTII – 0
ionization
ISBJT BJT injection saturation current A/m2 1e-06
ISDIF BOdy to source/drain injection saturation current A/m2 0
ISREC Recombinatin in depletion saturation current A/m2 1e-05
ISTUN Reverse tunneling saturation current A/m2 0
K1 First-order body effect coefficient V1/2 0
K1W1 First body effect width depenent parameter m 0
K1W2 Second body effect width depenent parameter m 0
K2 second-order body effect coefficient – 0
K3 Narrow width coefficient – 0
K3B Body effect coefficient of K3 V−1 0
KETA Body-bias coefficient of bulk charge effect V−1 -0.6
KETAS Surface potential adjustment for bulk charge effect V 0
LBJT0 Reference channel length for bipolar current m 2e-07
Channel length dependent parameter at threshold for
LII – 0
impact ionization current
LINT Length of offset fiting parameter from I-V without bias m 0
LN Electron/hole diffusion length m 2e-06
NBJT Power coefficient of channel length – 1
NDIODE Diode non-ideality factor – 1
NFACTOR Subthreshold swing factor – 1
NGATE Poly gate doping concentration cm−3 0

331
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
NGIDL GIDL Vds enhancement coefficient V 1.2
NLX Lateral non-uniform doping parameter m 1.74e-07
NRECF0 Recombination non-ideality factor at foward bias – 2
NRECR0 Recombination non-ideality factor at reverse bias – 10
NTUN Reverse tunneling non-ideality factor – 10
PCLM Channel length modulation parameter – 1.3
First output resistance DIBL effect correction
PDIBLC1 – 0.39
parameter
Second output resistance DIBL effect correction
PDIBLC2 – 0.0086
parameter
PDIBLCB Body effect coefficient of DIBL correction parameter V−1 0
PRWB Body effect coefficient of RDSW V−1/2 0
PRWG Gate-bias effect coefficient of RDSW V−1 0
PVAG Gate dependence of early voltage – 0
RBODY Intrinsic body contact sheet resistance ˙/□ 0
RBSH Intrinsic body contact sheet resistance ˙/□ 0
RDSW Parasitic resistance per unit width ˙ 𝜇m 100
RHALO Body halo sheet resistance ˙/m 1e+15
First Vgs dependent parameter of impact ionizatin
SII0 V−1 0.5
current
Second Vgs dependent parameter of impact ionizatin
SII1 V−1 0.1
current
Third Vgs dependent parameter of impact ionizatin
SII2 – 0
current
Vds dependent parameter of drain saturation voltage
SIID V−1 0
for impact ionizatin current
Temperature dependent parameter for impact
TII – 0
ionization current
UA First-order mobility degradation coefficient m/V 2.25e-09
UB First-order mobility degradation coefficient m2 /V2 5.87e-19
UC Body effect of mobility degridation coefficient m/V2 0
VABJT Early voltage for bipolar current V 10
Maximum applied body-bias in threshold voltage
VBM V -3
calculation
Normal drain saturatio voltage at threshold for impact
VDSATII0 V 0.9
ionization current
Offset voltage in the subthreshold region at large W
VOFF V -0.08
and L

332
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
Voltage dependent parameter for recombination
VREC0 V 0
current
VSAT Saturation velocity at temp = TNOM m/s 80000
VTH0 Threshold voltage at Vbs = 0 for large L V 0
VTUN0 Voltage dependent parameter for tunneling current V 0
W0 Narrow-width paameter m 2.5e-06
WINT Width-offset fitting parameter from I-V without bias m 0
WR Width offset from Weff for Rds Calculation – 1
Dependency Parameters
LA0 Length dependence of A0 m 0
LA1 Length dependence of A1 m/V 0
LA2 Length dependence of A2 m 0
LACDE Length dependence of ACDE m2 /V 0
LAELY Length dependence of AELY V 0
LAGIDL Length dependence of AGIDL m/˙ 0
LAGS Length dependence of AGS m/V 0
LAHLI Length dependence of AHLI m 0
LAIGC Length dependence of AIGC (F/g)1/2 sm/mV
0
LAIGSD Length dependence of AIGSD (F/g)1/2 sm/mV
0
LALPHA0 Length dependence of ALPHA0 m2 /V 0
LALPHAGB1 Length dependence of ALPHAGB1 m/V 0
LALPHAGB2 Length dependence of ALPHAGB2 m/V 0
LAT Length dependence of AT m2 /s 0
LB0 Length dependence of B0 m2 0
LB1 Length dependence of B1 m2 0
LBETA0 Length dependence of BETA0 Vm 0
LBETA1 Length dependence of BETA1 m 0
LBETA2 Length dependence of BETA2 Vm 0
LBETAGB1 Length dependence of BETAGB1 m/V2 0
LBETAGB2 Length dependence of BETAGB2 m/V2 0
LBGIDL Length dependence of BGIDL V 0
LBIGC Length dependence of BIGC (F/g)1/2 sm/mV
0
LBIGSD Length dependence of BIGSD (F/g)1/2 sm/mV
0
LCDSC Length dependence of CDSC F/m 0
LCDSCB Length dependence of CDSCB F/(Vm) 0
LCDSCD Length dependence of CDSCD F/(Vm) 0

333
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
LCGDL Length dependence of CGDL F 0
LCGSL Length dependence of CGSL F 0
LCIGC Length dependence of CIGC m/V 0
LCIGSD Length dependence of CIGSD m/V 0
LCIT Length dependence of CIT F/m 0
LCKAPPA Length dependence of CKAPPA F 0
LDELTA Length dependence of DELTA Vm 0
LDELVT Length dependence of DELVT Vm 0
LDROUT Length dependence of DROUT m 0
LDSUB Length dependence of DSUB m 0
LDVT0 Length dependence of DVT0 m 0
LDVT0W Length dependence of DVT0W – 0
LDVT1 Length dependence of DVT1 m 0
LDVT1W Length dependence of DVT1W – 0
LDVT2 Length dependence of DVT2 m/V 0
LDVT2W Length dependence of DVT2W m/V 0
LDWB Length dependence of DWB m2 /V1/2 0
LDWG Length dependence of DWG m2 /V1/2 0
LESATII Length dependence of ESATII V 0
LETA0 Length dependence of ETA0 m 0
LETAB Length dependence of ETAB m/V 0
LFBJTII Length dependence of FBJTII m 0
LISBJT Length dependence of ISBJT A/m 0
LISDIF Length dependence of ISDIF A/m 0
LISREC Length dependence of ISREC A/m 0
LISTUN Length dependence of ISTUN A/m 0
LK1 Length dependence of K1 V1/2 m 0
LK1W1 Length dependence of K1W1 m2 0
LK1W2 Length dependence of K1W2 m2 0
LK2 Length dependence of K2 m 0
LK3 Length dependence of K3 m 0
LK3B Length dependence of K3B m/V 0
LKB1 Length dependence of KB1 m 0
LKETA Length dependence of KETA m/V 0
LKETAS Length dependence of KETAS Vm 0
LKT1 Length dependence of KT1 Vm 0

334
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
LKT1L Length dependence of KT1L Vm2 0
LKT2 Length dependence of KT2 m 0
LLBJT0 Length dependence of LBJT0 m2 0
LLII Length dependence of LII m 0
LMOIN Length dependence of MOIN m 0
LNBJT Length dependence of NBJT m 0
LNCH Length dependence of NCH m/cm3 0
LNDIF Length dependence of NDIF m 0
LNDIODE Length dependence of NDIODE m 0
LNFACTOR Length dependence of NFACTOR m 0
LNGATE Length dependence of NGATE m/cm3 0
LNGIDL Length dependence of NGIDL Vm 0
LNIGC Length dependence of NIGC m 0
LNLX Length dependence of NLX m2 0
LNOFF Length dependence of NOFF m 0
LNRECF0 Length dependence of NRECF0 m 0
LNRECR0 Length dependence of NRECR0 m 0
LNSUB Length dependence of NSUB m/cm3 0
LNTRECF Length dependence of NTRECF m 0
LNTRECR Length dependence of NTRECR m 0
LNTUN Length dependence of NTUN m 0
LPCLM Length dependence of PCLM m 0
LPDIBLC1 Length dependence of PDIBLC1 m 0
LPDIBLC2 Length dependence of PDIBLC2 m 0
LPDIBLCB Length dependence of PDIBLCB m/V 0
LPIGCD Length dependence of PIGCD m 0
LPOXEDGE Length dependence of POXEDGE m 0
LPRT Length dependence of PRT ˙ 𝜇m m 0
LPRWB Length dependence of PRWB m/V1/2 0
LPRWG Length dependence of PRWG m/V 0
LPVAG Length dependence of PVAG m 0
LRDSW Length dependence of RDSW ˙ 𝜇m m 0
LSII0 Length dependence of SII0 m/V 0
LSII1 Length dependence of SII1 m/V 0
LSII2 Length dependence of SII2 m 0
LSIID Length dependence of SIID m/V 0

335
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
LU0 Length dependence of U0 m/(Vcm2 s) 0
LUA Length dependence of UA m2 /V 0
LUA1 Length dependence of UA1 m2 /V 0
LUB Length dependence of UB m3 /V2 0
LUB1 Length dependence of UB1 m3 /V2 0
LUC Length dependence of UC m2 /V2 0
LUC1 Length dependence of UC1 m2 /(◦ CV2 ) 0
LUTE Length dependence of UTE m 0
LVABJT Length dependence of VABJT Vm 0
LVDSATII0 Length dependence of VDSATII0 Vm 0
LVOFF Length dependence of VOFF Vm 0
LVREC0 Length dependence of VREC0 Vm 0
LVSAT Length dependence of VSAT m2 /s 0
LVSDFB Length dependence of VSDFB Vm 0
LVSDTH Length dependence of VSDTH Vm 0
LVTH0 Length dependence of VTH0 Vm 0
LVTUN0 Length dependence of VTUN0 Vm 0
LW0 Length dependence of W0 m2 0
LWR Length dependence of WR m 0
LXBJT Length dependence of XBJT m 0
LXDIF Length dependence of XDIF m 0
LXJ Length dependence of XJ m2 0
LXRCRG1 Length dependence of XRCRG1 m 0
LXRCRG2 Length dependence of XRCRG2 m 0
LXREC Length dependence of XREC m 0
LXTUN Length dependence of XTUN m 0
PA0 Cross-term dependence of A0 m2 0
PA1 Cross-term dependence of A1 m2 /V 0
PA2 Cross-term dependence of A2 m2 0
PACDE Cross-term dependence of ACDE m3 /V 0
PAELY Cross-term dependence of AELY Vm 0
PAGIDL Cross-term dependence of AGIDL m 2 /˙ 0
PAGS Cross-term dependence of AGS m2 /V 0
PAHLI Cross-term dependence of AHLI m2 0
PAIGC Cross-term dependence of AIGC (F/g)1/2 sm2 /mV
0
PAIGSD Cross-term dependence of AIGSD (F/g)1/2 sm2 /mV
0

336
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
PALPHA0 Cross-term dependence of ALPHA0 m3 /V 0
PALPHAGB1 Cross-term dependence of ALPHAGB1 m2 /V 0
PALPHAGB2 Cross-term dependence of ALPHAGB2 m2 /V 0
PAT Cross-term dependence of AT m3 /s 0
PB0 Cross-term dependence of B0 m3 0
PB1 Cross-term dependence of B1 m3 0
PBETA0 Cross-term dependence of BETA0 Vm2 0
PBETA1 Cross-term dependence of BETA1 m2 0
PBETA2 Cross-term dependence of BETA2 Vm2 0
PBETAGB1 Cross-term dependence of BETAGB1 m2 /V2 0
PBETAGB2 Cross-term dependence of BETAGB2 m2 /V2 0
PBGIDL Cross-term dependence of BGIDL Vm 0
PBIGC Cross-term dependence of BIGC (F/g)1/2 sm2 /mV
0
PBIGSD Cross-term dependence of BIGSD (F/g)1/2 sm2 /mV
0
PCDSC Cross-term dependence of CDSC F 0
PCDSCB Cross-term dependence of CDSCB F/V 0
PCDSCD Cross-term dependence of CDSCD F/V 0
PCGDL Cross-term dependence of CGDL Fm 0
PCGSL Cross-term dependence of CGSL Fm 0
PCIGC Cross-term dependence of CIGC m2 /V 0
PCIGSD Cross-term dependence of CIGSD m2 /V 0
PCIT Cross-term dependence of CIT F 0
PCKAPPA Cross-term dependence of CKAPPA Fm 0
PDELTA Cross-term dependence of DELTA Vm2 0
PDELVT Cross-term dependence of DELVT Vm2 0
PDROUT Cross-term dependence of DROUT m2 0
PDSUB Cross-term dependence of DSUB m2 0
PDVT0 Cross-term dependence of DVT0 m2 0
PDVT0W Cross-term dependence of DVT0W m 0
PDVT1 Cross-term dependence of DVT1 m2 0
PDVT1W Cross-term dependence of DVT1W m 0
PDVT2 Cross-term dependence of DVT2 m2 /V 0
PDVT2W Cross-term dependence of DVT2W m2 /V 0
PDWB Cross-term dependence of DWB m3 /V1/2 0
PDWG Cross-term dependence of DWG m3 /V1/2 0
PESATII Cross-term dependence of ESATII Vm 0

337
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
PETA0 Cross-term dependence of ETA0 m2 0
PETAB Cross-term dependence of ETAB m2 /V 0
PFBJTII Cross-term dependence of FBJTII m2 0
PISBJT Cross-term dependence of ISBJT A 0
PISDIF Cross-term dependence of ISDIF A 0
PISREC Cross-term dependence of ISREC A 0
PISTUN Cross-term dependence of ISTUN A 0
PK1 Cross-term dependence of K1 V1/2 m2 0
PK1W1 Cross-term dependence of K1W1 m3 0
PK1W2 Cross-term dependence of K1W2 m3 0
PK2 Cross-term dependence of K2 m2 0
PK3 Cross-term dependence of K3 m2 0
PK3B Cross-term dependence of K3B m2 /V 0
PKB1 Cross-term dependence of KB1 m2 0
PKETA Cross-term dependence of KETA m2 /V 0
PKETAS Cross-term dependence of KETAS Vm2 0
PKT1 Cross-term dependence of KT1 Vm2 0
PKT1L Cross-term dependence of KT1L Vm3 0
PKT2 Cross-term dependence of KT2 m2 0
PLBJT0 Cross-term dependence of LBJT0 m3 0
PLII Cross-term dependence of LII m2 0
PMOIN Cross-term dependence of MOIN m2 0
PNBJT Cross-term dependence of NBJT m2 0
PNCH Cross-term dependence of NCH m2 /cm3 0
PNDIF Cross-term dependence of NDIF m2 0
PNDIODE Cross-term dependence of NDIODE m2 0
PNFACTOR Cross-term dependence of NFACTOR m2 0
PNGATE Cross-term dependence of NGATE m2 /cm3 0
PNGIDL Cross-term dependence of NGIDL Vm2 0
PNIGC Cross-term dependence of NIGC m2 0
PNLX Cross-term dependence of NLX m3 0
PNOFF Cross-term dependence of NOFF m2 0
PNRECF0 Cross-term dependence of NRECF0 m2 0
PNRECR0 Cross-term dependence of NRECR0 m2 0
PNSUB Cross-term dependence of NSUB m2 /cm3 0
PNTRECF Cross-term dependence of NTRECF m2 0

338
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
PNTRECR Cross-term dependence of NTRECR m2 0
PNTUN Cross-term dependence of NTUN m2 0
PPCLM Cross-term dependence of PCLM m2 0
PPDIBLC1 Cross-term dependence of PDIBLC1 m2 0
PPDIBLC2 Cross-term dependence of PDIBLC2 m2 0
PPDIBLCB Cross-term dependence of PDIBLCB m2 /V 0
PPIGCD Cross-term dependence of PIGCD m2 0
PPOXEDGE Cross-term dependence of POXEDGE m2 0
PPRT Cross-term dependence of PRT ˙ 𝜇m m2 0
PPRWB Cross-term dependence of PRWB m2 /V1/2 0
PPRWG Cross-term dependence of PRWG m2 /V 0
PPVAG Cross-term dependence of PVAG m2 0
PRDSW Cross-term dependence of RDSW ˙ 𝜇m m2 0
PSII0 Cross-term dependence of SII0 m2 /V 0
PSII1 Cross-term dependence of SII1 m2 /V 0
PSII2 Cross-term dependence of SII2 m2 0
PSIID Cross-term dependence of SIID m2 /V 0
PU0 Cross-term dependence of U0 m2 /(Vcm2 s)0
PUA Cross-term dependence of UA m3 /V 0
PUA1 Cross-term dependence of UA1 m3 /V 0
PUB Cross-term dependence of UB m4 /V2 0
PUB1 Cross-term dependence of UB1 m4 /V2 0
PUC Cross-term dependence of UC m3 /V2 0
PUC1 Cross-term dependence of UC1 m3 /(◦ CV2 ) 0
PUTE Cross-term dependence of UTE m2 0
PVABJT Cross-term dependence of VABJT Vm2 0
PVDSATII0 Cross-term dependence of VDSATII0 Vm2 0
PVOFF Cross-term dependence of VOFF Vm2 0
PVREC0 Cross-term dependence of VREC0 Vm2 0
PVSAT Cross-term dependence of VSAT m3 /s 0
PVSDFB Cross-term dependence of VSDFB Vm2 0
PVSDTH Cross-term dependence of VSDTH Vm2 0
PVTH0 Cross-term dependence of VTH0 Vm2 0
PVTUN0 Cross-term dependence of VTUN0 Vm2 0
PW0 Cross-term dependence of W0 m3 0
PWR Cross-term dependence of WR m2 0

339
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
PXBJT Cross-term dependence of XBJT m2 0
PXDIF Cross-term dependence of XDIF m2 0
PXJ Cross-term dependence of XJ m3 0
PXRCRG1 Cross-term dependence of XRCRG1 m2 0
PXRCRG2 Cross-term dependence of XRCRG2 m2 0
PXREC Cross-term dependence of XREC m2 0
PXTUN Cross-term dependence of XTUN m2 0
WA0 Width dependence of A0 m 0
WA1 Width dependence of A1 m/V 0
WA2 Width dependence of A2 m 0
WACDE Width dependence of ACDE m2 /V 0
WAELY Width dependence of AELY V 0
WAGIDL Width dependence of AGIDL m/˙ 0
WAGS Width dependence of AGS m/V 0
WAHLI Width dependence of AHLI m 0
WAIGC Width dependence of AIGC (F/g)1/2 sm/mV
0
WAIGSD Width dependence of AIGSD (F/g)1/2 sm/mV
0
WALPHA0 Width dependence of ALPHA0 m2 /V 0
WALPHAGB1 Width dependence of ALPHAGB1 m/V 0
WALPHAGB2 Width dependence of ALPHAGB2 m/V 0
WAT Width dependence of AT m2 /s 0
WB0 Width dependence of B0 m2 0
WB1 Width dependence of B1 m2 0
WBETA0 Width dependence of BETA0 Vm 0
WBETA1 Width dependence of BETA1 m 0
WBETA2 Width dependence of BETA2 Vm 0
WBETAGB1 Width dependence of BETAGB1 m/V2 0
WBETAGB2 Width dependence of BETAGB2 m/V2 0
WBGIDL Width dependence of BGIDL V 0
WBIGC Width dependence of BIGC (F/g)1/2 sm/mV
0
WBIGSD Width dependence of BIGSD (F/g)1/2 sm/mV
0
WCDSC Width dependence of CDSC F/m 0
WCDSCB Width dependence of CDSCB F/(Vm) 0
WCDSCD Width dependence of CDSCD F/(Vm) 0
WCGDL Width dependence of CGDL F 0
WCGSL Width dependence of CGSL F 0

340
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
WCIGC Width dependence of CIGC m/V 0
WCIGSD Width dependence of CIGSD m/V 0
WCIT Width dependence of CIT F/m 0
WCKAPPA Width dependence of CKAPPA F 0
WDELTA Width dependence of DELTA Vm 0
WDELVT Width dependence of DELVT Vm 0
WDROUT Width dependence of DROUT m 0
WDSUB Width dependence of DSUB m 0
WDVT0 Width dependence of DVT0 m 0
WDVT0W Width dependence of DVT0W – 0
WDVT1 Width dependence of DVT1 m 0
WDVT1W Width dependence of DVT1W – 0
WDVT2 Width dependence of DVT2 m/V 0
WDVT2W Width dependence of DVT2W m/V 0
WDWB Width dependence of DWB m2 /V1/2 0
WDWG Width dependence of DWG m2 /V1/2 0
WESATII Width dependence of ESATII V 0
WETA0 Width dependence of ETA0 m 0
WETAB Width dependence of ETAB m/V 0
WFBJTII Width dependence of FBJTII m 0
WISBJT Width dependence of ISBJT A/m 0
WISDIF Width dependence of ISDIF A/m 0
WISREC Width dependence of ISREC A/m 0
WISTUN Width dependence of ISTUN A/m 0
WK1 Width dependence of K1 V1/2 m 0
WK1W1 Width dependence of K1W1 m2 0
WK1W2 Width dependence of K1W2 m2 0
WK2 Width dependence of K2 m 0
WK3 Width dependence of K3 m 0
WK3B Width dependence of K3B m/V 0
WKB1 Width dependence of KB1 m 0
WKETA Width dependence of KETA m/V 0
WKETAS Width dependence of KETAS Vm 0
WKT1 Width dependence of KT1 Vm 0
WKT1L Width dependence of KT1L Vm2 0
WKT2 Width dependence of KT2 m 0

341
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
WLBJT0 Width dependence of LBJT0 m2 0
WLII Width dependence of LII m 0
WMOIN Width dependence of MOIN m 0
WNBJT Width dependence of NBJT m 0
WNCH Width dependence of NCH m/cm3 0
WNDIF Width dependence of NDIF m 0
WNDIODE Width dependence of NDIODE m 0
WNFACTOR Width dependence of NFACTOR m 0
WNGATE Width dependence of NGATE m/cm3 0
WNGIDL Width dependence of NGIDL Vm 0
WNIGC Width dependence of NIGC m 0
WNLX Width dependence of NLX m2 0
WNOFF Width dependence of NOFF m 0
WNRECF0 Width dependence of NRECF0 m 0
WNRECR0 Width dependence of NRECR0 m 0
WNSUB Width dependence of NSUB m/cm3 0
WNTRECF Width dependence of NTRECF m 0
WNTRECR Width dependence of NTRECR m 0
WNTUN Width dependence of NTUN m 0
WPCLM Width dependence of PCLM m 0
WPDIBLC1 Width dependence of PDIBLC1 m 0
WPDIBLC2 Width dependence of PDIBLC2 m 0
WPDIBLCB Width dependence of PDIBLCB m/V 0
WPIGCD Width dependence of PIGCD m 0
WPOXEDGE Width dependence of POXEDGE m 0
WPRT Width dependence of PRT ˙ 𝜇m m 0
WPRWB Width dependence of PRWB m/V1/2 0
WPRWG Width dependence of PRWG m/V 0
WPVAG Width dependence of PVAG m 0
WRDSW Width dependence of RDSW ˙ 𝜇m m 0
WSII0 Width dependence of SII0 m/V 0
WSII1 Width dependence of SII1 m/V 0
WSII2 Width dependence of SII2 m 0
WSIID Width dependence of SIID m/V 0
WU0 Width dependence of U0 m/(Vcm2 s) 0
WUA Width dependence of UA m2 /V 0

342
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
WUA1 Width dependence of UA1 m2 /V 0
WUB Width dependence of UB m3 /V2 0
WUB1 Width dependence of UB1 m3 /V2 0
WUC Width dependence of UC m2 /V2 0
WUC1 Width dependence of UC1 m2 /(◦ CV2 ) 0
WUTE Width dependence of UTE m 0
WVABJT Width dependence of VABJT Vm 0
WVDSATII0 Width dependence of VDSATII0 Vm 0
WVOFF Width dependence of VOFF Vm 0
WVREC0 Width dependence of VREC0 Vm 0
WVSAT Width dependence of VSAT m2 /s 0
WVSDFB Width dependence of VSDFB Vm 0
WVSDTH Width dependence of VSDTH Vm 0
WVTH0 Width dependence of VTH0 Vm 0
WVTUN0 Width dependence of VTUN0 Vm 0
WW0 Width dependence of W0 m2 0
WWR Width dependence of WR m 0
WXBJT Width dependence of XBJT m 0
WXDIF Width dependence of XDIF m 0
WXJ Width dependence of XJ m2 0
WXRCRG1 Width dependence of XRCRG1 m 0
WXRCRG2 Width dependence of XRCRG2 m 0
WXREC Width dependence of XREC m 0
WXTUN Width dependence of XTUN m 0
Doping Parameters
NSUB Substrate doping density cm−3 6e+16
Flicker and Thermal Noise Parameters
AF Flicker noise exponent – 1
EF Flicker exponent – 1
EM Saturation field Vm−1 4.1e+07
KF Flicker noise coefficient – 0
NOIA Noise parameter a – 0
NOIB Noise parameter b – 0
NOIC Noise parameter c – 8.75e+09
Geometry Parameters

343
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
L Channel length m 5e-06
LL Coefficient of length dependence for length offset m 𝐿𝐿 𝑁 0
Coefficient of length dependence for CV channel
LLC m 𝐿𝐿 𝑁 0
length offset
LLN Power of length dependence for length offset – 1
LW Coefficient of width dependence for length offset m 𝐿𝑊 𝑁 0
Coefficient of width dependence for channel length
LWC m 𝐿𝑊 𝑁 0
offset
Coefficient of length and width cross term for length
LWL m 𝐿𝐿 𝑁 +𝐿𝑊 𝑁0
offset
Coefficient of length and width dependence for CV
LWLC m 𝐿𝐿 𝑁 +𝐿𝑊 𝑁0
channel length offset
LWN Power of width dependence for length offset – 1
TOX Gate oxide thickness m 1e-08
W Channel width m 5e-06
WL Coefficient of length dependence for width offset m𝑊 𝐿 𝑁 0
Coefficient of length dependence for CV channel
WLC m𝑊 𝐿 𝑁 0
width offset
WLN Power of length dependece of width offset – 1
WW Coefficient of width dependence for width offset m𝑊𝑊 𝑁 0
Coefficient of width dependence for CV channel
WWC m𝑊𝑊 𝑁 0
width offset
Coefficient of length and width cross term for width
WWL m𝑊 𝐿 𝑁 +𝑊𝑊 𝑁0
offset
Coefficient of length and width dependence for CV
WWLC m𝑊 𝐿 𝑁 +𝑊𝑊 𝑁0
channel width offset
WWN Power of width dependence of width offset – 1
XJ Junction depth m 0
Resistance Parameters
RSH Drain,source diffusion sheet resistance ˙ 0
Process Parameters
GAMMA1 Body effect coefficient near the surface V1/2 0
GAMMA2 Body effect coefficient in the bulk V1/2 0
NCH Channel doping concentration cm−3 1.7e+17
TBOX Buried oxide thickness m 3e-07
TOXM Gate oxide thickness used in extraction m 0
TSI Silicon film thickness m 1e-07
U0 Surface mobility 1/(Vcm2 s) 0

344
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
VBX Vbs at which the depetion region = XT V 0
XT Doping depth m 1.55e-07
RF Parameters
BUG1830FIX Voltage limter fix for bug 1830 – 0
NGCON Number of gate contacts – 1
RGATEMOD Gate resistance model selector – 0
XGL Offset of the gate length due to variations in patterning m 0
XGW Distance from the gate contact to the channel edge m 0
Parameter for distributed channel resistance effect for
XRCRG1 – 12
intrinsic input resistance
Parameter to account for the excess channel diffusion
XRCRG2 – 1
resistance for intrinsic input resistance
Temperature Parameters
AT Temperature coefficient for saturation velocity m/s 33000
CTH0 Thermal capacitance per unit width F/m 1e-05
KT1 Temperature coefficient for threshold voltage V -0.11
Channel length dependence of the temerature
KT1L Vm 0
coefficient for the threshold voltage
Body-bias coefficient fo the threshold voltage
KT2 – 0.022
temperature effect
NTRECF Temperature coefficient for NRECF – 0
NTRECR Temperature coefficient for NRECR – 0
PRT Temerature coefficient for RDSW ˙ 𝜇m 0
RTH0 Thermal resistance per unit width ˙/m 0
TCJSWG Temperature coefficient of Cjswg K−1 0
◦C Ambient
TNOM Nominal device temperature
Temperature
TPBSWG Temperature coefficient of Pbswg V/K 0
UA1 Temperature coefficient for UA m/V 4.31e-09
UB1 Temperature coefficient for UB m2 /V2 -7.61e-18
UC1 Temperature coefficient for UC m/(◦ CV2 ) 0
UTE Mobility temerature exponent – -1.5
WTH0 Minimum width for thermal resistance calculation m 0
XBJT Power dependence of JBJT on temperature – 1
XDIF Power dependence of JDIF on temperature – 0
XREC Power dependence of JREC on temperature – 1
XTUN Power dependence of JTUN on temperature – 0

345
 Table 2-102. BSIM3 SOI Device Model Parameters
Parameter Description Units Default
Tunnelling Parameters
First Vox dependent parameter for gate current in
ALPHAGB1 V−1 0.35
inversion
First Vox dependent parameter for gate current in
ALPHAGB2 V−1 0.43
accumulation
Second Vox dependent parameter for gate current in
BETAGB1 V−2 0.03
inversion
First Vox dependent parameter for gate current in
BETAGB2 V−2 0.05
accumulation
EBG Effective bandgap in gate current calculation V 1.2
IGMOD Gate current model selector – 0
NTOX Power term of gate current – 1
TOXQM Oxide thickness for Igb calculation m 0
TOXREF Target oxide thickness m 2.5e-09
Vaux parameter for conduction band electron
VECB – 0.026
tunneling
VEVB Vaux parameter for valence band electron tunneling – 0.075
Third Vox dependent parameter for gate current in
VGB1 V 300
inversion
Third Vox dependent parameter for gate current in
VGB2 V 17
accumulation
Built-in Potential Lowering Parameters
Third backgate body effect parameter for short
DK2B – 0
channel effect
DVBD0 First short channel effect parameter in FD module – 0
DVBD1 Second short channel effect parameter in FD module – 0
K1B First backgate body effect parameter – 1
Second backgate body effect parameter for short
K2B – 0
channel effect
Gate bias dependance coefficient of surface potential
MOINFD – 1000
in FD module
NOFFFD Smoothing parameter in FD module – 1
SIO model selector,SOIMOD=0:
SOIMOD BSIMPD,SOIMOD=1: undefined model for PD and – 0
FE,SOIMOD=2: ideal FD
VBSA Offset voltage due to non-idealities V 0
VOFFFD Smoothing parameter in FD module V 0

346
2.3.20.7. Level 14/54 MOSFET Tables (BSIM4)

The level 14 MOSFET device in Xyce is based on the Berkeley BSIM4 model version 4.6.1. (For HSPICE
compatibility, the Xyce BSIM4 model can also be specified as level 54.) The model’s parameters are given
in the following tables. Note that the parameters have not all been properly categorized with units in place.
For complete documentation of the BSIM4 model, see the BSIM4 User’s Manual, available for download at
http://bsim.berkeley.edu/models/bsim4/.
Note that the BSIM4 device in Xyce now supports multiple versions selectable with the VERSION parameter
in the model card. At this time versions 4.6.1, 4.7.0, and 4.8.2 are supported. This version parameter may be
specified either in legacy text format (“4.6.1” or “4.8.2”) or in the CMC standard floating point format
(“4.61” or “4.82”).
If a VERSION parameter is not given, the latest version supported is used.
If a VERSION parameter is given that is not one of the supported version numbers, the closest matching
supported version is used instead and a warning given. If a version older than the lowest supported version
is chosen, the lowest supported version (4.6.1) is used and a warning given. If a model lower than version
4.7.0 is requested, version 4.6.1 is used (and a warning given). If a version newer than 4.7.0 but older than
4.8.0 is requested, 4.7.0 is used and a warning given. If a version 4.8.0 or later is requested, 4.8.2 is used
with an appropriate warning.
Specifying any model parameter that is not supported in the chosen version results in a warning and the
parameter being ignored. Parameters that are only valid for specific ranges of versions are noted as such in
the tables 2-103 and 2-104.
At this time, the BSIM4 is the only device in Xyce that supports multiple versions in this manner. All other
devices that have multiple version in Xyce are handled by having a unique level number for each version.
NOTE: Many BSIM4 parameters listed in tables 2-103 and 2-104 as having default values of zero are
actually replaced with internally computed defaults if not given. Specifying zero in your model card
will override this internal computation. It is recommended that you only set model parameters that
you are actually changing from defaults and that you not generate model cards containing default
values from the tables.
Furthermore, the value of FGIDL changed from 0 to 1 with version 4.8.2 of the BSIM4. This change is
NOT reflected in the table, which is generated automatically, and which shows only the default value
of this parameter that applies to versions 4.6.1 and 4.7.0.

Table 2-103. BSIM4 Device Instance Parameters

Parameter Description Units Default
AD Drain area – 0
AS Source area – 0
IC2 – 0
IC3 – 0
L Length – 5e-06
M Number of parallel copies – 1
MIN Minimize either D or S – 0

347
 Table 2-103. BSIM4 Device Instance Parameters
Parameter Description Units Default
NF Number of fingers – 1
NGCON Number of gate contacts – 0
OFF Device is initially off – false
PD Drain perimeter – 0
PS Source perimeter – 0
RBDB Body resistance – 0
RBPB Body resistance – 0
RBPD Body resistance – 0
RBPS Body resistance – 0
RBSB Body resistance – 0
SA distance between OD edge to poly of one side – 0
SB distance between OD edge to poly of the other side – 0
SC Distance to a single well edge – 0
Integral of the first distribution function for scattered
SCA – 0
well dopant
Integral of the second distribution function for
SCB – 0
scattered well dopant
Integral of the third distribution function for scattered
SCC – 0
well dopant
SD distance between neighbour fingers – 0
W Width – 5e-06
XGW Distance from gate contact center to device edge – 0
Basic Parameters
DELVT0 Zero bias threshold voltage variation V 0
DELVTO Zero bias threshold voltage variation V 0
Control Parameters
ACNQSMOD AC NQS model selector – 0
GEOMOD Geometry dependent parasitics model selector – 0
RBODYMOD Distributed body R model selector – 0
RGATEMOD Gate resistance model selector – 0
RGEOMOD S/D resistance and contact model selector – 0
TRNQSMOD Transient NQS model selector – 0
Temperature Parameters

◦C Ambient
TEMP Device temperature
Temperature
Voltage Parameters

348
 Table 2-103. BSIM4 Device Instance Parameters
Parameter Description Units Default
IC1 Vector of initial values: Vds,Vgs,Vbs V 0
Asymmetric and Bias-Dependent 𝑅 𝑑𝑠 Parameters
NRD Number of squares in drain – 1
NRS Number of squares in source – 1

Table 2-104. BSIM4 Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent – 1
AIGSD Parameter for Igs,d – 0.0136
AT Temperature coefficient of vsat – 33000
BIGSD Parameter for Igs,d – 0.00171
BVD Drain diode breakdown voltage – 10
BVS Source diode breakdown voltage – 10
CIGSD Parameter for Igs,d – 0.075
CJD Drain bottom junction capacitance per unit area – 0.0005
CJS Source bottom junction capacitance per unit area – 0.0005
CJSWD Drain sidewall junction capacitance per unit periphery – 5e-10
Drain (gate side) sidewall junction capacitance per
CJSWGD – 0
unit width
Source (gate side) sidewall junction capacitance per
CJSWGS – 0
unit width
Source sidewall junction capacitance per unit
CJSWS – 5e-10
periphery
DLCIG Delta L for Ig model – 0
DMCG Distance of Mid-Contact to Gate edge – 0
Distance of Mid-Contact to Gate edge in Test
DMCGT – 0
structures
DMCI Distance of Mid-Contact to Isolation – 0
DMDG Distance of Mid-Diffusion to Gate edge – 0
DWJ Delta W for S/D junctions – 0
EF Flicker noise frequency exponent – 1
EM Flicker noise parameter – 4.1e+07
EPSRGATE Dielectric constant of gate relative to vacuum – 11.7
GBMIN Minimum body conductance ˙ −1 1e-12
GIDLCLAMP
gidl clamp value – -1e-05
[Only for versions starting with 4.82]

349
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
IDOVVDS
noise clamping limit parameter – 1e-09
[Only for versions starting with 4.82]
IJTHDFWD Forward drain diode forward limiting current – 0.1
IJTHDREV Reverse drain diode forward limiting current – 0.1
IJTHSFWD Forward source diode forward limiting current – 0.1
IJTHSREV Reverse source diode forward limiting current – 0.1
Bottom drain junction reverse saturation current
JSD – 0.0001
density
Bottom source junction reverse saturation current
JSS – 0.0001
density
Isolation edge sidewall drain junction reverse
JSWD – 0
saturation current density
Gate edge drain junction reverse saturation current
JSWGD – 0
density
Gate edge source junction reverse saturation current
JSWGS – 0
density
Isolation edge sidewall source junction reverse
JSWS – 0
saturation current density
JTSD Drain bottom trap-assisted saturation current density – 0
JTSS Source bottom trap-assisted saturation current density – 0
Drain STI sidewall trap-assisted saturation current
JTSSWD – 0
density
Drain gate-edge sidewall trap-assisted saturation
JTSSWGD – 0
current density
Source gate-edge sidewall trap-assisted saturation
JTSSWGS – 0
current density
Source STI sidewall trap-assisted saturation current
JTSSWS – 0
density
JTWEFF
TAT current width dependence m 0
[Only for versions starting with 4.7]
K2WE K2 shift factor for well proximity effect – 0
K3B Body effect coefficient of k3 – 0
KF Flicker noise coefficient – 0
KT1 Temperature coefficient of Vth – -0.11
KT1L Temperature coefficient of Vth – 0
KT2 Body-coefficient of kt1 – 0.022
Mobility degradation/enhancement coefficient for
KU0 – 0
LOD
KU0WE Mobility degradation factor for well proximity effect – 0

350
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
Saturation velocity degradation/enhancement
KVSAT – 0
parameter for LOD
Threshold degradation/enhancement parameter for
KVTH0 – 0
LOD
KVTH0WE Threshold shift factor for well proximity effect – 0
LA0 Length dependence of a0 – 0
LA1 Length dependence of a1 – 0
LA2 Length dependence of a2 – 0
LACDE Length dependence of acde – 0
LAGIDL Length dependence of agidl – 0
LAGISL Length dependence of agisl – 0
LAGS Length dependence of ags – 0
LAIGBACC Length dependence of aigbacc – 0
LAIGBINV Length dependence of aigbinv – 0
LAIGC Length dependence of aigc – 0
LAIGD Length dependence of aigd – 0
LAIGS Length dependence of aigs – 0
LAIGSD Length dependence of aigsd – 0
LALPHA0 Length dependence of alpha0 – 0
LALPHA1 Length dependence of alpha1 – 0
LAT Length dependence of at – 0
LB0 Length dependence of b0 – 0
LB1 Length dependence of b1 – 0
LBETA0 Length dependence of beta0 – 0
LBGIDL Length dependence of bgidl – 0
LBGISL Length dependence of bgisl – 0
LBIGBACC Length dependence of bigbacc – 0
LBIGBINV Length dependence of bigbinv – 0
LBIGC Length dependence of bigc – 0
LBIGD Length dependence of bigd – 0
LBIGS Length dependence of bigs – 0
LBIGSD Length dependence of bigsd – 0
LCDSC Length dependence of cdsc – 0
LCDSCB Length dependence of cdscb – 0
LCDSCD Length dependence of cdscd – 0
LCF Length dependence of cf – 0

351
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
LCGDL Length dependence of cgdl – 0
LCGIDL Length dependence of cgidl – 0
LCGISL Length dependence of cgisl – 0
LCGSL Length dependence of cgsl – 0
LCIGBACC Length dependence of cigbacc – 0
LCIGBINV Length dependence of cigbinv – 0
LCIGC Length dependence of cigc – 0
LCIGD Length dependence of cigd – 0
LCIGS Length dependence of cigs – 0
LCIGSD Length dependence of cigsd – 0
LCIT Length dependence of cit – 0
LCKAPPAD Length dependence of ckappad – 0
LCKAPPAS Length dependence of ckappas – 0
LCLC Length dependence of clc – 0
LCLE Length dependence of cle – 0
LDELTA Length dependence of delta – 0
LDROUT Length dependence of drout – 0
LDSUB Length dependence of dsub – 0
LDVT0 Length dependence of dvt0 – 0
LDVT0W Length dependence of dvt0w – 0
LDVT1 Length dependence of dvt1 – 0
LDVT1W Length dependence of dvt1w – 0
LDVT2 Length dependence of dvt2 – 0
LDVT2W Length dependence of dvt2w – 0
LDVTP0 Length dependence of dvtp0 – 0
LDVTP1 Length dependence of dvtp1 – 0
LDVTP2
Length dependence of dvtp2 – 0
[Only for versions starting with 4.7]
LDVTP3
Length dependence of dvtp3 – 0
[Only for versions starting with 4.7]
LDVTP4
Length dependence of dvtp4 – 0
[Only for versions starting with 4.7]
LDVTP5
Length dependence of dvtp5 – 0
[Only for versions starting with 4.7]
LDWB Length dependence of dwb – 0
LDWG Length dependence of dwg – 0
LEGIDL Length dependence of egidl – 0

352
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
LEGISL Length dependence of egisl – 0
LEIGBINV Length dependence for eigbinv – 0
LETA0 Length dependence of eta0 – 0
LETAB Length dependence of etab – 0
LEU Length dependence of eu – 0
LFGIDL
Length dependence of fgidl – 0
[Only for versions starting with 4.7]
LFGISL
Length dependence of fgisl – 0
[Only for versions starting with 4.7]
LFPROUT Length dependence of pdiblcb – 0
LGAMMA1 Length dependence of gamma1 – 0
LGAMMA2 Length dependence of gamma2 – 0
LINTNOI lint offset for noise calculation – 0
LK1 Length dependence of k1 – 0
LK2 Length dependence of k2 – 0
LK2WE Length dependence of k2we – 0
LK3 Length dependence of k3 – 0
LK3B Length dependence of k3b – 0
LKETA Length dependence of keta – 0
LKGIDL
Length dependence of kgidl – 0
[Only for versions starting with 4.7]
LKGISL
Length dependence of kgisl – 0
[Only for versions starting with 4.7]
LKT1 Length dependence of kt1 – 0
LKT1L Length dependence of kt1l – 0
LKT2 Length dependence of kt2 – 0
LKU0 Length dependence of ku0 – 0
LKU0WE Length dependence of ku0we – 0
LKVTH0 Length dependence of kvth0 – 0
LKVTH0WE Length dependence of kvth0we – 0
LL Length reduction parameter – 0
LLAMBDA Length dependence of lambda – 0
LLC Length reduction parameter for CV – 0
LLN Length reduction parameter – 1
LLODKU0 Length parameter for u0 LOD effect – 0
LLODVTH Length parameter for vth LOD effect – 0
LLP Length dependence of lp – 0

353
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
LLPE0 Length dependence of lpe0 – 0
LLPEB Length dependence of lpeb – 0
LMAX Maximum length for the model – 1
LMIN Minimum length for the model – 0
LMINV Length dependence of minv – 0
LMINVCV Length dependence of minvcv – 0
LMOIN Length dependence of moin – 0
LNDEP Length dependence of ndep – 0
LNFACTOR Length dependence of nfactor – 0
LNGATE Length dependence of ngate – 0
LNIGBACC Length dependence of nigbacc – 0
LNIGBINV Length dependence of nigbinv – 0
LNIGC Length dependence of nigc – 0
LNOFF Length dependence of noff – 0
LNSD Length dependence of nsd – 0
LNSUB Length dependence of nsub – 0
LNTOX Length dependence of ntox – 0
LODETA0 eta0 shift modification factor for stress effect – 1
LODK2 K2 shift modification factor for stress effect – 1
LPCLM Length dependence of pclm – 0
LPDIBLC1 Length dependence of pdiblc1 – 0
LPDIBLC2 Length dependence of pdiblc2 – 0
LPDIBLCB Length dependence of pdiblcb – 0
LPDITS Length dependence of pdits – 0
LPDITSD Length dependence of pditsd – 0
LPHIN Length dependence of phin – 0
LPIGCD Length dependence for pigcd – 0
LPOXEDGE Length dependence for poxedge – 0
LPRT Length dependence of prt – 0
LPRWB Length dependence of prwb – 0
LPRWG Length dependence of prwg – 0
LPSCBE1 Length dependence of pscbe1 – 0
LPSCBE2 Length dependence of pscbe2 – 0
LPVAG Length dependence of pvag – 0
LRDSW Length dependence of rdsw – 0
LRDW Length dependence of rdw – 0

354
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
LRGIDL
Length dependence of rgidl – 0
[Only for versions starting with 4.7]
LRGISL
Length dependence of rgisl – 0
[Only for versions starting with 4.7]
LRSW Length dependence of rsw – 0
LTETA0
Length dependence of teta0 – 0
[Only for versions starting with 4.7]
LTNFACTOR
Length dependence of tnfactor – 0
[Only for versions starting with 4.7]
LTVFBSDOFF Length dependence of tvfbsdoff – 0
LTVOFF Length dependence of tvoff – 0
LTVOFFCV
Length dependence of tvoffcv – 0
[Only for versions starting with 4.7]
LU0 Length dependence of u0 – 0
LUA Length dependence of ua – 0
LUA1 Length dependence of ua1 – 0
LUB Length dependence of ub – 0
LUB1 Length dependence of ub1 – 0
LUC Length dependence of uc – 0
LUC1 Length dependence of uc1 – 0
LUCS
Length dependence of ucs – 0
[Only for versions starting with 4.7]
LUCSTE
Length dependence of ucste – 0
[Only for versions starting with 4.7]
LUD Length dependence of ud – 0
LUD1 Length dependence of ud1 – 0
LUP Length dependence of up – 0
LUTE Length dependence of ute – 0
LVBM Length dependence of vbm – 0
LVBX Length dependence of vbx – 0
LVFB Length dependence of vfb – 0
LVFBCV Length dependence of vfbcv – 0
LVFBSDOFF Length dependence of vfbsdoff – 0
LVOFF Length dependence of voff – 0
LVOFFCV Length dependence of voffcv – 0
LVSAT Length dependence of vsat – 0
LVTH0 – 0
LVTL Length dependence of vtl – 0

355
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
LW Length reduction parameter – 0
LW0 Length dependence of w0 – 0
LWC Length reduction parameter for CV – 0
LWL Length reduction parameter – 0
LWLC Length reduction parameter for CV – 0
LWN Length reduction parameter – 1
LWR Length dependence of wr – 0
LXJ Length dependence of xj – 0
LXN Length dependence of xn – 0
LXRCRG1 Length dependence of xrcrg1 – 0
LXRCRG2 Length dependence of xrcrg2 – 0
LXT Length dependence of xt – 0
MJD Drain bottom junction capacitance grading coefficient – 0.5
Source bottom junction capacitance grading
MJS – 0.5
coefficient
Drain sidewall junction capacitance grading
MJSWD – 0.33
coefficient
Drain (gate side) sidewall junction capacitance
MJSWGD – 0.33
grading coefficient
Source (gate side) sidewall junction capacitance
MJSWGS – 0.33
grading coefficient
Source sidewall junction capacitance grading
MJSWS – 0.33
coefficient
NGCON Number of gate contacts – 1
NJD Drain junction emission coefficient – 1
NJS Source junction emission coefficient – 1
NJTS Non-ideality factor for bottom junction – 20
NJTSD Non-ideality factor for bottom junction drain side – 20
NJTSSW Non-ideality factor for STI sidewall junction – 20
NJTSSWD Non-ideality factor for STI sidewall junction drain side – 20
NJTSSWG Non-ideality factor for gate-edge sidewall junction – 20
Non-ideality factor for gate-edge sidewall junction
NJTSSWGD – 20
drain side
NTNOI Thermal noise parameter – 1
PA0 Cross-term dependence of a0 – 0
PA1 Cross-term dependence of a1 – 0
PA2 Cross-term dependence of a2 – 0
PACDE Cross-term dependence of acde – 0

356
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
PAGIDL Cross-term dependence of agidl – 0
PAGISL Cross-term dependence of agisl – 0
PAGS Cross-term dependence of ags – 0
PAIGBACC Cross-term dependence of aigbacc – 0
PAIGBINV Cross-term dependence of aigbinv – 0
PAIGC Cross-term dependence of aigc – 0
PAIGD Cross-term dependence of aigd – 0
PAIGS Cross-term dependence of aigs – 0
PAIGSD Cross-term dependence of aigsd – 0
PALPHA0 Cross-term dependence of alpha0 – 0
PALPHA1 Cross-term dependence of alpha1 – 0
PAT Cross-term dependence of at – 0
PB0 Cross-term dependence of b0 – 0
PB1 Cross-term dependence of b1 – 0
PBD Drain junction built-in potential – 1
PBETA0 Cross-term dependence of beta0 – 0
PBGIDL Cross-term dependence of bgidl – 0
PBGISL Cross-term dependence of bgisl – 0
PBIGBACC Cross-term dependence of bigbacc – 0
PBIGBINV Cross-term dependence of bigbinv – 0
PBIGC Cross-term dependence of bigc – 0
PBIGD Cross-term dependence of bigd – 0
PBIGS Cross-term dependence of bigs – 0
PBIGSD Cross-term dependence of bigsd – 0
PBS Source junction built-in potential – 1
PBSWD Drain sidewall junction capacitance built in potential – 1
Drain (gate side) sidewall junction capacitance built in
PBSWGD – 0
potential
Source (gate side) sidewall junction capacitance built
PBSWGS – 0
in potential
PBSWS Source sidewall junction capacitance built in potential – 1
PCDSC Cross-term dependence of cdsc – 0
PCDSCB Cross-term dependence of cdscb – 0
PCDSCD Cross-term dependence of cdscd – 0
PCF Cross-term dependence of cf – 0
PCGDL Cross-term dependence of cgdl – 0

357
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
PCGIDL Cross-term dependence of cgidl – 0
PCGISL Cross-term dependence of cgisl – 0
PCGSL Cross-term dependence of cgsl – 0
PCIGBACC Cross-term dependence of cigbacc – 0
PCIGBINV Cross-term dependence of cigbinv – 0
PCIGC Cross-term dependence of cigc – 0
PCIGD Cross-term dependence of cigd – 0
PCIGS Cross-term dependence of cigs – 0
PCIGSD Cross-term dependence of cigsd – 0
PCIT Cross-term dependence of cit – 0
PCKAPPAD Cross-term dependence of ckappad – 0
PCKAPPAS Cross-term dependence of ckappas – 0
PCLC Cross-term dependence of clc – 0
PCLE Cross-term dependence of cle – 0
PDELTA Cross-term dependence of delta – 0
PDROUT Cross-term dependence of drout – 0
PDSUB Cross-term dependence of dsub – 0
PDVT0 Cross-term dependence of dvt0 – 0
PDVT0W Cross-term dependence of dvt0w – 0
PDVT1 Cross-term dependence of dvt1 – 0
PDVT1W Cross-term dependence of dvt1w – 0
PDVT2 Cross-term dependence of dvt2 – 0
PDVT2W Cross-term dependence of dvt2w – 0
PDVTP0 Cross-term dependence of dvtp0 – 0
PDVTP1 Cross-term dependence of dvtp1 – 0
PDVTP2
Cross-term dependence of dvtp2 – 0
[Only for versions starting with 4.7]
PDVTP3
Cross-term dependence of dvtp3 – 0
[Only for versions starting with 4.7]
PDVTP4
Cross-term dependence of dvtp4 – 0
[Only for versions starting with 4.7]
PDVTP5
Cross-term dependence of dvtp5 – 0
[Only for versions starting with 4.7]
PDWB Cross-term dependence of dwb – 0
PDWG Cross-term dependence of dwg – 0
PEGIDL Cross-term dependence of egidl – 0
PEGISL Cross-term dependence of egisl – 0

358
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
PEIGBINV Cross-term dependence for eigbinv – 0
PETA0 Cross-term dependence of eta0 – 0
PETAB Cross-term dependence of etab – 0
PEU Cross-term dependence of eu – 0
PFGIDL
Cross-term dependence of fgidl – 0
[Only for versions starting with 4.7]
PFGISL
Cross-term dependence of fgisl – 0
[Only for versions starting with 4.7]
PFPROUT Cross-term dependence of pdiblcb – 0
PGAMMA1 Cross-term dependence of gamma1 – 0
PGAMMA2 Cross-term dependence of gamma2 – 0
PHIG Work Function of gate – 4.05
PK1 Cross-term dependence of k1 – 0
PK2 Cross-term dependence of k2 – 0
PK2WE Cross-term dependence of k2we – 0
PK3 Cross-term dependence of k3 – 0
PK3B Cross-term dependence of k3b – 0
PKETA Cross-term dependence of keta – 0
PKGIDL
Cross-term dependence of kgidl – 0
[Only for versions starting with 4.7]
PKGISL
Cross-term dependence of kgisl – 0
[Only for versions starting with 4.7]
PKT1 Cross-term dependence of kt1 – 0
PKT1L Cross-term dependence of kt1l – 0
PKT2 Cross-term dependence of kt2 – 0
PKU0 Cross-term dependence of ku0 – 0
PKU0WE Cross-term dependence of ku0we – 0
PKVTH0 Cross-term dependence of kvth0 – 0
PKVTH0WE Cross-term dependence of kvth0we – 0
PLAMBDA Cross-term dependence of lambda – 0
PLP Cross-term dependence of lp – 0
PLPE0 Cross-term dependence of lpe0 – 0
PLPEB Cross-term dependence of lpeb – 0
PMINV Cross-term dependence of minv – 0
PMINVCV Cross-term dependence of minvcv – 0
PMOIN Cross-term dependence of moin – 0
PNDEP Cross-term dependence of ndep – 0

359
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
PNFACTOR Cross-term dependence of nfactor – 0
PNGATE Cross-term dependence of ngate – 0
PNIGBACC Cross-term dependence of nigbacc – 0
PNIGBINV Cross-term dependence of nigbinv – 0
PNIGC Cross-term dependence of nigc – 0
PNOFF Cross-term dependence of noff – 0
PNSD Cross-term dependence of nsd – 0
PNSUB Cross-term dependence of nsub – 0
PNTOX Cross-term dependence of ntox – 0
PPCLM Cross-term dependence of pclm – 0
PPDIBLC1 Cross-term dependence of pdiblc1 – 0
PPDIBLC2 Cross-term dependence of pdiblc2 – 0
PPDIBLCB Cross-term dependence of pdiblcb – 0
PPDITS Cross-term dependence of pdits – 0
PPDITSD Cross-term dependence of pditsd – 0
PPHIN Cross-term dependence of phin – 0
PPIGCD Cross-term dependence for pigcd – 0
PPOXEDGE Cross-term dependence for poxedge – 0
PPRT Cross-term dependence of prt – 0
PPRWB Cross-term dependence of prwb – 0
PPRWG Cross-term dependence of prwg – 0
PPSCBE1 Cross-term dependence of pscbe1 – 0
PPSCBE2 Cross-term dependence of pscbe2 – 0
PPVAG Cross-term dependence of pvag – 0
PRDSW Cross-term dependence of rdsw – 0
PRDW Cross-term dependence of rdw – 0
PRGIDL
Cross-term dependence of rgidl – 0
[Only for versions starting with 4.7]
PRGISL
Cross-term dependence of rgisl – 0
[Only for versions starting with 4.7]
PRSW Cross-term dependence of rsw – 0
PRT Temperature coefficient of parasitic resistance – 0
PTETA0
Cross-term dependence of teta0 – 0
[Only for versions starting with 4.7]
PTNFACTOR
Cross-term dependence of tnfactor – 0
[Only for versions starting with 4.7]
PTVFBSDOFF Cross-term dependence of tvfbsdoff – 0

360
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
PTVOFF Cross-term dependence of tvoff – 0
PTVOFFCV
Cross-term dependence of tvoffcv – 0
[Only for versions starting with 4.7]
PU0 Cross-term dependence of u0 – 0
PUA Cross-term dependence of ua – 0
PUA1 Cross-term dependence of ua1 – 0
PUB Cross-term dependence of ub – 0
PUB1 Cross-term dependence of ub1 – 0
PUC Cross-term dependence of uc – 0
PUC1 Cross-term dependence of uc1 – 0
PUCS
Cross-term dependence of ucs – 0
[Only for versions starting with 4.7]
PUCSTE
Cross-term dependence of ucste – 0
[Only for versions starting with 4.7]
PUD Cross-term dependence of ud – 0
PUD1 Cross-term dependence of ud1 – 0
PUP Cross-term dependence of up – 0
PUTE Cross-term dependence of ute – 0
PVAG Gate dependence of output resistance parameter – 0
PVBM Cross-term dependence of vbm – 0
PVBX Cross-term dependence of vbx – 0
PVFB Cross-term dependence of vfb – 0
PVFBCV Cross-term dependence of vfbcv – 0
PVFBSDOFF Cross-term dependence of vfbsdoff – 0
PVOFF Cross-term dependence of voff – 0
PVOFFCV Cross-term dependence of voffcv – 0
PVSAT Cross-term dependence of vsat – 0
PVTH0 – 0
PVTL Cross-term dependence of vtl – 0
PW0 Cross-term dependence of w0 – 0
PWR Cross-term dependence of wr – 0
PXJ Cross-term dependence of xj – 0
PXN Cross-term dependence of xn – 0
PXRCRG1 Cross-term dependence of xrcrg1 – 0
PXRCRG2 Cross-term dependence of xrcrg2 – 0
PXT Cross-term dependence of xt – 0
RBDB Resistance between bNode and dbNode ˙ 50

361
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
RBDBX0 Body resistance RBDBX scaling – 100
RBDBY0 Body resistance RBDBY scaling – 100
RBPB Resistance between bNodePrime and bNode ˙ 50
RBPBX0 Body resistance RBPBX scaling – 100
RBPBXL Body resistance RBPBX L scaling – 0
RBPBXNF Body resistance RBPBX NF scaling – 0
RBPBXW Body resistance RBPBX W scaling – 0
RBPBY0 Body resistance RBPBY scaling – 100
RBPBYL Body resistance RBPBY L scaling – 0
RBPBYNF Body resistance RBPBY NF scaling – 0
RBPBYW Body resistance RBPBY W scaling – 0
RBPD Resistance between bNodePrime and bNode ˙ 50
RBPD0 Body resistance RBPD scaling – 50
RBPDL Body resistance RBPD L scaling – 0
RBPDNF Body resistance RBPD NF scaling – 0
RBPDW Body resistance RBPD W scaling – 0
RBPS Resistance between bNodePrime and sbNode ˙ 50
RBPS0 Body resistance RBPS scaling – 50
RBPSL Body resistance RBPS L scaling – 0
RBPSNF Body resistance RBPS NF scaling – 0
RBPSW Body resistance RBPS W scaling – 0
RBSB Resistance between bNode and sbNode ˙ 50
RBSBX0 Body resistance RBSBX scaling – 100
RBSBY0 Body resistance RBSBY scaling – 100
RBSDBXL Body resistance RBSDBX L scaling – 0
RBSDBXNF Body resistance RBSDBX NF scaling – 0
RBSDBXW Body resistance RBSDBX W scaling – 0
RBSDBYL Body resistance RBSDBY L scaling – 0
RBSDBYNF Body resistance RBSDBY NF scaling – 0
RBSDBYW Body resistance RBSDBY W scaling – 0
RNOIA Thermal noise coefficient – 0.577
RNOIB Thermal noise coefficient – 0.5164
RNOIC
Thermal noise coefficient – 0.395
[Only for versions starting with 4.7]
Reference distance between OD edge to poly of one
SAREF – 1e-06
side

362
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
Reference distance between OD edge to poly of the
SBREF – 1e-06
other side
SCREF Reference distance to calculate SCA,SCB and SCC – 1e-06
STETA0 eta0 shift factor related to stress effect on vth – 0
STK2 K2 shift factor related to stress effect on vth – 0
TCJ Temperature coefficient of cj – 0
TCJSW Temperature coefficient of cjsw – 0
TCJSWG Temperature coefficient of cjswg – 0
TETA0
Temperature parameter for eta0 – 0
[Only for versions starting with 4.7]
TKU0 Temperature coefficient of KU0 – 0
TNFACTOR
Temperature parameter for nfactor – 0
[Only for versions starting with 4.7]
TNJTS Temperature coefficient for NJTS – 0
TNJTSD Temperature coefficient for NJTSD – 0
TNJTSSW Temperature coefficient for NJTSSW – 0
TNJTSSWD Temperature coefficient for NJTSSWD – 0
TNJTSSWG Temperature coefficient for NJTSSWG – 0
TNJTSSWGD Temperature coefficient for NJTSSWGD – 0
TNOIA Thermal noise parameter – 1.5
TNOIB Thermal noise parameter – 3.5
TNOIC
Thermal noise parameter – 0
[Only for versions starting with 4.7]
Ambient
TNOM Parameter measurement temperature –
Temperature
TPB Temperature coefficient of pb – 0
TPBSW Temperature coefficient of pbsw – 0
TPBSWG Temperature coefficient of pbswg – 0
TVFBSDOFF Temperature parameter for vfbsdoff – 0
TVOFF Temperature parameter for voff – 0
TVOFFCV
Temperature parameter for tvoffcv – 0
[Only for versions starting with 4.7]
UA1 Temperature coefficient of ua – 1e-09
UB1 Temperature coefficient of ub – -1e-18
UC1 Temperature coefficient of uc – 0
UCSTE
Temperature coefficient of colombic mobility – -0.004775
[Only for versions starting with 4.7]
UD1 Temperature coefficient of ud – 0

363
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
UTE Temperature coefficient of mobility – -1.5
Drain bottom trap-assisted voltage dependent
VTSD – 10
parameter
Source bottom trap-assisted voltage dependent
VTSS – 10
parameter
Drain STI sidewall trap-assisted voltage dependent
VTSSWD – 10
parameter
Drain gate-edge sidewall trap-assisted voltage
VTSSWGD – 10
dependent parameter
Source gate-edge sidewall trap-assisted voltage
VTSSWGS – 10
dependent parameter
Source STI sidewall trap-assisted voltage dependent
VTSSWS – 10
parameter
WA0 Width dependence of a0 – 0
WA1 Width dependence of a1 – 0
WA2 Width dependence of a2 – 0
WACDE Width dependence of acde – 0
WAGIDL Width dependence of agidl – 0
WAGISL Width dependence of agisl – 0
WAGS Width dependence of ags – 0
WAIGBACC Width dependence of aigbacc – 0
WAIGBINV Width dependence of aigbinv – 0
WAIGC Width dependence of aigc – 0
WAIGD Width dependence of aigd – 0
WAIGS Width dependence of aigs – 0
WAIGSD Width dependence of aigsd – 0
WALPHA0 Width dependence of alpha0 – 0
WALPHA1 Width dependence of alpha1 – 0
WAT Width dependence of at – 0
WB0 Width dependence of b0 – 0
WB1 Width dependence of b1 – 0
WBETA0 Width dependence of beta0 – 0
WBGIDL Width dependence of bgidl – 0
WBGISL Width dependence of bgisl – 0
WBIGBACC Width dependence of bigbacc – 0
WBIGBINV Width dependence of bigbinv – 0
WBIGC Width dependence of bigc – 0
WBIGD Width dependence of bigd – 0

364
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
WBIGS Width dependence of bigs – 0
WBIGSD Width dependence of bigsd – 0
WCDSC Width dependence of cdsc – 0
WCDSCB Width dependence of cdscb – 0
WCDSCD Width dependence of cdscd – 0
WCF Width dependence of cf – 0
WCGDL Width dependence of cgdl – 0
WCGIDL Width dependence of cgidl – 0
WCGISL Width dependence of cgisl – 0
WCGSL Width dependence of cgsl – 0
WCIGBACC Width dependence of cigbacc – 0
WCIGBINV Width dependence of cigbinv – 0
WCIGC Width dependence of cigc – 0
WCIGD Width dependence of cigd – 0
WCIGS Width dependence of cigs – 0
WCIGSD Width dependence of cigsd – 0
WCIT Width dependence of cit – 0
WCKAPPAD Width dependence of ckappad – 0
WCKAPPAS Width dependence of ckappas – 0
WCLC Width dependence of clc – 0
WCLE Width dependence of cle – 0
WDELTA Width dependence of delta – 0
WDROUT Width dependence of drout – 0
WDSUB Width dependence of dsub – 0
WDVT0 Width dependence of dvt0 – 0
WDVT0W Width dependence of dvt0w – 0
WDVT1 Width dependence of dvt1 – 0
WDVT1W Width dependence of dvt1w – 0
WDVT2 Width dependence of dvt2 – 0
WDVT2W Width dependence of dvt2w – 0
WDVTP0 Width dependence of dvtp0 – 0
WDVTP1 Width dependence of dvtp1 – 0
WDVTP2
Width dependence of dvtp2 – 0
[Only for versions starting with 4.7]
WDVTP3
Width dependence of dvtp3 – 0
[Only for versions starting with 4.7]

365
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
WDVTP4
Width dependence of dvtp4 – 0
[Only for versions starting with 4.7]
WDVTP5
Width dependence of dvtp5 – 0
[Only for versions starting with 4.7]
WDWB Width dependence of dwb – 0
WDWG Width dependence of dwg – 0
WEB Coefficient for SCB – 0
WEC Coefficient for SCC – 0
WEGIDL Width dependence of egidl – 0
WEGISL Width dependence of egisl – 0
WEIGBINV Width dependence for eigbinv – 0
WETA0 Width dependence of eta0 – 0
WETAB Width dependence of etab – 0
WEU Width dependence of eu – 0
WFGIDL
Width dependence of fgidl – 0
[Only for versions starting with 4.7]
WFGISL
Width dependence of fgisl – 0
[Only for versions starting with 4.7]
WFPROUT Width dependence of pdiblcb – 0
WGAMMA1 Width dependence of gamma1 – 0
WGAMMA2 Width dependence of gamma2 – 0
WK1 Width dependence of k1 – 0
WK2 Width dependence of k2 – 0
WK2WE Width dependence of k2we – 0
WK3 Width dependence of k3 – 0
WK3B Width dependence of k3b – 0
WKETA Width dependence of keta – 0
WKGIDL
Width dependence of kgidl – 0
[Only for versions starting with 4.7]
WKGISL
Width dependence of kgisl – 0
[Only for versions starting with 4.7]
WKT1 Width dependence of kt1 – 0
WKT1L Width dependence of kt1l – 0
WKT2 Width dependence of kt2 – 0
WKU0 Width dependence of ku0 – 0
WKU0WE Width dependence of ku0we – 0
WKVTH0 Width dependence of kvth0 – 0
WKVTH0WE Width dependence of kvth0we – 0

366
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
WL Width reduction parameter – 0
WLAMBDA Width dependence of lambda – 0
WLC Width reduction parameter for CV – 0
WLN Width reduction parameter – 1
WLOD Width parameter for stress effect – 0
WLODKU0 Width parameter for u0 LOD effect – 0
WLODVTH Width parameter for vth LOD effect – 0
WLP Width dependence of lp – 0
WLPE0 Width dependence of lpe0 – 0
WLPEB Width dependence of lpeb – 0
WMAX Maximum width for the model – 1
WMIN Minimum width for the model – 0
WMINV Width dependence of minv – 0
WMINVCV Width dependence of minvcv – 0
WMOIN Width dependence of moin – 0
WNDEP Width dependence of ndep – 0
WNFACTOR Width dependence of nfactor – 0
WNGATE Width dependence of ngate – 0
WNIGBACC Width dependence of nigbacc – 0
WNIGBINV Width dependence of nigbinv – 0
WNIGC Width dependence of nigc – 0
WNOFF Width dependence of noff – 0
WNSD Width dependence of nsd – 0
WNSUB Width dependence of nsub – 0
WNTOX Width dependence of ntox – 0
WPCLM Width dependence of pclm – 0
WPDIBLC1 Width dependence of pdiblc1 – 0
WPDIBLC2 Width dependence of pdiblc2 – 0
WPDIBLCB Width dependence of pdiblcb – 0
WPDITS Width dependence of pdits – 0
WPDITSD Width dependence of pditsd – 0
Flag for WPE model (WPEMOD=1 to activate this
WPEMOD – 0
model)
WPHIN Width dependence of phin – 0
WPIGCD Width dependence for pigcd – 0
WPOXEDGE Width dependence for poxedge – 0

367
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
WPRT Width dependence of prt – 0
WPRWB Width dependence of prwb – 0
WPRWG Width dependence of prwg – 0
WPSCBE1 Width dependence of pscbe1 – 0
WPSCBE2 Width dependence of pscbe2 – 0
WPVAG Width dependence of pvag – 0
WRDSW Width dependence of rdsw – 0
WRDW Width dependence of rdw – 0
WRGIDL
Width dependence of rgidl – 0
[Only for versions starting with 4.7]
WRGISL
Width dependence of rgisl – 0
[Only for versions starting with 4.7]
WRSW Width dependence of rsw – 0
WTETA0
Width dependence of teta0 – 0
[Only for versions starting with 4.7]
WTNFACTOR
Width dependence of tnfactor – 0
[Only for versions starting with 4.7]
WTVFBSDOFF Width dependence of tvfbsdoff – 0
WTVOFF Width dependence of tvoff – 0
WTVOFFCV
Width dependence of tvoffcv – 0
[Only for versions starting with 4.7]
WU0 Width dependence of u0 – 0
WUA Width dependence of ua – 0
WUA1 Width dependence of ua1 – 0
WUB Width dependence of ub – 0
WUB1 Width dependence of ub1 – 0
WUC Width dependence of uc – 0
WUC1 Width dependence of uc1 – 0
WUCS
Width dependence of ucs – 0
[Only for versions starting with 4.7]
WUCSTE
Width dependence of ucste – 0
[Only for versions starting with 4.7]
WUD Width dependence of ud – 0
WUD1 Width dependence of ud1 – 0
WUP Width dependence of up – 0
WUTE Width dependence of ute – 0
WVBM Width dependence of vbm – 0
WVBX Width dependence of vbx – 0

368
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
WVFB Width dependence of vfb – 0
WVFBCV Width dependence of vfbcv – 0
WVFBSDOFF Width dependence of vfbsdoff – 0
WVOFF Width dependence of voff – 0
WVOFFCV Width dependence of voffcv – 0
WVSAT Width dependence of vsat – 0
WVTH0 – 0
WVTL Width dependence of vtl – 0
WW Width reduction parameter – 0
WW0 Width dependence of w0 – 0
WWC Width reduction parameter for CV – 0
WWL Width reduction parameter – 0
WWLC Width reduction parameter for CV – 0
WWN Width reduction parameter – 1
WWR Width dependence of wr – 0
WXJ Width dependence of xj – 0
WXN Width dependence of xn – 0
WXRCRG1 Width dependence of xrcrg1 – 0
WXRCRG2 Width dependence of xrcrg2 – 0
WXT Width dependence of xt – 0
XGL Variation in Ldrawn – 0
XGW Distance from gate contact center to device edge – 0
XJBVD Fitting parameter for drain diode breakdown current – 1
XJBVS Fitting parameter for source diode breakdown current – 1
XL L offset for channel length due to mask/etch effect – 0
XRCRG1 First fitting parameter the bias-dependent Rg – 12
XRCRG2 Second fitting parameter the bias-dependent Rg – 1
XTID Drainjunction current temperature exponent – 3
XTIS Source junction current temperature exponent – 3
XTSD Power dependence of JTSD on temperature – 0.02
XTSS Power dependence of JTSS on temperature – 0.02
XTSSWD Power dependence of JTSSWD on temperature – 0.02
XTSSWGD Power dependence of JTSSWGD on temperature – 0.02
XTSSWGS Power dependence of JTSSWGS on temperature – 0.02
XTSSWS Power dependence of JTSSWS on temperature – 0.02
XW W offset for channel width due to mask/etch effect – 0

369
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
Basic Parameters
A0 Non-uniform depletion width effect coefficient. – 1
A1 Non-saturation effect coefficient V−1 0
A2 Non-saturation effect coefficient – 1
ADOS Charge centroid parameter – 1
AGS Gate bias coefficient of Abulk. V−1 0
B0 Abulk narrow width parameter m 0
B1 Abulk narrow width parameter m 0
BDOS Charge centroid parameter – 1
BG0SUB Band-gap of substrate at T=0K eV 1.16
CDSC Drain/Source and channel coupling capacitance F/m2 0.00024
CDSCB Body-bias dependence of cdsc F/(Vm2 ) 0
CDSCD Drain-bias dependence of cdsc F/(Vm2 ) 0
CIT Interface state capacitance F/m2 0
DELTA Effective Vds parameter V 0.01
DROUT DIBL coefficient of output resistance – 0.56
DSUB DIBL coefficient in the subthreshold region – 0
DVT0 Short channel effect coeff. 0 – 2.2
DVT0W Narrow Width coeff. 0 – 0
DVT1 Short channel effect coeff. 1 – 0.53
DVT1W Narrow Width effect coeff. 1 m−1 5.3e+06
DVT2 Short channel effect coeff. 2 V−1 -0.032
DVT2W Narrow Width effect coeff. 2 V−1 -0.032
DVTP0 First parameter for Vth shift due to pocket m 0
DVTP1 Second parameter for Vth shift due to pocket V−1 0
DVTP2
3rd parameter for Vth shift due to pocket Vm𝑋 0
[Only for versions starting with 4.7]
DVTP3
4th parameter for Vth shift due to pocket – 0
[Only for versions starting with 4.7]
DVTP4
5th parameter for Vth shift due to pocket V−1 0
[Only for versions starting with 4.7]
DVTP5
6th parameter for Vth shift due to pocket V 0
[Only for versions starting with 4.7]
DWB Width reduction parameter m/V1/2 0
DWG Width reduction parameter m/V 0
EASUB Electron affinity of substrate V 4.05
EPSRSUB Dielectric constant of substrate relative to vacuum – 11.7

370
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
ETA0 Subthreshold region DIBL coefficient – 0.08
ETAB Subthreshold region DIBL coefficient V−1 -0.07
EU Mobility exponent – 0
FPROUT Rout degradation coefficient for pocket devices V/m1/2 0
K1 Bulk effect coefficient 1 V−1/2 0
K2 Bulk effect coefficient 2 – 0
K3 Narrow width effect coefficient – 80
Body-bias coefficient of non-uniform depletion width
KETA V−1 -0.047
effect.
LAMBDA Velocity overshoot parameter – 0
LC back scattering parameter m 5e-09
LEFFEOT
Effective length for extraction of EOT m 1e-06
[Only for versions starting with 4.7]
LINT Length reduction parameter m 0
LP Channel length exponential factor of mobility m 1e-08
LPE0 Equivalent length of pocket region at zero bias m 1.74e-07
Equivalent length of pocket region accounting for
LPEB m 0
body bias
MINV Fitting parameter for moderate inversion in Vgsteff – 0
NFACTOR Subthreshold swing Coefficient – 1
NI0SUB Intrinsic carrier concentration of substrate at 300.15K cm−3 1.45e+10
PCLM Channel length modulation Coefficient – 1.3
PDIBLC1 Drain-induced barrier lowering coefficient – 0.39
PDIBLC2 Drain-induced barrier lowering coefficient – 0.0086
PDIBLCB Body-effect on drain-induced barrier lowering V−1 0
PDITS Coefficient for drain-induced Vth shifts V−1 0
PDITSD Vds dependence of drain-induced Vth shifts V−1 0
PDITSL Length dependence of drain-induced Vth shifts m−1 0
Adjusting parameter for surface potential due to
PHIN V 0
non-uniform vertical doping
PSCBE1 Substrate current body-effect coefficient Vm−1 4.24e+08
PSCBE2 Substrate current body-effect coefficient m/V 1e-05
TBGASUB First parameter of band-gap change due to temperature eV/K 0.000702
Second parameter of band-gap change due to
TBGBSUB K 1108
temperature
TEMPEOT
Temperature for extraction of EOT – 300.15
[Only for versions starting with 4.7]
U0 Low-field mobility at Tnom m2 /(Vs) 0

371
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
UA Linear gate dependence of mobility m/V 0
UB Quadratic gate dependence of mobility m2 /V2 1e-19
UC Body-bias dependence of mobility V−1 0
UCS
Colombic scattering exponent – 1.67
[Only for versions starting with 4.7]
UD Coulomb scattering factor of mobility m−2 0
UP Channel length linear factor of mobility m−2 0
VBM Maximum body voltage V -3
Voltage for extraction of equivalent gate oxide
VDDEOT V 1.5
thickness
VFB Flat Band Voltage V -1
VOFF Threshold voltage offset V -0.08
VOFFL Length dependence parameter for Vth offset V 0
VSAT Saturation velocity at tnom m/s 80000
VTH0 V 0
VTL thermal velocity m/s 200000
W0 Narrow width effect parameter m 2.5e-06
WEFFEOT
Effective width for extraction of EOT m 1e-05
[Only for versions starting with 4.7]
WINT Width reduction parameter m 0
XN back scattering parameter – 3
Capacitance Parameters
ACDE Exponential coefficient for finite charge thickness m/V 1
CF Fringe capacitance parameter F/m 0
CGBO Gate-bulk overlap capacitance per length – 0
CGDL New C-V model parameter F/m 0
CGDO Gate-drain overlap capacitance per width F/m 0
CGSL New C-V model parameter F/m 0
CGSO Gate-source overlap capacitance per width F/m 0
CKAPPAD D/G overlap C-V parameter V 0.6
CKAPPAS S/G overlap C-V parameter V 0.6
CLC Vdsat parameter for C-V model m 1e-07
CLE Vdsat parameter for C-V model – 0.6
DLC Delta L for C-V model m 0
DWC Delta W for C-V model m 0
MINVCV Fitting parameter for moderate inversion in Vgsteffcv – 0
MOIN Coefficient for gate-bias dependent surface potential – 15

372
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
NOFF C-V turn-on/off parameter – 1
VFBCV Flat Band Voltage parameter for capmod=0 only V -1
VOFFCV C-V lateral-shift parameter V 0
VOFFCVL Length dependence parameter for Vth offset in CV – 0
XPART Channel charge partitioning F/m 0
Control Parameters
ACNQSMOD AC NQS model selector – 0
BINUNIT Bin unit selector – 1
CAPMOD Capacitance model selector – 2
CVCHARGEMOD Capacitance charge model selector – 0
DIOMOD Diode IV model selector – 1
FNOIMOD Flicker noise model selector – 1
GEOMOD Geometry dependent parasitics model selector – 0
GIDLMOD
parameter for GIDL selector – 0
[Only for versions starting with 4.7]
IGBMOD Gate-to-body Ig model selector – 0
IGCMOD Gate-to-channel Ig model selector – 0
MOBMOD Mobility model selector – 0
MTRLCOMPATMOD
New material Mod backward compatibility selector – 0
[Only for versions starting with 4.7]
parameter for nonm-silicon substrate or metal gate
MTRLMOD – 0
selector
PARAMCHK Model parameter checking selector – 1
PERMOD Pd and Ps model selector – 1
RBODYMOD Distributed body R model selector – 0
RDSMOD Bias-dependent S/D resistance model selector – 0
RGATEMOD Gate R model selector – 0
RGEOMOD S/D resistance and contact model selector – 0
TEMPMOD Temperature model selector – 0
TNOIMOD Thermal noise model selector – 0
TRNQSMOD Transient NQS model selector – 0
VERSION parameter for model version – ’4.8.2’
Flicker and Thermal Noise Parameters
NOIA Flicker Noise parameter a – 0
NOIB Flicker Noise parameter b – 0
NOIC Flicker Noise parameter c – 0
Process Parameters

373
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
DTOX Defined as (toxe - toxp) m 0
EOT Equivalent gate oxide thickness in meters m 1.5e-09
Dielectric constant of the gate oxide relative to
EPSROX – 3.9
vacuum
GAMMA1 Vth body coefficient V1/2 0
GAMMA2 Vth body coefficient V1/2 0
NDEP Channel doping concentration at the depletion edge cm−3 1.7e+17
NGATE Poly-gate doping concentration cm−3 0
NSD S/D doping concentration cm−3 1e+20
NSUB Substrate doping concentration cm−3 6e+16
RSH Source-drain sheet resistance ˙/□ 0
RSHG Gate sheet resistance ˙/□ 0.1
TOXE Electrical gate oxide thickness in meters m 3e-09
TOXM Gate oxide thickness at which parameters are extracted m 3e-09
TOXP Physical gate oxide thickness in meters m 3e-09
VBX Vth transition body Voltage V 0
XJ Junction depth in meters m 1.5e-07
XT Doping depth m 1.55e-07
Tunnelling Parameters
AIGBACC Parameter for Igb (Fs2 /g)1/2 /m0.0136
AIGBINV Parameter for Igb (Fs2 /g)1/2 /m0.0111
AIGC Parameter for Igc (Fs2 /g)1/2 /m0.0136
AIGD Parameter for Igd (Fs2 /g)1/2 /m0.0136
AIGS Parameter for Igs (Fs2 /g)1/2 /m0.0136
BIGBACC Parameter for Igb (Fs2 /g)1/2 /mV
0.00171
BIGBINV Parameter for Igb (Fs2 /g)1/2 /mV
0.000949
BIGC Parameter for Igc (Fs2 /g)1/2 /mV
0.00171
BIGD Parameter for Igd (Fs2 /g)1/2 /mV
0.00171
BIGS Parameter for Igs (Fs2 /g)1/2 /mV
0.00171
CIGBACC Parameter for Igb V−1 0.075
CIGBINV Parameter for Igb V−1 0.006
CIGC Parameter for Igc V−1 0.075
CIGD Parameter for Igd V−1 0.075
CIGS Parameter for Igs V−1 0.075
DLCIGD Delta L for Ig model drain side m 0
EIGBINV Parameter for the Si bandgap for Igbinv V 1.1

374
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
NIGBACC Parameter for Igbacc slope – 1
NIGBINV Parameter for Igbinv slope – 3
NIGC Parameter for Igc slope – 1
NTOX Exponent for Tox ratio – 1
PIGCD Parameter for Igc partition – 1
POXEDGE Factor for the gate edge Tox – 1
TOXREF Target tox value m 3e-09
VFBSDOFF S/D flatband voltage offset V 0
Asymmetric and Bias-Dependent 𝑅 𝑑𝑠 Parameters
PRWB Body-effect on parasitic resistance V−1 0
PRWG Gate-bias effect on parasitic resistance V−1 1
RDSW Source-drain resistance per width ˙ 𝜇m 200
RDSWMIN Source-drain resistance per width at high Vg ˙ 𝜇m 0
RDW Drain resistance per width ˙ 𝜇m 100
RDWMIN Drain resistance per width at high Vg ˙ 𝜇m 0
RSW Source resistance per width ˙ 𝜇m 100
RSWMIN Source resistance per width at high Vg ˙ 𝜇m 0
WR Width dependence of rds – 1
Impact Ionization Current Parameters
ALPHA0 substrate current model parameter m/V 0
ALPHA1 substrate current model parameter V−1 0
BETA0 substrate current model parameter V−1 0
Gate-induced Drain Leakage Model Parameters
AGIDL Pre-exponential constant for GIDL ˙ −1 0
AGISL Pre-exponential constant for GISL ˙ −1 0
BGIDL Exponential constant for GIDL Vm−1 2.3e+09
BGISL Exponential constant for GISL Vm−1 2.3e-09
CGIDL Parameter for body-bias dependence of GIDL V3 0.5
CGISL Parameter for body-bias dependence of GISL V3 0.5
EGIDL Fitting parameter for Bandbending V 0.8
EGISL Fitting parameter for Bandbending V 0.8
FGIDL
GIDL vb parameter V 0
[Only for versions starting with 4.7]
FGISL
Parameter for GISL body bias dependence V 0
[Only for versions starting with 4.7]

375
 Table 2-104. BSIM4 Device Model Parameters
Parameter Description Units Default
KGIDL
GIDL vb parameter V 0
[Only for versions starting with 4.7]
KGISL
Parameter for GISL body bias dependence V 0
[Only for versions starting with 4.7]
RGIDL
GIDL vg parameter – 1
[Only for versions starting with 4.7]
RGISL
Parameter for GISL gate bias dependence – 1
[Only for versions starting with 4.7]

376
2.3.20.8. Level 18 MOSFET Tables (VDMOS)

The vertical double-diffused power MOSFET model is based on the uniform charge control model (UCCM)
developed at Rensselaer Polytechnic Institute [15]. The VDMOS current-voltage characteristics are
described by a single, continuous analytical expression for all regimes of operation. The physics-based
model includes effects such as velocity saturation in the channel, drain induced barrier lowering, finite
output conductance in saturation, the quasi-saturation effect through a bias dependent drain parasitic
resistance, effects of bulk charge, and bias dependent low-field mobility. An important feature of the
implementation is the utilization of a single continuous expression for the drain current, which is valid
below and above threshold, effectively removing discontinuities and improving convergence properties.
The following tables give parameters for the level 18 MOSFET.

Table 2-105. Power MOSFET Device Instance Parameters

Parameter Description Units Default
AD Drain diffusion area m2 0
AS Source diffusion area m2 0
L Channel length m 0
M Multiplier for M devices connected in parallel – 1
Multiplier for RSH to yield parasitic resistance of
NRD □ 1
drain
Multiplier for RSH to yield parasitic resistance of
NRS □ 1
source
PD Drain diffusion perimeter m 0
PS Source diffusion perimeter m 0
◦C Ambient
TEMP Device temperature
Temperature
W Channel width m 0

Table 2-106. Power MOSFET Device Model Parameters

Parameter Description Units Default
AI – 2e+09
Parameter accounting for the threshold dependence on
ALPHA – 0
the channel potential
ARTD – 0
BI – 8e+08
BRTD – 0.035
CBD Zero-bias bulk-drain p-n capacitance F 0
CBS Zero-bias bulk-source p-n capacitance F 0
CGBO Gate-bulk overlap capacitance/channel length F/m 0
CGDO Gate-drain overlap capacitance/channel width F/m 0

377
 Table 2-106. Power MOSFET Device Model Parameters
Parameter Description Units Default
CGSO Gate-source overlap capacitance/channel width F/m 0
CJ Bulk p-n zero-bias bottom capacitance/area F/m2 0
CJSW Bulk p-n zero-bias sidewall capacitance/area F/m2 0
CRTD – 0.1472
CV Charge model storage selector – 1
CVE Meyer-like capacitor model selector – 1
D1AF Drain-source diode flicker noise exponent – 1
D1BV Drain-source diode reverse breakdown voltage V 1e+99
D1CJO Drain-source diode junction capacitance F 0
D1EG Drain-source diode activation energy eV 1.11
D1FC Drain-source diode forward bias depletion capacitance – 0.5
D1IBV Drain-source diode current at breakdown voltage A 0.001
D1IKF Drain-source diode high injection knee currrent A 0
D1IS Drain-Source diode saturation current A 1e-14
D1ISR Drain-source diode recombination saturation current A 0
D1KF Drain-source diode flicker noise coefficient – 0
D1M Drain-source diode grading coefficient – 0.5
D1N Drain-source diode emission coefficient – 1
Drain-source diode recombination emission
D1NR – 2
coefficient
D1RS Drain-source diode ohmic resistance ˙ 0
D1TNOM Drain-source diode nominal temperature ◦C 300.15
D1TT Drain-source diode transit time s 0
D1VJ Drain-source diode junction potential V 1
D1XTI Drain-source diode sat. current temperature exponent – 3
DELMAX – 0.9
DELTA Transition width parameter – 5
DRIFTPARAMA Drift region resistance intercept parameter ˙ 0.08
DRIFTPARAMB Drift region resistance slope parameter ˙ V −1 0.013
DRTD – 0.0052
ETA Subthreshold ideality factor – 1.32
Coefficient for forward-bias depletion capacitance
FC – 0.5
formula
FPE Charge partitioning scheme selector – 1
GAMMAL0 Body effect constant in front of linear term – 0
GAMMAS0 Body effect constant in front of square root term V−1/2 0.5

378
 Table 2-106. Power MOSFET Device Model Parameters
Parameter Description Units Default
IS Bulk p-n saturation current A 1e-14
ISUBMOD – 0
JS Bulk p-n saturation current density A/m2 0
K – 0
KVS – 0
KVT – 0
L0 Gate length of nominal device m 0
LAMBDA Output conductance parameter V−1 0.048
LD Lateral diffusion length m 0
LGAMMAL Sensitivity of gL on device length – 0
LGAMMAS Sensitivity of gS on device length V−1/2 0
LS – 3.5e-08
M Knee shape parameter – 4
MC – 3
Transition width parameter used by the charge
MCV – 10
partitioning scheme
MD – 2
MDTEMP – 0
MJ Bulk p-n bottom grading coefficient – 0.5
MJSW Bulk p-n sidewall grading coefficient – 0.5
MTH – 0
N2 – 1
NRTD – 0.115
NSS Surface state density cm−2 0
NSUB Substrate doping density cm−3 0
PB Bulk p-n bottom potential V 0.8
PHI Surface potential V 0.6
RD Drain ohmic resistance ˙ 0
RDSSHUNT Drain-source shunt resistance ˙ 0
RG Gate ohmic resistance ˙ 0
RS Source ohmic resistance ˙ 0
RSH Drain,source diffusion sheet resistance ˙ 0
RSUB – 0
SIGMA0 DIBL parameter – 0.048
Specifies the type of parameter interpolation over
TEMPMODEL – ’NONE’
temperature

379
 Table 2-106. Power MOSFET Device Model Parameters
Parameter Description Units Default
THETA Mobility degradation parameter m/V 0
◦C Ambient
TNOM Nominal device temperature
Temperature
TOX Gate oxide thickness m 1e-07
Gate material type (-1 = same as substrate, 0 =
TPG – 1
aluminum,1 = opposite of substrate)
TS – 0
TVS – 0
U0 Surface mobility 1/(Vcm2 s) 280
UO Surface mobility 1/(Vcm2 s) 280
VFB Flat band voltage V 0
VMAX Maximum drift velocity for carriers m/s 40000
VP – 0
VSIGMA DIBL parameter V 0.2
VSIGMAT DIBL parameter V 1.7
VTO Zero-bias threshold voltage V 0
W0 Gate width of nominal device m 0
WGAMMAL Sensitivity of gL on device width – 0
WGAMMAS Sensitivity of gS on device width V−1/2 0
XJ Metallurgical junction depth m 0
XQC Charge partitioning factor – 0.6

380
2.3.20.9. Levels 70 and 70450 MOSFET Tables (BSIM-SOI 4.6.1 and 4.5.0)

For complete documentation of the BSIM-SOI model, see the users’ manual for the BSIM-SOI, available
for download at http://bsim.berkeley.edu/models/bsimsoi/. Xyce implements Version 4.6.1 of the
BSIM-SOI as the level 70 device and version 4.5.0 as level 70450.
Instance and model parameters of the level 70 MOSFET are given in tables 2-107 and 2-108.
Beginning with Xyce 7.2, the BSIM-SOI models level 70 and 70450 have limited support for the optional
5th, 6th, and 7th nodes. See the BSIM-SOI technical manual at the BSIM web site for details of what
configurations the full device supports. Only some of these use cases are supported: Use of the BSIM-SOI
4.x with TNODEOUT=0 (the default) is supported in 4-, 5-, 6-, and 7-node configurations. TNODEOUT=1 is
supported only in the 7-node configuration, with the 7th node being temperature. No access to the external
temperature node is available in 5- or 6- node configuration.
When TNODEOUT=0, the temperature node is an internal node of the device even when not specified on the
instance line, and its value may still be printed using the N() notation (see section 2.1.31.12). This
somewhat minimizes the impact of the lack of support for TNODEOUT=1 in Xyce — the temperature rise due
to self-heating is always available for printing, but it is not available for creation of a thermal coupling
network except in the 7-node configuration.
Note that with some choices of model parameters, the BSIM-SOI devices attempt to “collapse” the “P” and
“B” nodes (external and internal body nodes, 5th and 6th netlist nodes if given, internal nodes if not given).
Xyce is unable to perform such collapse when the nodes are externally specified, and will issue warnings
when it finds the model trying to do so. Depending on the actual nodes used for P and B, the device may fail
to converge or produce invalid results; as an example, if P and B are actually specified on the netlist line to
be the same node, this failure to collapse will not matter — the nodes are already the same. But if two
different node names are used for the 5th and 6th nodes, the failure to collapse will leave one node floating
and the simulation will likely fail if the printed warnings are ignored.
A similar problem exists for other choices of model parameter: in some cases neither the “P” nor “B” nodes
are used, and if the nodes are specified on the netlist line the BSIM-SOI code attempts to collapse them to
ground. This is not something Xyce can do, and therefore instead Xyce ignores the specified nodes. This can
leave those nodes floating and lead to convergence failures unless the specified nodes are already the
ground node (node 0). Xyce will issue appropriate warnings when this condition exists and suggest removal
of the unused external nodes from the instance line.
The BSIM SOI 4.6.1 device supports output of the internal variables in table 2-109 on the .PRINT line of a
netlist. To access them from a print line, use the syntax N(<instance>:<variable>) where
“<instance>” refers to the name of the specific level 70 M device in your netlist.
NOTE: It has been observed that the gate capacitance model of BSIM-SOI 4.6.1 behaves differently than
earlier versions, and the team has seen significant disagreement of gate currents when comparing identical
simulations with other simulators that have only earlier BSIM-SOI models. For this reason, we are also
providing BSIM-SOI 4.5.0 as the level 70450 MOSFET. This model does agree with these other simulators.
The parameters and output variables are given in tables 2-110, 2-111, and 2-112. Unlike BSIM-SOI 4.6.1,
the 4.5.0 model’s original Verilog-A source code does not contain descriptions and units for the parameters,
and these appear blank in the tables. For descriptions and units, see the corresponding parameters in the
level 70 tables.

381
 Table 2-107. BSIM-SOI 4.6.1 Device Instance Parameters
Parameter Description Units Default
AD Drain area m2 0
AEBCP Substrate to body overlap area for bc parasitics m2 0
AGBCP Gate to body overlap area for bc parasitics m2 0
Parasitic Gate to body overlap area for bc parasitics /*
AGBCP2 m2 0
v4.1 improvement on BC
AGBCPD Gate to body overlap area for bc parasitics in DC m2 0
AS Source area m2 0
BJTOFF BJT on/off flag — 0
CTH0 Instance Thermal Capacitance — 1e-05
DEBUG DEBUG on/off flag — 0
DELVTO Zero bias threshold voltage variation V 0
DTEMP device temperature offset from ambient — 0
FRBODY layout dependent body-resistance coefficient — 1
L Length m 5e-06
M multiplicity factor — 1
NBC Number of body contact isolation edge — 0
NF Number of fingers — 1
NRB Number of squares in body — 1
NRD Number of squares in drain — 1
NRS Number of squares in source — 1
NSEG Number segments for width partitioning — 1
OFF Device is initially off — 0
PD Drain perimeter m 0
PDBCP Perimeter length for bc parasitics at drain side m 0
PS Source perimeter m 0
PSBCP Perimeter length for bc parasitics at source side m 0
RBDB Body resistance — 50
RBSB Body resistance — 50
RTH0 Instance Thermal Resistance — 0
SA distance between OD edge to poly of one side m 0
SB distance between OD edge to poly of the other side m 0
SD distance between neighbor fingers m 0
SHMOD Self heating mode selector — 0
SOIMOD Instance model selector for PD/FD operation /* v3.2 — 0
TNODEOUT Flag indicating external temp node — 0
W Width m 5e-06

382
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
A0 Non-uniform depletion width effect coefficient. — 1
A1 Non-saturation effect coefficient — 0
A2 Non-saturation effect coefficient — 1
ABJTII Exponent factor for avalanche current — 0
Exponential coefficient for charge thickness in
ACDE m/V 1
capMod=3 for accumulation and depletion regions
AD Drain area m2 0
ADOS Charge centroid parameter — 1
AEBCP Substrate to body overlap area for bc parasitics m2 0
Channel length dependency of early voltage for
AELY — 0
bipolar current
AF Flicker noise exponent — 1
AGB1 ’A’ for Igb1 Tunneling current model — 3.7622e-07
AGB2 ’A’ for Igb2 Tunneling current model — 4.9758e-07
AGBC2N NMOS ’A’ for tunneling current model — 3.4254e-07
AGBC2P PMOS ’A’ for tunneling current model — 4.9723e-07
AGBCP Gate to body overlap area for bc parasitics m2 0
Parasitic Gate to body overlap area for bc parasitics /*
AGBCP2 m2 0
v4.1 improvement on BC
AGBCPD Gate to body overlap area for bc parasitics in DC m2 0
AGIDL GIDL second parameter — 0
AGISL GISL second parameter — 0
AGS Gate bias coefficient of Abulk. — 0
High level injection parameter for bipolar current /*
AHLI — 0
v4.0
High level injection parameter for bipolar current /*
AHLID — 0
v4.0
First Vgp dependent parameter for gate current in
AIGBCP2 — 0.043
accumulation in AGBCP2 region
AIGC Parameter for Igc — 0
AIGSD Parameter for Igs,d — 0
ALPHA0 substrate current model parameter m/V 0
First Vox dependent parameter for gate current in
ALPHAGB1 — 0.35
inversion
First Vox dependent parameter for gate current in
ALPHAGB2 — 0.43
accumulation
AS Source area m2 0
ASD Source/drain bottom diffusion smoothing parameter — 0.3

383
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
AT Temperature coefficient of vsat — 33000
B0 Abulk narrow width parameter m 0
B1 Abulk narrow width parameter m 0
BDOS Charge centroid parameter — 1
First Vds dependent parameter of impact ionization
BETA0 — 0
current
Second Vds dependent parameter of impact ionization
BETA1 — 0
current
Third Vds dependent parameter of impact ionization
BETA2 V 0.1
current
Second Vox dependent parameter for gate current in
BETAGB1 — 0.03
inversion
Second Vox dependent parameter for gate current in
BETAGB2 — 0.05
accumulation
BF Flicker noise length dependence exponent — 2
BG0SUB Band-gap of substrate at T=0K — 1.16
BGB1 ’B’ for Igb1 Tunneling current model — -3.1051e+10
BGB2 ’B’ for Igb2 Tunneling current model — -2.357e+10
BGBC2N NMOS ’B’ for tunneling current model — 1.1665e+12
BGBC2P PMOS ’B’ for tunneling current model — 7.4567e+11
BGIDL GIDL third parameter — 2.3e+09
BGISL GISL third parameter — 0
Second Vgp dependent parameter for gate current in
BIGBCP2 — 0.0054
accumulation in AGBCP2 region
BIGC Parameter for Igc — 0
BIGSD Parameter for Igs,d — 0
BINUNIT Bin unit selector — 1
BJTOFF BJT on/off flag — 0
CAPMOD Capacitance model selector — 2
CBJTII Length scaling parameter for II BJT part m 0
CDSBS coupling from Vd to Vbs for improved dVbi model — 0
CDSC Drain/Source and channel coupling capacitance — 0.00024
CDSCB Body-bias dependence of cdsc — 0
CDSCD Drain-bias dependence of cdsc — 0
CF Fringe capacitance parameter — 0
CFRCOEFF Fringe Cap parameter /* v4.4 — 1
CGDL New C-V model parameter — 0
CGDO Gate-drain overlap capacitance per width — 0

384
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
Gate substrate overlap capacitance per unit channel
CGEO — 0
length
CGIDL GIDL vb parameter — 0.5
CGISL GISL vb parameter — 0
CGSL New C-V model parameter — 0
CGSO Gate-source overlap capacitance per width — 0
Third Vgp dependent parameter for gate current in
CIGBCP2 — 0.0075
accumulation in AGBCP2 region
CIGC Parameter for Igc — 0
CIGSD Parameter for Igs,d — 0
CIT Interface state capacitance — 0
Source(gate side) sidewall junction capacitance per
CJSWG — 1e-10
unit width /* v4.0
Drain (gate side) sidewall junction capacitance per
CJSWGD — 0
unit width /* v4.0
CKAPPA New C-V model parameter — 0.6
CLC Vdsat parameter for C-V model m 1e-08
CLE Vdsat parameter for C-V model — 0
Source/drain sidewall fringing capacitance per unit
CSDESW — 0
length
CSDMIN Source/drain bottom diffusion minimum capacitance V 0
CTH0 Instance Thermal Capacitance — 1e-05
DELTA Effective Vds parameter — 0.01
the smoothing parameter in the Vox smoothing
DELTAVOX — 0.005
function
DELVT Threshold voltage adjust for CV V 0
DELVTO Zero bias threshold voltage variation V 0
third backgate body effect parameter for short channel
DK2B — 0
effect
DLBG Length offset fitting parameter for backgate charge m 0
DLC Delta L for C-V model m 0
DLCB Length offset fitting parameter for body charge m 0
DLCIG Delta L for Ig model m 0
DROUT DIBL coefficient of output resistance — 0.56
DSUB DIBL coefficient in the subthreshold region — 0
DTOXCV Delta oxide thickness in meters in CapMod3 /* v2.2.3 m 0
DVBD0 first short-channel effect parameter in FD module — 0
DVBD1 second short-channel effect parameter in FD module — 0

385
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
DVT0 Short channel effect coeff. 0 — 2.2
DVT0W Narrow Width coeff. 0 — 0
DVT1 Short channel effect coeff. 1 — 0.53
DVT1W Narrow Width effect coeff. 1 — 5.3e+06
DVT2 Short channel effect coeff. 2 — -0.032
DVT2W Narrow Width effect coeff. 2 — -0.032
DVTP0 First parameter for Vth shift due to pocket m 0
DVTP1 Second parameter for Vth shift due to pocket — 0
DVTP2 Third parameter for Vth shift due to pocket — 0
DVTP3 Third parameter for Vth shift due to pocket — 0
DVTP4 Forth parameter for Vth shift due to pocket — 0
DWB Width reduction parameter — 0
DWBC Width offset for body contact isolation edge m 0
DWC Delta W for C-V model m 0
DWG Width reduction parameter m/V 0
EASUB Electron affinity of substrate — 4.05
EBG effective bandgap in gate current calculation — 1.2
EBJTII Impact ionization parameter for BJT part — 0
EF Flicker noise frequency exponent — 1
EGGBCP2 Bandgap in Agbcp2 region — 1.12
EGGDEP Bandgap for gate depletion effect — 1.12
EGIDL GIDL first parameter V 0
EGISL GISL first parameter V 0
EM Flicker noise parameter — 4.1e+07
EOT Effective SiO2 thickness m 1e-08
EPSRGATE Dielectric constant of gate relative to vacuum — 11.7
Dielectric constant of the gate oxide relative to
EPSROX — 3.9
vacuum
EPSRSUB Dielectric constant of substrate relative to vacuum — 11.7
ESATII Saturation electric field for impact ionization — 1e+07
ETA0 Subthreshold region DIBL coefficient for I-V — 0.08
ETA0CV Subthreshold region DIBL coefficient for C-V — 0
ETAB Subthreshold region DIBL coefficient for I-V — -0.07
ETABCV Subthreshold region DIBL coefficient for C-V — 0
ETSI Effective Silicon-on-insulator thickness in meters m 1e-07
EU Mobility exponent — 0

386
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
Fraction of bipolar current affecting the impact
FBJTII — 0
ionization
FBODY Scaling factor for body charge — 1
FDMOD Improved dVbi model selector — 0
FGIDL GIDL vb parameter — 0
FGISL GISL vb parameter V 0
FNOIMOD Flicker noise model selector — 1
FPROUT Rout degradation coefficient for pocket devices — 0
FRBODY layout dependent body-resistance coefficient — 1
GAMMA1 Vth body coefficient — 0
GAMMA2 Vth body coefficient — 0
GBMIN Minimum body conductance — 1e-12
GIDLMOD parameter for GIDL selector — 0
IDBJT BJT injection saturation current — 0
Body to source/drain injection saturation current /*
IDDIF — 0
v4.0
IDREC Recombination in depletion saturation current — 0
IDTUN Reverse tunneling saturation current — 0
IGBMOD gate-body tunneling current model selector /* v3.0 — 0
IGCMOD gate-channel tunneling current model selector /* v3.0 — 0
IGMOD gate-body tunneling current model selector /* v3.1.1 — 0
IIIMOD parameter for III selector — 0
ISBJT BJT injection saturation current — 1e-06
ISDIF Body to source/drain injection saturation current — 0
ISREC Recombination in depletion saturation current — 1e-05
ISTUN Reverse tunneling saturation current — 0
K1 Bulk effect coefficient 1 — 0.53
K1B first backgate body effect parameter — 1
K1W1 First Body effect width dependent parameter m 0
K1W2 Second Body effect width dependent parameter m 0
K2 Bulk effect coefficient 2 — -0.0186
second backgate body effect parameter for short
K2B — 0
channel effect
K3 Narrow width effect coefficient — 0
K3B Body effect coefficient of k3 — 0
KB1 Scaling factor for backgate charge — 1

387
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
Body-bias coefficient of non-uniform depletion width
KETA — -0.6
effect.
KETAS Surface potential adjustment for bulk charge effect V 0
KF Flicker noise coefficient — 0
KGIDL GIDL vb parameter V 0
KGISL GISL vb parameter V 0
KT1 Temperature coefficient of Vth V -0.11
KT1L Temperature coefficient of Vth — 0
KT2 Body-coefficient of kt1 — 0.022
Mobility degradation/enhancement coefficient for
KU0 m 0
LOD
Saturation velocity degradation/enhancement
KVSAT m 0
parameter for LOD
Threshold degradation/enhancement parameter for
KVTH0 Vm 0
LOD
L Length m 5e-06
LA0 Length dependence of a0 — 0
LA1 Length dependence of a1 — 0
LA2 Length dependence of a2 — 0
LABJTII Length dependence of abjtii — 0
LACDE Length dependence of acde — 0
LAELY Length dependence of aely — 0
LAGIDL Length dependence of agidl — 0
LAGISL Length dependence of agisl — 0
LAGS Length dependence of ags — 0
LAHLI Length dependence of ahli /*v4.0 — 0
LAHLID Length dependence of ahlid /*v4.0 — 0
LAIGBCP2 Length dependence of aigbcp2 — 0
LAIGC Length dependence of aigc — 0
LAIGSD Length dependence of aigsd — 0
LALPHA0 Length dependence of alpha0 — 0
LALPHAGB1 Length dependence of alphagb1 — 0
LALPHAGB2 Length dependence of alphagb2 — 0
LAT Length dependence of at — 0
LB0 Length dependence of b0 — 0
LB1 Length dependence of b1 — 0
LBETA0 Length dependence of beta0 — 0

388
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
LBETA1 Length dependence of beta1 — 0
LBETA2 Length dependence of beta2 — 0
LBETAGB1 Length dependence of betagb1 — 0
LBETAGB2 Length dependence of betagb2 — 0
LBGIDL Length dependence of bgidl — 0
LBGISL Length dependence of bgisl — 0
LBIGBCP2 Length dependence of bigbcp2 — 0
LBIGC Length dependence of bigc — 0
LBIGSD Length dependence of bigsd — 0
LBJT0 Reference channel length for bipolar current m 2e-07
LCBJTII Length dependence of cbjtii — 0
LCDSBS Length dependence of cdsbs — 0
LCDSC Length dependence of cdsc — 0
LCDSCB Length dependence of cdscb — 0
LCDSCD Length dependence of cdscd — 0
LCGDL Length dependence of cgdl — 0
LCGIDL Length dependence of cgidl — 0
LCGISL Length dependence of cgisl — 0
LCGSL Length dependence of cgsl — 0
LCIGBCP2 Length dependence of cigbcp2 — 0
LCIGC Length dependence of cigc — 0
LCIGSD Length dependence of cigsd — 0
LCIT Length dependence of cit — 0
LCKAPPA Length dependence of ckappa — 0
LDELTA Length dependence of delta — 0
LDELVT Length dependence of delvt — 0
Channel-length dependency coefficient of diffusion
LDIF0 — 1
cap
LDK2B Length dependence of dk2b — 0
LDROUT Length dependence of drout — 0
LDSUB Length dependence of dsub — 0
LDVBD0 Length dependence of dvbd0 — 0
LDVBD1 Length dependence of dvbd1 — 0
LDVT0 Length dependence of dvt0 — 0
LDVT0W Length dependence of dvt0w — 0
LDVT1 Length dependence of dvt1 — 0

389
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
LDVT1W Length dependence of dvt1w — 0
LDVT2 Length dependence of dvt2 — 0
LDVT2W Length dependence of dvt2w — 0
LDVTP0 Length dependence of dvtp0 — 0
LDVTP1 Length dependence of dvtp1 — 0
LDVTP2 Length dependence of dvtp2 — 0
LDVTP3 Length dependence of dvtp3 — 0
LDVTP4 Length dependence of dvtp4 — 0
LDWB Length dependence of dwb — 0
LDWG Length dependence of dwg — 0
LEBJTII Length dependence of ebjtii — 0
LEFFEOT Effective length for extraction of EOT — 1
LEGIDL Length dependence of egidl — 0
LEGISL Length dependence of egisl — 0
LESATII Length dependence of esatii — 0
LETA0 Length dependence of eta0 — 0
LETA0CV Length dependence of eta0cv — 0
LETAB Length dependence of etab — 0
LETABCV Length dependence of etabcv — 0
LEU Length dependence of eu — 0
LFBJTII Length dependence of fbjtii — 0
LFGIDL Length dependence of fgidl — 0
LFGISL Length dependence of fgisl — 0
LFPROUT Length dependence of pdiblcb — 0
LIDBJT Length dependence of idbjt — 0
LIDDIF Length dependence of iddif — 0
LIDREC Length dependence of idrec — 0
LIDTUN Length dependence of idtun — 0
Channel length dependent parameter at threshold for
LII — 0
impact ionization current
LINT Length reduction parameter m 0
LISBJT Length dependence of isbjt — 0
LISDIF Length dependence of isdif — 0
LISREC Length dependence of isrec — 0
LISTUN Length dependence of istun — 0
LK1 Length dependence of k1 — 0

390
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
LK1B Length dependence of k1b — 0
LK1W1 Length dependence of k1w1 — 0
LK1W2 Length dependence of k1w2 — 0
LK2 Length dependence of k2 — 0
LK2B Length dependence of k2b — 0
LK3 Length dependence of k3 — 0
LK3B Length dependence of k3b — 0
LKB1 Length dependence of kb1 — 0
LKETA Length dependence of keta — 0
LKETAS Length dependence of ketas — 0
LKGIDL Length dependence of kgidl — 0
LKGISL Length dependence of kgisl — 0
LKT1 Length dependence of kt1 — 0
LKT1L Length dependence of kt1l — 0
LKT2 Length dependence of kt2 — 0
LKU0 Length dependence of ku0 — 0
LKVTH0 Length dependence of kvth0 — 0
LL Length reduction parameter m 0
LLBJT0 Length dependence of lbjt0 — 0
LLC Length reduction parameter /* v2.2.3 — 0
LLII Length dependence of lii — 0
LLN Length reduction parameter — 1
LLODKU0 Length parameter for u0 LOD effect — 0
LLODVTH Length parameter for vth LOD effect — 0
LLPE0 Length dependence of lpe0 — 0
LLPEB Length dependence of lpeb — 0
LMBJTII Length dependence of mbjtii — 0
LMINV Length dependence of minv — 0
LMINVCV Length dependence of minvcv — 0
LMOIN Length dependence of moin — 0
LMOINFD Length dependence of moinfd — 0
LN Electron/hole diffusion length m 2e-06
LNBJT Length dependence of nbjt — 0
LNCH Length dependence of nch — 0
LNDIF Length dependence of ndif — 0
LNDIODE Length dependence of ndiode — 0

391
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
LNDIODED Length dependence of ndioded — 0
LNFACTOR Length dependence of nfactor — 0
LNGATE Length dependence of ngate — 0
LNGIDL Length dependence of ngidl — 0
LNIGC Length dependence of nigc — 0
LNLX Length dependence of nlx — 0
LNOFF Length dependence of noff /* v3.2 — 0
LNOFF2 Length dependence of noff2 /* v4.6 — 0
LNOFFFD Length dependence of nofffd — 0
LNRECF0 Length dependence of nrecf0 — 0
LNRECF0D Length dependence of nrecf0d — 0
LNRECR0 Length dependence of nrecr0 — 0
LNRECR0D Length dependence of nrecr0d — 0
LNSD Length dependence of nsd — 0
LNSUB Length dependence of nsub — 0
LNTRECF Length dependence of ntrecf — 0
LNTRECR Length dependence of ntrecr — 0
LNTUN Length dependence of ntun — 0
LNTUND Length dependence of ntund — 0
LODETA0 eta0 shift modification factor for stress effect — 1
LODETA0CV eta0cv shift modification factor for stress effect — 0
LODK2 K2 shift modification factor for stress effect — 1
LPCLM Length dependence of pclm — 0
LPDIBLC1 Length dependence of pdiblc1 — 0
LPDIBLC2 Length dependence of pdiblc2 — 0
LPDIBLCB Length dependence of pdiblcb — 0
LPDITS Length dependence of pdits — 0
LPDITSD Length dependence of pditsd — 0
LPE0 Lateral non-uniform doping effect m 0
LPEB Lateral non-uniform doping effect for body bias m 0
LPIGCD Length dependence for pigcd — 0
LPOXEDGE Length dependence for poxedge — 0
LPRT Length dependence of prt — 0
LPRWB Length dependence of prwb — 0
LPRWG Length dependence of prwg — 0
LPVAG Length dependence of pvag — 0

392
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
LRDSW Length dependence of rdsw — 0
LRDW Length dependence of rdw /* v4.0 — 0
LRGIDL Length dependence of rgidl — 0
LRGISL Length dependence of rgisl — 0
LRSW Length dependence of rsw /* v4.0 — 0
LSII0 Length dependence of sii0 — 0
LSII1 Length dependence of sii1 — 0
LSII2 Length dependence of sii2 — 0
LSIID Length dependence of siid — 0
LU0 Length dependence of u0 — 0
LUA Length dependence of ua — 0
LUA1 Length dependence of ua1 — 0
LUB Length dependence of ub — 0
LUB1 Length dependence of ub1 — 0
LUC Length dependence of uc — 0
LUC1 Length dependence of uc1 — 0
LUCS Length dependence of lucs — 0
LUCSTE Length dependence of ucste — 0
LUD Length dependence of ud — 0
LUD1 Length dependence of ud1 — 0
LUTE Length dependence of ute — 0
LVABJT Length dependence of vabjt — 0
LVBCI Length dependence of vbci — 0
LVBS0FD Length dependence of vbs0fd — 0
LVBS0PD Length dependence of vbs0pd — 0
LVBSA Length dependence of vbsa — 0
LVDSATII0 Length dependence of vdsatii0 — 0
LVFB Length dependence of vfb /* v4.1 — 0
LVOFF Length dependence of voff — 0
LVOFFCV Length dependence of voffcv — 0
LVOFFFD Length dependence of vofffd — 0
LVREC0 Length dependence of vrec0 — 0
LVREC0D Length dependence of vrec0d — 0
LVSAT Length dependence of vsat — 0
LVSCE Length dependence of vsce — 0
LVSDFB Length dependence of vsdfb — 0

393
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
LVSDTH Length dependence of vsdth — 0
LVTH0 Length dependence of vto — 0
LVTUN0 Length dependence of vtun0 — 0
LVTUN0D Length dependence of vtun0d — 0
LW Length reduction parameter m 0
LW0 Length dependence of w0 — 0
LWC Length reduction parameter /* v2.2.3 — 0
LWL Length reduction parameter — 0
LWLC Length reduction parameter /* v2.2.3 — 0
LWN Length reduction parameter — 1
LWR Length dependence of wr — 0
LXBJT Length dependence of xbjt — 0
LXDIF Length dependence of xdif — 0
LXDIFD Length dependence of xdifd — 0
LXJ Length dependence of xj — 0
LXRCRG1 Length dependence of xrcrg1 — 0
LXRCRG2 Length dependence of xrcrg2 — 0
LXREC Length dependence of xrec — 0
LXRECD Length dependence of xrecd — 0
LXTUN Length dependence of xtun — 0
LXTUND Length dependence of xtund — 0
MBJTII Internal B-C grading coefficient — 0.4
MINV For moderate inversion in Vgsteff — 0
MINVCV For moderate inversion in VgsteffCV — 0
Source (gate side) sidewall junction capacitance
MJSWG V 0.5
grading coefficient /* v4.0
Drain (gate side) sidewall junction capacitance
MJSWGD V 0
grading coefficient /* v4.0
MOBMOD Mobility model selector — 1
Coefficient for the gate-bias dependent surface
MOIN — 15
potential
Coefficient for the gate-bias dependent surface
MOINFD — 1000
potential in FD
parameter for non-silicon substrate or metal gate
MTRLMOD — 0
selector
NBC Number of body contact isolation edge — 0
Power coefficient of channel length dependency for
NBJT — 1
bipolar current

394
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
NCH Channel doping concentration — 1.7e+17
Power coefficient of channel length dependency for
NDIF — -1
diffusion capacitance
NDIODE Diode non-ideality factor /*v4.0 — 1
NDIODED Diode non-ideality factor /*v4.0 — 0
NF Number of fingers — 1
NFACTOR Subthreshold swing Coefficient — 1
NGATE Poly-gate doping concentration — 0
NGCON Number of gate contacts — 1
NGIDL GIDL first parameter — 1.2
NI0SUB Intrinsic carrier concentration of substrate at Tnom — 1.45e+10
NIGC Parameter for Igc slope — 1
NLX Lateral non-uniform doping effect m 1.74e-07
NODECHK NODE checking flag — 1
NOFF C-V turn-on/off parameter /* v3.2 — 1
NOFF2 C-V turn-on/off parameter /* v4.6 — 0
NOFFFD smoothing parameter in FD module — 1
NOIA Flicker noise parameter — 0
NOIB Flicker noise parameter — 0
NOIC Flicker noise parameter — 8.75e+09
NOIF Floating body excess noise ideality factor — 1
NRB Number of squares in body — 1
NRD Number of squares in drain — 1
NRECF0 Recombination non-ideality factor at forward bias — 2
NRECF0D Recombination non-ideality factor at forward bias — 0
NRECR0 Recombination non-ideality factor at reversed bias — 10
NRECR0D Recombination non-ideality factor at reversed bias — 0
NRS Number of squares in source — 1
NSD S/D doping concentration — 1e+20
NSEG Number segments for width partitioning — 1
NSUB Substrate doping concentration with polarity — 6e+16
NTNOI Thermal noise parameter — 1
NTOX power term of gate current — 1
NTRECF Temperature coefficient for Nrecf — 0
NTRECR Temperature coefficient for Nrecr — 0
NTUN Reverse tunneling non-ideality factor — 10

395
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
NTUND Reverse tunneling non-ideality factor — 0
OFF Device is initially off — 0
PA0 Cross-term dependence of a0 — 0
PA1 Cross-term dependence of a1 — 0
PA2 Cross-term dependence of a2 — 0
PABJTII Cross-term dependence of abjtii — 0
PACDE Cross-term dependence of acde — 0
PAELY Cross-term dependence of aely — 0
PAGIDL Cross-term dependence of agidl — 0
PAGISL Cross-term dependence of agisl — 0
PAGS Cross-term dependence of ags — 0
PAHLI X-term dependence of ahli /* v4.0 — 0
PAHLID X-term dependence of ahlid /* v4.0 — 0
PAIGBCP2 Cross-term dependence of aigbcp2 — 0
PAIGC Cross-term dependence of aigc — 0
PAIGSD Cross-term dependence of aigsd — 0
PALPHA0 Cross-term dependence of alpha0 — 0
PALPHAGB1 Cross-term dependence of alphagb1 — 0
PALPHAGB2 Cross-term dependence of alphagb2 — 0
PARAMCHK Model parameter checking selector — 0
PAT Cross-term dependence of at — 0
PB0 Cross-term dependence of b0 — 0
PB1 Cross-term dependence of b1 — 0
PBETA0 Cross-term dependence of beta0 — 0
PBETA1 Cross-term dependence of beta1 — 0
PBETA2 Cross-term dependence of beta2 — 0
PBETAGB1 Cross-term dependence of betagb1 — 0
PBETAGB2 Cross-term dependence of betagb2 — 0
PBGIDL Cross-term dependence of bgidl — 0
PBGISL Cross-term dependence of bgisl — 0
PBIGBCP2 Cross-term dependence of bigbcp2 — 0
PBIGC Cross-term dependence of bigc — 0
PBIGSD Cross-term dependence of bigsd — 0
Source(gate side) sidewall junction capacitance built
PBSWG V 0.7
in potential /* v4.0

396
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
Drain(gate side) sidewall junction capacitance built in
PBSWGD V 0
potential /* v4.0
PCBJTII Cross-term dependence of cbjtii — 0
PCDSBS Cross-term dependence of cdsbs — 0
PCDSC Cross-term dependence of cdsc — 0
PCDSCB Cross-term dependence of cdscb — 0
PCDSCD Cross-term dependence of cdscd — 0
PCGDL Cross-term dependence of cgdl — 0
PCGIDL Cross-term dependence of cgidl — 0
PCGISL Cross-term dependence of cgisl — 0
PCGSL Cross-term dependence of cgsl — 0
PCIGBCP2 Cross-term dependence of cigbcp2 — 0
PCIGC Cross-term dependence of cigc — 0
PCIGSD Cross-term dependence of cigsd — 0
PCIT Cross-term dependence of cit — 0
PCKAPPA Cross-term dependence of ckappa — 0
PCLM Channel length modulation Coefficient — 1.3
PD Drain perimeter m 0
PDBCP Perimeter length for bc parasitics at drain side m 0
PDELTA Cross-term dependence of delta — 0
PDELVT Cross-term dependence of delvt — 0
PDIBLC1 Drain-induced barrier lowering coefficient — 0.39
PDIBLC2 Drain-induced barrier lowering coefficient — 0.0086
PDIBLCB Body-effect on drain-induced barrier lowering — 0
PDITS Coefficient for drain-induced Vth shifts — 1e-20
PDITSD Vds dependence of drain-induced Vth shifts — 0
PDITSL Length dependence of drain-induced Vth shifts — 0
PDK2B Cross-term dependence of dk2b — 0
PDROUT Cross-term dependence of drout — 0
PDSUB Cross-term dependence of dsub — 0
PDVBD0 Cross-term dependence of dvbd0 — 0
PDVBD1 Cross-term dependence of dvbd1 — 0
PDVT0 Cross-term dependence of dvt0 — 0
PDVT0W Cross-term dependence of dvt0w — 0
PDVT1 Cross-term dependence of dvt1 — 0
PDVT1W Cross-term dependence of dvt1w — 0

397
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
PDVT2 Cross-term dependence of dvt2 — 0
PDVT2W Cross-term dependence of dvt2w — 0
PDVTP0 Cross-term dependence of dvtp0 — 0
PDVTP1 Cross-term dependence of dvtp1 — 0
PDVTP2 Cross-term dependence of dvtp2 — 0
PDVTP3 Cross-term dependence of dvtp3 — 0
PDVTP4 Cross-term dependence of dvtp4 — 0
PDWB Cross-term dependence of dwb — 0
PDWG Cross-term dependence of dwg — 0
PEBJTII Cross-term dependence of ebjtii — 0
PEGIDL Cross-term dependence of egidl — 0
PEGISL Cross-term dependence of egisl — 0
PESATII Cross-term dependence of esatii — 0
PETA0 Cross-term dependence of eta0 — 0
PETA0CV Cross-term dependence of eta0cv — 0
PETAB Cross-term dependence of etab — 0
PETABCV Cross-term dependence of etabcv — 0
PEU Cross-term dependence of eu — 0
PFBJTII Cross-term dependence of fbjtii — 0
PFGIDL Cross-term dependence of fgidl — 0
PFGISL Cross-term dependence of fgisl — 0
PFPROUT Cross-term dependence of pdiblcb — 0
PHIG Work function of gate — 4.05
PIDBJT Cross-term dependence of idbjt — 0
PIDDIF Cross-term dependence of iddif — 0
PIDREC Cross-term dependence of idrec — 0
PIDTUN Cross-term dependence of idtun — 0
PIGCD Parameter for Igc partition — 1
PISBJT Cross-term dependence of isbjt — 0
PISDIF Cross-term dependence of isdif — 0
PISREC Cross-term dependence of isrec — 0
PISTUN Cross-term dependence of istun — 0
PK1 Cross-term dependence of k1 — 0
PK1B Cross-term dependence of k1b — 0
PK1W1 Cross-term dependence of k1w1 — 0
PK1W2 Cross-term dependence of k1w2 — 0

398
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
PK2 Cross-term dependence of k2 — 0
PK2B Cross-term dependence of k2b — 0
PK3 Cross-term dependence of k3 — 0
PK3B Cross-term dependence of k3b — 0
PKB1 Cross-term dependence of kb1 — 0
PKETA Cross-term dependence of keta — 0
PKETAS Cross-term dependence of ketas — 0
PKGIDL Cross-term dependence of kgidl — 0
PKGISL Cross-term dependence of kgisl — 0
PKT1 Cross-term dependence of kt1 — 0
PKT1L Cross-term dependence of kt1l — 0
PKT2 Cross-term dependence of kt2 — 0
PKU0 Cross-term dependence of ku0 — 0
PKVTH0 Cross-term dependence of kvth0 — 0
PLBJT0 Cross-term dependence of lbjt0 — 0
PLII Cross-term dependence of lii — 0
PLPE0 Cross-term dependence of lpe0 — 0
PLPEB Cross-term dependence of lpeb — 0
PMBJTII Cross-term dependence of mbjtii — 0
PMINV Cross-term dependence of minv — 0
PMINVCV Cross-term dependence of minvcv — 0
PMOIN Cross-term dependence of moin — 0
PMOINFD Cross-term dependence of moinfd — 0
PNBJT Cross-term dependence of nbjt — 0
PNCH Cross-term dependence of nch — 0
PNDIF Cross-term dependence of ndif — 0
PNDIODE Cross-term dependence of ndiode — 0
PNDIODED Cross-term dependence of ndiode — 0
PNFACTOR Cross-term dependence of nfactor — 0
PNGATE Cross-term dependence of ngate — 0
PNGIDL Cross-term dependence of ngidl — 0
PNIGC Cross-term dependence of nigc — 0
PNLX Cross-term dependence of nlx — 0
PNOFF Cross-term dependence of noff /* v3.2 — 0
PNOFF2 Cross-term dependence of noff2 /* v4.6 — 0
PNOFFFD Cross-term dependence of nofffd — 0

399
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
PNRECF0 Cross-term dependence of nrecf0 — 0
PNRECF0D Cross-term dependence of nrecf0 — 0
PNRECR0 Cross-term dependence of nrecr0 — 0
PNRECR0D Cross-term dependence of nrecr0 — 0
PNSD Cross-term dependence of nsd — 0
PNSUB Cross-term dependence of nsub — 0
PNTRECF Cross-term dependence of ntrecf — 0
PNTRECR Cross-term dependence of ntrecr — 0
PNTUN Cross-term dependence of ntun — 0
PNTUND Cross-term dependence of ntund — 0
POXEDGE Factor for the gate edge Tox — 1
PPCLM Cross-term dependence of pclm — 0
PPDIBLC1 Cross-term dependence of pdiblc1 — 0
PPDIBLC2 Cross-term dependence of pdiblc2 — 0
PPDIBLCB Cross-term dependence of pdiblcb — 0
PPDITS Cross-term dependence of pdits — 0
PPDITSD Cross-term dependence of pditsd — 0
PPIGCD Cross-term dependence for pigcd — 0
PPOXEDGE Cross-term dependence for poxedge — 0
PPRT Cross-term dependence of prt — 0
PPRWB Cross-term dependence of prwb — 0
PPRWG Cross-term dependence of prwg — 0
PPVAG Cross-term dependence of pvag — 0
PRDSW Cross-term dependence of rdsw — 0
PRDW Cross-term dependence of rdw /*v4.0 — 0
PRGIDL Cross-term dependence of rgidl — 0
PRGISL Cross-term dependence of rgisl — 0
PRSW Cross-term dependence of rsw /*v4.0 — 0
PRT Temperature coefficient of parasitic resistance — 0
PRWB Body-effect on parasitic resistance — 0
PRWG Gate-bias effect on parasitic resistance — 0
PS Source perimeter m 0
PSBCP Perimeter length for bc parasitics at source side m 0
PSII0 Cross-term dependence of sii0 — 0
PSII1 Cross-term dependence of sii1 — 0
PSII2 Cross-term dependence of sii2 — 0

400
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
PSIID Cross-term dependence of siid — 0
PU0 Cross-term dependence of u0 — 0
PUA Cross-term dependence of ua — 0
PUA1 Cross-term dependence of ua1 — 0
PUB Cross-term dependence of ub — 0
PUB1 Cross-term dependence of ub1 — 0
PUC Cross-term dependence of uc — 0
PUC1 Cross-term dependence of uc1 — 0
PUCS Cross-term dependence of ucs — 0
PUCSTE Cross-term dependence of ucste — 0
PUD Cross-term dependence of ud — 0
PUD1 Cross-term dependence of ud1 — 0
PUTE Cross-term dependence of ute — 0
PVABJT Cross-term dependence of vabjt — 0
PVAG Gate dependence of output resistance parameter — 0
PVBCI Cross-term dependence of vbci — 0
PVBS0FD Cross-term dependence of vbs0fd — 0
PVBS0PD Cross-term dependence of vbs0pd — 0
PVBSA Cross-term dependence of vbsa — 0
PVDSATII0 Cross-term dependence of vdsatii0 — 0
PVFB Cross-term dependence of vfb /* v4.1 — 0
PVOFF Cross-term dependence of voff — 0
PVOFFCV Cross-term dependence of voffcv — 0
PVOFFFD Cross-term dependence of vofffd — 0
PVREC0 Cross-term dependence of vrec0 — 0
PVREC0D Cross-term dependence of vrec0d — 0
PVSAT Cross-term dependence of vsat — 0
PVSCE Cross-term dependence of vsce — 0
PVSDFB Cross-term dependence of vsdfb — 0
PVSDTH Cross-term dependence of vsdth — 0
PVTH0 Cross-term dependence of vto — 0
PVTUN0 Cross-term dependence of vtun0 — 0
PVTUN0D Cross-term dependence of vtun0d — 0
PW0 Cross-term dependence of w0 — 0
PWR Cross-term dependence of wr — 0
PXBJT Cross-term dependence of xbjt — 0

401
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
PXDIF Cross-term dependence of xdif — 0
PXDIFD Cross-term dependence of xdifd — 0
PXJ Cross-term dependence of xj — 0
PXRCRG1 Cross-term dependence of xrcrg1 — 0
PXRCRG2 Cross-term dependence of xrcrg2 — 0
PXREC Cross-term dependence of xrec — 0
PXRECD Cross-term dependence of xrecd — 0
PXTUN Cross-term dependence of xtun — 0
PXTUND Cross-term dependence of xtund — 0
RBDB Body resistance — 50
RBODY Intrinsic body contact sheet resistance — 0
RBODYMOD Body R model selector /* v4.0 — 0
RBSB Body resistance — 50
RBSH Extrinsic body contact sheet resistance — 0
RDSMOD Bias-dependent S/D resistance model selector /* v4.0 — 0
RDSW Source-drain resistance per width — 100
RDW Drain resistance per width /* v4.0 — 50
RDWMIN Drain resistance per width at hight Vg — 0
RGATEMOD Gate resistance model selector — 0
RGIDL GIDL vg parameter — 1
RGISL GISL vg parameter — 0
RHALO body halo sheet resistance — 1e+15
RNOIA Thermal noise coefficient — 0.577
RNOIB Thermal noise coefficient — 0.37
RSH Source-drain sheet resistance — 0
RSHG Gate sheet resistance — 0.1
RSW Source resistance per width /* v4.0 — 50
RSWMIN Source resistance per width at high Vg — 0
RTH0 Instance Thermal Resistance — 0
SA distance between OD edge to poly of one side m 0
Reference distance between OD edge to poly of one
SAREF m 1e-06
side
SB distance between OD edge to poly of the other side m 0
Reference distance between OD edge to poly of the
SBREF m 1e-06
other side
SD distance between neighbor fingers m 0

402
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
SHMOD Self heating mode selector — 0
First Vgs dependent parameter for impact ionization
SII0 — 0.5
current
Second Vgs dependent parameter for impact
SII1 — 0.1
ionization current
Third Vgs dependent parameter for impact ionization
SII2 — 0
current
Vds dependent parameter of drain saturation voltage
SIID — 0
for impact ionization current
SOIMOD Instance model selector for PD/FD operation /* v3.2 — 0
STETA0 eta0 shift factor related to stress effect on vth m 0
STETA0CV eta0cv shift factor related to stress effect on vth — 0
STK2 K2 shift factor related to stress effect on vth m 0
TBGASUB First parameter of band-gap change due to temperature — 0.000702
Second parameter of band-gap change due to
TBGBSUB — 1108
temperature
TBOX Back gate oxide thickness in meters m 3e-07
TCJSWG Temperature coefficient of Cjswgs 1/K 0
TCJSWGD Temperature coefficient of Cjswgd 1/K 0
TEMPEOT Temperature for extraction of EOT K 300.15
Temperature dependent parameter for impact
TII — 0
ionization
TKU0 Temperature coefficient of KU0 — 0
TNOIA Thermal noise parameter — 1.5
TNOIB Thermal noise parameter — 3.5
TNOIMOD Thermal noise model selector — 0
TNOM Parameter measurement temperature — 27
TOX Gate oxide thickness in meters m 1e-08
TOXM Gate oxide thickness used in extraction /* v3.2 m 0
TOXP Physical gate oxide thickness m 0
TOXQM effective oxide thickness considering quantum effect m 0
TOXREF target oxide thickness m 2.5e-09
TPBSWG Temperature coefficient of Pbswgs V/K 0
TPBSWGD Temperature coefficient of Pbswgd V/K 0
TSI Silicon-on-insulator thickness in meters m 1e-07
TT Diffusion capacitance transit time coefficient s 1e-12
TVBCI Temperature coefficient for VBCI — 0
TYPE +1 = NMOS, -1 = PMOS — 1

403
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
U0 Low-field mobility at Tnom — 0
UA Linear gate dependence of mobility — 2.25e-09
UA1 Temperature coefficient of ua m/V 4.31e-09
UB Quadratic gate dependence of mobility — 5.87e-19
UB1 Temperature coefficient of ub — -7.61e-18
UC Body-bias dependence of mobility — 0
UC1 Temperature coefficient of uc — 0
UCS Mobility exponent — 0
UCSTE Temperature coefficient of UCS — -0.004775
UD Coulomb scattering factor of mobility — 0
UD1 Temperature coefficient of ud — 0
UTE Temperature coefficient of mobility — -1.5
VABJT Early voltage for bipolar current V 10
VBCI Internal B-C built-in potential V 0
VBM Maximum body voltage V -3
Lower bound of built-in potential lowering for FD
VBS0FD V 0.5
operation /* v3.2
Upper bound of built-in potential lowering for PD
VBS0PD V 0
operation /* v3.2
VBSA vbsa offset voltage V 0
VBSUSR Vbs specified by user V 0
VBX Vth transition body Voltage — 0
VDDEOT Voltage for extraction of EOT — 0
Nominal drain saturation voltage at threshold for
VDSATII0 — 0.9
impact ionization current
Vaux parameter for conduction-band electron
VECB — 0.026
tunneling
VERSION parameter for model version — 4.6
VEVB Vaux parameter for valence-band electron tunneling — 0.075
VFB Flat Band Voltage /* v4.1 V -1
Third Vox dependent parameter for gate current in
VGB1 V 300
inversion
Third Vox dependent parameter for gate current in
VGB2 V 17
accumulation
VGSTCVMOD Improved VgsteffCV selector — 0
VOFF Threshold voltage offset V -0.08
CV Threshold voltage offset // NOT -0.08 for
VOFFCV V 0
backwards-compatibility

404
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
VOFFFD smoothing parameter in FD module V 0
VOXH the limit of Vox in gate current calculation V 5
Voltage dependent parameter for recombination
VREC0 V 0
current
Voltage dependent parameter for recombination
VREC0D V 0
current
VSAT Saturation velocity at tnom — 80000
VSCE SCE parameter for improved dVbi model — 0
Source/drain bottom diffusion capacitance flatband
VSDFB V 0
voltage
Source/drain bottom diffusion capacitance threshold
VSDTH V 0
voltage
VTH0 Threshold voltage V 0
VTHO Threshold voltage V 0
VTM00 Hard coded 25 degC thermal voltage V 0.026
VTUN0 Voltage dependent parameter for tunneling current V 0
VTUN0D Voltage dependent parameter for tunneling current V 0
W Width m 5e-06
W0 Narrow width effect parameter m 2.5e-06
W0FLK Flicker noise width dependence — 1e-05
WA0 Width dependence of a0 — 0
WA1 Width dependence of a1 — 0
WA2 Width dependence of a2 — 0
WABJTII Width dependence of abjtii — 0
WACDE Width dependence of acde — 0
WAELY Width dependence of aely — 0
WAGIDL Width dependence of agidl — 0
WAGISL Width dependence of agisl — 0
WAGS Width dependence of ags — 0
WAHLI Width dependence of ahli /* v4.0 — 0
WAHLID Width dependence of ahlid /* v4.0 — 0
WAIGBCP2 Width dependence of aigbcp2 — 0
WAIGC Width dependence of aigc — 0
WAIGSD Width dependence of aigsd — 0
WALPHA0 Width dependence of alpha0 — 0
WALPHAGB1 Width dependence of alphagb1 — 0
WALPHAGB2 Width dependence of alphagb2 — 0

405
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
WAT Width dependence of at — 0
WB0 Width dependence of b0 — 0
WB1 Width dependence of b1 — 0
WBETA0 Width dependence of beta0 — 0
WBETA1 Width dependence of beta1 — 0
WBETA2 Width dependence of beta2 — 0
WBETAGB1 Width dependence of betagb1 — 0
WBETAGB2 Width dependence of betagb2 — 0
WBGIDL Width dependence of bgidl — 0
WBGISL Width dependence of bgisl — 0
WBIGBCP2 Width dependence of bigbcp2 — 0
WBIGC Width dependence of bigc — 0
WBIGSD Width dependence of bigsd — 0
WCBJTII Width dependence of cbjtii — 0
WCDSBS Width dependence of cdsbs — 0
WCDSC Width dependence of cdsc — 0
WCDSCB Width dependence of cdscb — 0
WCDSCD Width dependence of cdscd — 0
WCGDL Width dependence of cgdl — 0
WCGIDL Width dependence of cgidl — 0
WCGISL Width dependence of cgisl — 0
WCGSL Width dependence of cgsl — 0
WCIGBCP2 Width dependence of cigbcp2 — 0
WCIGC Width dependence of cigc — 0
WCIGSD Width dependence of cigsd — 0
WCIT Width dependence of cit — 0
WCKAPPA Width dependence of ckappa — 0
WDELTA Width dependence of delta — 0
WDELVT Width dependence of delvt — 0
WDK2B Width dependence of dk2b — 0
WDROUT Width dependence of drout — 0
WDSUB Width dependence of dsub — 0
WDVBD0 Width dependence of dvbd0 — 0
WDVBD1 Width dependence of dvbd1 — 0
WDVT0 Width dependence of dvt0 — 0
WDVT0W Width dependence of dvt0w — 0

406
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
WDVT1 Width dependence of dvt1 — 0
WDVT1W Width dependence of dvt1w — 0
WDVT2 Width dependence of dvt2 — 0
WDVT2W Width dependence of dvt2w — 0
WDVTP0 Width dependence of dvtp0 — 0
WDVTP1 Width dependence of dvtp1 — 0
WDVTP2 Width dependence of dvtp2 — 0
WDVTP3 Width dependence of dvtp3 — 0
WDVTP4 Width dependence of dvtp4 — 0
WDWB Width dependence of dwb — 0
WDWG Width dependence of dwg — 0
WEBJTII Width dependence of ebjtii — 0
WEFFEOT Effective width for extraction of EOT — 10
WEGIDL Width dependence of egidl — 0
WEGISL Width dependence of egisl — 0
WESATII Width dependence of esatii — 0
WETA0 Width dependence of eta0 — 0
WETA0CV Width dependence of eta0cv — 0
WETAB Width dependence of etab — 0
WETABCV Width dependence of etabcv — 0
WEU Width dependence of eu — 0
WFBJTII Width dependence of fbjtii — 0
WFGIDL Width dependence of fgidl — 0
WFGISL Width dependence of fgisl — 0
WFPROUT Width dependence of pdiblcb — 0
WIDBJT Width dependence of idbjt — 0
WIDDIF Width dependence of iddif — 0
WIDREC Width dependence of idrec — 0
WIDTUN Width dependence of idtun — 0
WINT Width reduction parameter m 0
WISBJT Width dependence of isbjt — 0
WISDIF Width dependence of isdif — 0
WISREC Width dependence of isrec — 0
WISTUN Width dependence of istun — 0
WK1 Width dependence of k1 — 0
WK1B Width dependence of k1b — 0

407
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
WK1W1 Width dependence of k1w1 — 0
WK1W2 Width dependence of k1w2 — 0
WK2 Width dependence of k2 — 0
WK2B Width dependence of k2b — 0
WK3 Width dependence of k3 — 0
WK3B Width dependence of k3b — 0
WKB1 Width dependence of kb1 — 0
WKETA Width dependence of keta — 0
WKETAS Width dependence of ketas — 0
WKGIDL Width dependence of kgidl — 0
WKGISL Width dependence of kgisl — 0
WKT1 Width dependence of kt1 — 0
WKT1L Width dependence of kt1l — 0
WKT2 Width dependence of kt2 — 0
WKU0 Width dependence of ku0 — 0
WKVTH0 Width dependence of kvth0 — 0
WL Width reduction parameter m 0
WLBJT0 Width dependence of lbjt0 — 0
WLC Width reduction parameter /* v2.2.3 — 0
WLII Width dependence of lii — 0
WLN Width reduction parameter — 1
WLOD Width parameter for stress effect m 0
WLODKU0 Width parameter for u0 LOD effect — 0
WLODVTH Width parameter for vth LOD effect — 0
WLPE0 Width dependence of lpe0 — 0
WLPEB Width dependence of lpeb — 0
WMBJTII Width dependence of mbjtii — 0
WMINV width dependence of minv — 0
WMINVCV width dependence of minvcv — 0
WMOIN Width dependence of moin — 0
WMOINFD Width dependence of moinfd — 0
WNBJT Width dependence of nbjt — 0
WNCH Width dependence of nch — 0
WNDIF Width dependence of ndif — 0
WNDIODE Width dependence of ndiode — 0
WNDIODED Width dependence of ndioded — 0

408
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
WNFACTOR Width dependence of nfactor — 0
WNGATE Width dependence of ngate — 0
WNGIDL Width dependence of ngidl — 0
WNIGC Width dependence of nigc — 0
WNLX Width dependence of nlx — 0
WNOFF Width dependence of noff /* v3.2 — 0
WNOFF2 Width dependence of noff2 /* v4.6 — 0
WNOFFFD Width dependence of nofffd — 0
WNRECF0 Width dependence of nrecf0 — 0
WNRECF0D Width dependence of nrecf0d — 0
WNRECR0 Width dependence of nrecr0 — 0
WNRECR0D Width dependence of nrecr0d — 0
WNSD Width dependence of nsd — 0
WNSUB Width dependence of nsub — 0
WNTRECF Width dependence of ntrecf — 0
WNTRECR Width dependence of ntrecr — 0
WNTUN Width dependence of ntun — 0
WNTUND Width dependence of ntund — 0
WPCLM Width dependence of pclm — 0
WPDIBLC1 Width dependence of pdiblc1 — 0
WPDIBLC2 Width dependence of pdiblc2 — 0
WPDIBLCB Width dependence of pdiblcb — 0
WPDITS Width dependence of pdits — 0
WPDITSD Width dependence of pditsd — 0
WPIGCD Width dependence for pigcd — 0
WPOXEDGE Width dependence for poxedge — 0
WPRT Width dependence of prt — 0
WPRWB Width dependence of prwb — 0
WPRWG Width dependence of prwg — 0
WPVAG Width dependence of pvag — 0
WR Width dependence of rds — 1
WRDSW Width dependence of rdsw — 0
WRDW Width dependence of rdw /* v4.0 — 0
WRGIDL Width dependence of rgidl — 0
WRGISL Width dependence of rgisl — 0
WRSW Width dependence of rsw /* v4.0 — 0

409
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
WSII0 Width dependence of sii0 — 0
WSII1 Width dependence of sii1 — 0
WSII2 Width dependence of sii2 — 0
WSIID Width dependence of siid — 0
WTH0 Minimum width for thermal resistance calculation m 0
WU0 Width dependence of u0 — 0
WUA Width dependence of ua — 0
WUA1 Width dependence of ua1 — 0
WUB Width dependence of ub — 0
WUB1 Width dependence of ub1 — 0
WUC Width dependence of uc — 0
WUC1 Width dependence of uc1 — 0
WUCS Width dependence of ucs — 0
WUCSTE Width dependence of ucste — 0
WUD Width dependence of ud — 0
WUD1 Width dependence of ud1 — 0
WUTE Width dependence of ute — 0
WVABJT Width dependence of vabjt — 0
WVBCI Width dependence of vbci — 0
WVBS0FD Width dependence of vbs0fd — 0
WVBS0PD Width dependence of vbs0pd — 0
WVBSA Width dependence of vbsa — 0
WVDSATII0 Width dependence of vdsatii0 — 0
WVFB Width dependence of vfb /* v4.1 — 0
WVOFF Width dependence of voff — 0
WVOFFCV Width dependence of voffcv — 0
WVOFFFD Width dependence of vofffd — 0
WVREC0 Width dependence of vrec0 — 0
WVREC0D Width dependence of vrec0d — 0
WVSAT Width dependence of vsat — 0
WVSCE Width dependence of vsce — 0
WVSDFB Width dependence of vsdfb — 0
WVSDTH Width dependence of vsdth — 0
WVTH0 Width dependence of vto — 0
WVTUN0 Width dependence of vtun0 — 0
WVTUN0D Width dependence of vtun0d — 0

410
 Table 2-108. BSIM-SOI 4.6.1 Device Model Parameters
Parameter Description Units Default
WW Width reduction parameter m 0
WW0 Width dependence of w0 — 0
WWC Width reduction parameter /* v2.2.3 — 0
WWL Width reduction parameter m 0
WWLC Width reduction parameter /* v2.2.3 — 0
WWN Width reduction parameter — 1
WWR Width dependence of wr — 0
WXBJT Width dependence of xbjt — 0
WXDIF Width dependence of xdif — 0
WXDIFD Width dependence of xdifd — 0
WXJ Width dependence of xj — 0
WXRCRG1 Width dependence of xrcrg1 — 0
WXRCRG2 Width dependence of xrcrg2 — 0
WXREC Width dependence of xrec — 0
WXRECD Width dependence of xrecd — 0
WXTUN Width dependence of xtun — 0
WXTUND Width dependence of xtund — 0
XBJT Temperature coefficient for Isbjt — 1
XDIF Temperature coefficient for Isdif — 0
XDIFD Temperature coefficient for Iddif — 0
XGL Variation in Ldrawn m 0
XGW Distance from gate contact center to device edge m 0
XJ Junction Depth m 0
XPART Channel charge partitioning — 0
XRCRG1 First fitting parameter the bias-dependent Rg — 12
XRCRG2 Second fitting parameter the bias-dependent Rg — 1
XREC Temperature coefficient for Isrec — 1
XRECD Temperature coefficient for Idrec — 0
XT Doping depth m 1.55e-07
XTUN Temperature coefficient for Istun — 0
XTUND Temperature coefficient for Idtun — 0

Table 2-109. MOSFET level 70 Output Variables

Parameter Description Units Default
VDSAT V none

411
 Table 2-109. MOSFET level 70 Output Variables
Parameter Description Units Default
VTH Threshold Voltage V none
IDS Drain-Source current A none
GM mho none
GDS mho none
GMBS mho none
IC Collector Current A none
IBD A none
IBS A none
IGIDL A none
IGISL A none
IGS A none
IGD A none
IGB A none
IGCS A none
IGCD A none
CGG g-g MOSFET capacitance F none
CGS g-s MOSFET capacitance F none
CGD g-d MOSFET capacitance F none
CBG b-g MOSFET capacitance F none
CBS b-s MOSFET capacitance F none
CBD b-d MOSFET capacitance F none
CDG d-g MOSFET capacitance F none
CDD d-d MOSFET capacitance F none
CDS d-s MOSFET capacitance F none
CAPBD MOSFET capacitance F none
CAPBS MOSFET capacitance F none
QG Gate Charge C none
QB Body Charge C none
QD Drain Charge C none
QS Source Charge C none
QJD Drain Junction Charge C none
QJS Source Junction Charge C none
T_TOTAL_K K none
T_TOTAL_C K none
T_DELTA_SH K none

412
 Table 2-110. BSIM-SOI 4.5.0 Device Instance Parameters
Parameter Description Units Default
AD – 0
AEBCP – 0
AGBCP – 0
AGBCP2 – 0
AGBCPD – 0
AS – 0
BJTOFF – 0
CTH0 – 1e-05
DELVTO – 0
DTEMP – 0
FRBODY – 1
L – 5e-06
M multiplicity factor – 1
NBC – 0
NF – 1
NRB – 1
NRD – 1
NRS – 1
NSEG – 1
OFF – 0
PD – 0
PDBCP – 0
PS – 0
PSBCP – 0
RBDB – 50
RBSB – 50
RTH0 – 0
SA – 0
SB – 0
SD – 0
SHMOD – 0
SOIMOD – 0
TNODEOUT – 0
W – 5e-06

413
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
A0 – 1
A1 – 0
A2 – 1
ABJTII – 0
ACDE – 1
AD – 0
ADOS – 1
AEBCP – 0
AELY – 0
AF – 1
AGB1 – 3.7622e-07
AGB2 – 4.9758e-07
AGBC2N – 3.4254e-07
AGBC2P – 4.9723e-07
AGBCP – 0
AGBCP2 – 0
AGBCPD – 0
AGIDL – 0
AGISL – 0
AGS – 0
AHLI – 0
AHLID – 0
AIGBCP2 – 0.043
AIGC – 0
AIGSD – 0
ALPHA0 – 0
ALPHAGB1 – 0.35
ALPHAGB2 – 0.43
AS – 0
ASD – 0.3
AT – 33000
B0 – 0
B1 – 0
BDOS – 1
BETA0 – 0
BETA1 – 0

414
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
BETA2 – 0.1
BETAGB1 – 0.03
BETAGB2 – 0.05
BF – 2
BG0SUB – 1.16
BGB1 – -3.1051e+10
BGB2 – -2.357e+10
BGBC2N – 1.1665e+12
BGBC2P – 7.4567e+11
BGIDL – 2.3e+09
BGISL – 0
BIGBCP2 – 0.0054
BIGC – 0
BIGSD – 0
BINUNIT – 1
BJTOFF – 0
CAPMOD – 2
CBJTII – 0
CDSBS – 0
CDSC – 0.00024
CDSCB – 0
CDSCD – 0
CF – 0
CFRCOEFF – 1
CGDL – 0
CGDO – 0
CGEO – 0
CGIDL – 0.5
CGISL – 0
CGSL – 0
CGSO – 0
CIGBCP2 – 0.0075
CIGC – 0
CIGSD – 0
CIT – 0
CJSWG – 1e-10

415
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
CJSWGD – 0
CKAPPA – 0.6
CLC – 1e-08
CLE – 0
CSDESW – 0
CSDMIN – 0
CTH0 – 1e-05
DEBUG – 0
DELTA – 0.01
DELTAVOX – 0.005
DELVT – 0
DELVTO – 0
DK2B – 0
DLBG – 0
DLC – 0
DLCB – 0
DLCIG – 0
DROUT – 0.56
DSUB – 0
DTOXCV – 0
DVBD0 – 0
DVBD1 – 0
DVT0 – 2.2
DVT0W – 0
DVT1 – 0.53
DVT1W – 5.3e+06
DVT2 – -0.032
DVT2W – -0.032
DVTP0 – 0
DVTP1 – 0
DVTP2 – 0
DVTP3 – 0
DVTP4 – 0
DWB – 0
DWBC – 0
DWC – 0

416
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
DWG – 0
EASUB – 4.05
EBG – 1.2
EBJTII – 0
EF – 1
EGGBCP2 – 1.12
EGGDEP – 1.12
EGIDL – 0
EGISL – 0
EM – 4.1e+07
EOT – 1e-08
EPSRGATE – 11.7
EPSROX – 3.9
EPSRSUB – 11.7
ESATII – 1e+07
ETA0 – 0.08
ETA0CV – 0
ETAB – -0.07
ETABCV – 0
ETSI – 1e-07
EU – 0
FBJTII – 0
FBODY – 1
FDMOD – 0
FGIDL – 0
FGISL – 0
FNOIMOD – 1
FPROUT – 0
FRBODY – 1
GAMMA1 – 0
GAMMA2 – 0
GBMIN – 1e-12
GIDLMOD – 0
IDBJT – 0
IDDIF – 0
IDREC – 0

417
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
IDTUN – 0
IGBMOD – 0
IGCMOD – 0
IGMOD – 0
IIIMOD – 0
ISBJT – 1e-06
ISDIF – 0
ISREC – 1e-05
ISTUN – 0
K1 – 0.53
K1B – 1
K1W1 – 0
K1W2 – 0
K2 – -0.0186
K2B – 0
K3 – 0
K3B – 0
KB1 – 1
KETA – -0.6
KETAS – 0
KF – 0
KGIDL – 0
KGISL – 0
KT1 – -0.11
KT1L – 0
KT2 – 0.022
KU0 – 0
KVSAT – 0
KVTH0 – 0
L – 5e-06
LA0 – 0
LA1 – 0
LA2 – 0
LABJTII – 0
LACDE – 0
LAELY – 0

418
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
LAGIDL – 0
LAGISL – 0
LAGS – 0
LAHLI – 0
LAHLID – 0
LAIGBCP2 – 0
LAIGC – 0
LAIGSD – 0
LALPHA0 – 0
LALPHAGB1 – 0
LALPHAGB2 – 0
LAT – 0
LB0 – 0
LB1 – 0
LBETA0 – 0
LBETA1 – 0
LBETA2 – 0
LBETAGB1 – 0
LBETAGB2 – 0
LBGIDL – 0
LBGISL – 0
LBIGBCP2 – 0
LBIGC – 0
LBIGSD – 0
LBJT0 – 2e-07
LCBJTII – 0
LCDSC – 0
LCDSCB – 0
LCDSCD – 0
LCGDL – 0
LCGIDL – 0
LCGISL – 0
LCGSL – 0
LCIGBCP2 – 0
LCIGC – 0
LCIGSD – 0

419
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
LCIT – 0
LCKAPPA – 0
LDELTA – 0
LDELVT – 0
LDIF0 – 1
LDROUT – 0
LDSUB – 0
LDVT0 – 0
LDVT0W – 0
LDVT1 – 0
LDVT1W – 0
LDVT2 – 0
LDVT2W – 0
LDVTP0 – 0
LDVTP1 – 0
LDVTP2 – 0
LDVTP3 – 0
LDVTP4 – 0
LDWB – 0
LDWG – 0
LEBJTII – 0
LEFFEOT – 1
LEGIDL – 0
LEGISL – 0
LESATII – 0
LETA0 – 0
LETA0CV – 0
LETAB – 0
LETABCV – 0
LEU – 0
LFBJTII – 0
LFGIDL – 0
LFGISL – 0
LFPROUT – 0
LIDBJT – 0
LIDDIF – 0

420
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
LIDREC – 0
LIDTUN – 0
LII – 0
LINT – 0
LISBJT – 0
LISDIF – 0
LISREC – 0
LISTUN – 0
LK1 – 0
LK1W1 – 0
LK1W2 – 0
LK2 – 0
LK3 – 0
LK3B – 0
LKB1 – 0
LKETA – 0
LKETAS – 0
LKGIDL – 0
LKGISL – 0
LKT1 – 0
LKT1L – 0
LKT2 – 0
LKU0 – 0
LKVTH0 – 0
LL – 0
LLBJT0 – 0
LLC – 0
LLII – 0
LLN – 1
LLODKU0 – 0
LLODVTH – 0
LLPE0 – 0
LLPEB – 0
LMBJTII – 0
LMINV – 0
LMINVCV – 0

421
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
LMOIN – 0
LN – 2e-06
LNBJT – 0
LNCH – 0
LNDIF – 0
LNDIODE – 0
LNDIODED – 0
LNFACTOR – 0
LNGATE – 0
LNGIDL – 0
LNIGC – 0
LNLX – 0
LNOFF – 0
LNRECF0 – 0
LNRECF0D – 0
LNRECR0 – 0
LNRECR0D – 0
LNSD – 0
LNSUB – 0
LNTRECF – 0
LNTRECR – 0
LNTUN – 0
LNTUND – 0
LODETA0 – 1
LODETA0CV – 0
LODK2 – 1
LPCLM – 0
LPDIBLC1 – 0
LPDIBLC2 – 0
LPDIBLCB – 0
LPDITS – 0
LPDITSD – 0
LPE0 – 0
LPEB – 0
LPIGCD – 0
LPOXEDGE – 0

422
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
LPRT – 0
LPRWB – 0
LPRWG – 0
LPVAG – 0
LRDSW – 0
LRDW – 0
LRGIDL – 0
LRGISL – 0
LRSW – 0
LSII0 – 0
LSII1 – 0
LSII2 – 0
LSIID – 0
LU0 – 0
LUA – 0
LUA1 – 0
LUB – 0
LUB1 – 0
LUC – 0
LUC1 – 0
LUCS – 0
LUCSTE – 0
LUD – 0
LUD1 – 0
LUTE – 0
LVABJT – 0
LVBCI – 0
LVDSATII0 – 0
LVFB – 0
LVOFF – 0
LVOFFCV – 0
LVREC0 – 0
LVREC0D – 0
LVSAT – 0
LVSDFB – 0
LVSDTH – 0

423
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
LVTH0 – 0
LVTUN0 – 0
LVTUN0D – 0
LW – 0
LW0 – 0
LWC – 0
LWL – 0
LWLC – 0
LWN – 1
LWR – 0
LXBJT – 0
LXDIF – 0
LXDIFD – 0
LXJ – 0
LXRCRG1 – 0
LXRCRG2 – 0
LXREC – 0
LXRECD – 0
LXTUN – 0
LXTUND – 0
MBJTII – 0.4
MINV – 0
MINVCV – 0
MJSWG – 0.5
MJSWGD – 0
MOBMOD – 1
MOIN – 15
MOINFD – 1000
MTRLMOD – 0
NBC – 0
NBJT – 1
NCH – 1.7e+17
NDIF – -1
NDIODE – 1
NDIODED – 0
NF – 1

424
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
NFACTOR – 1
NGATE – 0
NGCON – 1
NGIDL – 1.2
NI0SUB – 1.45e+10
NIGC – 1
NLX – 1.74e-07
NOFF – 1
NOFFFD – 1
NOIA – 0
NOIB – 0
NOIC – 8.75e+09
NOIF – 1
NRB – 1
NRD – 1
NRECF0 – 2
NRECF0D – 0
NRECR0 – 10
NRECR0D – 0
NRS – 1
NSD – 1e+20
NSEG – 1
NSUB – 6e+16
NTNOI – 1
NTOX – 1
NTRECF – 0
NTRECR – 0
NTUN – 10
NTUND – 0
OFF – 0
PA0 – 0
PA1 – 0
PA2 – 0
PABJTII – 0
PACDE – 0
PAELY – 0

425
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
PAGIDL – 0
PAGISL – 0
PAGS – 0
PAHLI – 0
PAHLID – 0
PAIGBCP2 – 0
PAIGC – 0
PAIGSD – 0
PALPHA0 – 0
PALPHAGB1 – 0
PALPHAGB2 – 0
PARAMCHK – 0
PAT – 0
PB0 – 0
PB1 – 0
PBETA0 – 0
PBETA1 – 0
PBETA2 – 0
PBETAGB1 – 0
PBETAGB2 – 0
PBGIDL – 0
PBGISL – 0
PBIGBCP2 – 0
PBIGC – 0
PBIGSD – 0
PBSWG – 0.7
PBSWGD – 0
PCBJTII – 0
PCDSC – 0
PCDSCB – 0
PCDSCD – 0
PCGDL – 0
PCGIDL – 0
PCGISL – 0
PCGSL – 0
PCIGBCP2 – 0

426
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
PCIGC – 0
PCIGSD – 0
PCIT – 0
PCKAPPA – 0
PCLM – 1.3
PD – 0
PDBCP – 0
PDELTA – 0
PDELVT – 0
PDIBLC1 – 0.39
PDIBLC2 – 0.0086
PDIBLCB – 0
PDITS – 1e-20
PDITSD – 0
PDITSL – 0
PDROUT – 0
PDSUB – 0
PDVT0 – 0
PDVT0W – 0
PDVT1 – 0
PDVT1W – 0
PDVT2 – 0
PDVT2W – 0
PDVTP0 – 0
PDVTP1 – 0
PDVTP2 – 0
PDVTP3 – 0
PDVTP4 – 0
PDWB – 0
PDWG – 0
PEBJTII – 0
PEGIDL – 0
PEGISL – 0
PESATII – 0
PETA0 – 0
PETA0CV – 0

427
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
PETAB – 0
PETABCV – 0
PEU – 0
PFBJTII – 0
PFGIDL – 0
PFGISL – 0
PFPROUT – 0
PHIG – 4.05
PIDBJT – 0
PIDDIF – 0
PIDREC – 0
PIDTUN – 0
PIGCD – 1
PISBJT – 0
PISDIF – 0
PISREC – 0
PISTUN – 0
PK1 – 0
PK1W1 – 0
PK1W2 – 0
PK2 – 0
PK3 – 0
PK3B – 0
PKB1 – 0
PKETA – 0
PKETAS – 0
PKGIDL – 0
PKGISL – 0
PKT1 – 0
PKT1L – 0
PKT2 – 0
PKU0 – 0
PKVTH0 – 0
PLBJT0 – 0
PLII – 0
PLPE0 – 0

428
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
PLPEB – 0
PMBJTII – 0
PMINV – 0
PMINVCV – 0
PMOIN – 0
PNBJT – 0
PNCH – 0
PNDIF – 0
PNDIODE – 0
PNDIODED – 0
PNFACTOR – 0
PNGATE – 0
PNGIDL – 0
PNIGC – 0
PNLX – 0
PNOFF – 0
PNRECF0 – 0
PNRECF0D – 0
PNRECR0 – 0
PNRECR0D – 0
PNSD – 0
PNSUB – 0
PNTRECF – 0
PNTRECR – 0
PNTUN – 0
PNTUND – 0
POXEDGE – 1
PPCLM – 0
PPDIBLC1 – 0
PPDIBLC2 – 0
PPDIBLCB – 0
PPDITS – 0
PPDITSD – 0
PPIGCD – 0
PPOXEDGE – 0
PPRT – 0

429
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
PPRWB – 0
PPRWG – 0
PPVAG – 0
PRDSW – 0
PRDW – 0
PRGIDL – 0
PRGISL – 0
PRSW – 0
PRT – 0
PRWB – 0
PRWG – 0
PS – 0
PSBCP – 0
PSII0 – 0
PSII1 – 0
PSII2 – 0
PSIID – 0
PU0 – 0
PUA – 0
PUA1 – 0
PUB – 0
PUB1 – 0
PUC – 0
PUC1 – 0
PUCS – 0
PUCSTE – 0
PUD – 0
PUD1 – 0
PUTE – 0
PVABJT – 0
PVAG – 0
PVBCI – 0
PVDSATII0 – 0
PVFB – 0
PVOFF – 0
PVOFFCV – 0

430
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
PVREC0 – 0
PVREC0D – 0
PVSAT – 0
PVSDFB – 0
PVSDTH – 0
PVTH0 – 0
PVTUN0 – 0
PVTUN0D – 0
PW0 – 0
PWR – 0
PXBJT – 0
PXDIF – 0
PXDIFD – 0
PXJ – 0
PXRCRG1 – 0
PXRCRG2 – 0
PXREC – 0
PXRECD – 0
PXTUN – 0
PXTUND – 0
RBDB – 50
RBODY – 0
RBODYMOD – 0
RBSB – 50
RBSH – 0
RDSMOD – 0
RDSW – 100
RDW – 50
RDWMIN – 0
RGATEMOD – 0
RGIDL – 1
RGISL – 0
RHALO – 1e+15
RNOIA – 0.577
RNOIB – 0.37
RSH – 0

431
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
RSHG – 0.1
RSW – 50
RSWMIN – 0
RTH0 – 0
SA – 0
SAREF – 1e-06
SB – 0
SBREF – 1e-06
SD – 0
SHMOD – 0
SII0 – 0.5
SII1 – 0.1
SII2 – 0
SIID – 0
SOIMOD – 0
STETA0 – 0
STETA0CV – 0
STK2 – 0
TBGASUB – 0.000702
TBGBSUB – 1108
TBOX – 3e-07
TCJSWG – 0
TCJSWGD – 0
TEMPEOT – 300.15
TII – 0
TKU0 – 0
TNOIA – 1.5
TNOIB – 3.5
TNOIMOD – 0
TNOM – 27
TOX – 1e-08
TOXM – 0
TOXP – 0
TOXQM – 0
TOXREF – 2.5e-09
TPBSWG – 0

432
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
TPBSWGD – 0
TSI – 1e-07
TT – 1e-12
TVBCI – 0
TYPE – 1
U0 – 0
UA – 2.25e-09
UA1 – 4.31e-09
UB – 5.87e-19
UB1 – -7.61e-18
UC – 0
UC1 – 0
UCS – 0
UCSTE – -0.004775
UD – 0
UD1 – 0
UTE – -1.5
VABJT – 10
VBCI – 0
VBM – -3
VBS0FD – 0.5
VBS0PD – 0
VBSA – 0
VBSUSR – 0
VBX – 0
VDDEOT – 0
VDSATII0 – 0.9
VECB – 0.026
VERSION – 4.4
VEVB – 0.075
VFB – -1
VGB1 – 300
VGB2 – 17
VGSTCVMOD – 0
VOFF – -0.08
VOFFCV – 0

433
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
VOFFFD – 0
VOXH – 5
VREC0 – 0
VREC0D – 0
VSAT – 80000
VSCE – 0
VSDFB – 0
VSDTH – 0
VTH0 – 0
VTHO – 0
VTM00 – 0.026
VTUN0 – 0
VTUN0D – 0
W – 5e-06
W0 – 2.5e-06
W0FLK – 1e-05
WA0 – 0
WA1 – 0
WA2 – 0
WABJTII – 0
WACDE – 0
WAELY – 0
WAGIDL – 0
WAGISL – 0
WAGS – 0
WAHLI – 0
WAHLID – 0
WAIGBCP2 – 0
WAIGC – 0
WAIGSD – 0
WALPHA0 – 0
WALPHAGB1 – 0
WALPHAGB2 – 0
WAT – 0
WB0 – 0
WB1 – 0

434
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
WBETA0 – 0
WBETA1 – 0
WBETA2 – 0
WBETAGB1 – 0
WBETAGB2 – 0
WBGIDL – 0
WBGISL – 0
WBIGBCP2 – 0
WBIGC – 0
WBIGSD – 0
WCBJTII – 0
WCDSC – 0
WCDSCB – 0
WCDSCD – 0
WCGDL – 0
WCGIDL – 0
WCGISL – 0
WCGSL – 0
WCIGBCP2 – 0
WCIGC – 0
WCIGSD – 0
WCIT – 0
WCKAPPA – 0
WDELTA – 0
WDELVT – 0
WDROUT – 0
WDSUB – 0
WDVT0 – 0
WDVT0W – 0
WDVT1 – 0
WDVT1W – 0
WDVT2 – 0
WDVT2W – 0
WDVTP0 – 0
WDVTP1 – 0
WDVTP2 – 0

435
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
WDVTP3 – 0
WDVTP4 – 0
WDWB – 0
WDWG – 0
WEBJTII – 0
WEFFEOT – 10
WEGIDL – 0
WEGISL – 0
WESATII – 0
WETA0 – 0
WETA0CV – 0
WETAB – 0
WETABCV – 0
WEU – 0
WFBJTII – 0
WFGIDL – 0
WFGISL – 0
WFPROUT – 0
WIDBJT – 0
WIDDIF – 0
WIDREC – 0
WIDTUN – 0
WINT – 0
WISBJT – 0
WISDIF – 0
WISREC – 0
WISTUN – 0
WK1 – 0
WK1W1 – 0
WK1W2 – 0
WK2 – 0
WK3 – 0
WK3B – 0
WKB1 – 0
WKETA – 0
WKETAS – 0

436
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
WKGIDL – 0
WKGISL – 0
WKT1 – 0
WKT1L – 0
WKT2 – 0
WKU0 – 0
WKVTH0 – 0
WL – 0
WLBJT0 – 0
WLC – 0
WLII – 0
WLN – 1
WLOD – 0
WLODKU0 – 0
WLODVTH – 0
WLPE0 – 0
WLPEB – 0
WMBJTII – 0
WMINV – 0
WMINVCV – 0
WMOIN – 0
WNBJT – 0
WNCH – 0
WNDIF – 0
WNDIODE – 0
WNDIODED – 0
WNFACTOR – 0
WNGATE – 0
WNGIDL – 0
WNIGC – 0
WNLX – 0
WNOFF – 0
WNRECF0 – 0
WNRECF0D – 0
WNRECR0 – 0
WNRECR0D – 0

437
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
WNSD – 0
WNSUB – 0
WNTRECF – 0
WNTRECR – 0
WNTUN – 0
WNTUND – 0
WPCLM – 0
WPDIBLC1 – 0
WPDIBLC2 – 0
WPDIBLCB – 0
WPDITS – 0
WPDITSD – 0
WPIGCD – 0
WPOXEDGE – 0
WPRT – 0
WPRWB – 0
WPRWG – 0
WPVAG – 0
WR – 1
WRDSW – 0
WRDW – 0
WRGIDL – 0
WRGISL – 0
WRSW – 0
WSII0 – 0
WSII1 – 0
WSII2 – 0
WSIID – 0
WTH0 – 0
WU0 – 0
WUA – 0
WUA1 – 0
WUB – 0
WUB1 – 0
WUC – 0
WUC1 – 0

438
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
WUCS – 0
WUCSTE – 0
WUD – 0
WUD1 – 0
WUTE – 0
WVABJT – 0
WVBCI – 0
WVDSATII0 – 0
WVFB – 0
WVOFF – 0
WVOFFCV – 0
WVREC0 – 0
WVREC0D – 0
WVSAT – 0
WVSDFB – 0
WVSDTH – 0
WVTH0 – 0
WVTUN0 – 0
WVTUN0D – 0
WW – 0
WW0 – 0
WWC – 0
WWL – 0
WWLC – 0
WWN – 1
WWR – 0
WXBJT – 0
WXDIF – 0
WXDIFD – 0
WXJ – 0
WXRCRG1 – 0
WXRCRG2 – 0
WXREC – 0
WXRECD – 0
WXTUN – 0
WXTUND – 0

439
 Table 2-111. BSIM-SOI 4.5.0 Device Model Parameters
Parameter Description Units Default
XBJT – 1
XDIF – 0
XDIFD – 0
XGL – 0
XGW – 0
XJ – 0
XPART – 0
XRCRG1 – 12
XRCRG2 – 1
XREC – 1
XRECD – 0
XT – 1.55e-07
XTUN – 0
XTUND – 0

Table 2-112. MOSFET level 70450 Output Variables

Parameter Description Units Default
VDSAT VDSAT – none
VTH VTH – none
IDS IDS – none
GM GM – none
GDS GDS – none
GMBS GMBS – none
IC IC – none
IBD IBD – none
IBS IBS – none
IGIDL IGIDL – none
IGISL IGISL – none
IGS IGS – none
IGD IGD – none
IGB IGB – none
IGCS IGCS – none
IGCD IGCD – none
CGG CGG – none
CGS CGS – none

440
 Table 2-112. MOSFET level 70450 Output Variables
Parameter Description Units Default
CGD CGD – none
CBG CBG – none
CBD CBD – none
CBS CBS – none
CDG CDG – none
CDD CDD – none
CDS CDS – none
CAPBD CAPBD – none
CAPBS CAPBS – none
QG QG – none
QB QB – none
QD QD – none
QS QS – none
QJD QJD – none
QJS QJS – none

441
2.3.20.10. Level 77 MOSFET Tables (BSIM6 version 6.1.1)

Xyce includes the BSIM6 MOSFET model, version 6.1.1. Full documentation of the BSIM6 is available at
its web site, http://bsim.berkeley.edu/models/bsim6/. Instance and model parameters for the
BSIM6 are given in tables 2-113 and 2-114. These tables are generated directly from information present in
the original Verilog-A implementation of the BSIM6, and lack many descriptions for the parameters.
Consult the BSIM6 technical manual from the BSIM group for further details about these parameters.
Beginning with version 7.2 of Xyce, an optional fifth node may be specified for BSIM6 devices. If
specified, it is the temperature node, which is used by the self-heating model and is internal if not specified
on the instance line.
The BSIM6 device supports output of the internal variables in table 2-115 on the .PRINT line of a netlist.
To access them from a print line, use the syntax N(<instance>:<variable>) where “<instance>”
refers to the name of the specific level 77 M device in your netlist.

Table 2-113. BSIM6 Device Instance Parameters

Parameter Description Units Default
AD Drain to Substrate Junction Area — 0
AS Source to Substrate Junction Area — 0
GEOMOD Geo dependent parasitics model — 0
L m 1e-05
M multiplicity factor — 1
MINZ Minimize either D or S — 0
NF Number of fingers — 1
NGCON Number of gate contacts — 1
NRD Number of squares in drain — 1
NRS Number of squares in source — 1
PD Drain to Substrate Junction Perimeter — 0
PS Source to Substrate Junction Perimeter — 0
RBDB Resistance between bNode and dbNode — 50
RBODYMOD Distributed body R model — 0
RBPB Resistance between bNodePrime and bNode — 50
RBPD Resistance between bNodePrime and bNode — 50
RBPS Resistance between bNodePrime and sbNode — 50
RBSB Resistance between bNode and sbNode — 50
RGATEMOD Gate resistance model selector — 0
Geometry-dependent source/drain resistance, 0:
RGEOMOD — 0
RSH-based, 1: Holistic
SA Distance between OD edge from Poly from one side — 0
SB Distance between OD edge from Poly from other side — 0
SC Distance to a single well edge if <=0.0, turn off WPE — 0

442
 Table 2-113. BSIM6 Device Instance Parameters
Parameter Description Units Default
SCA — 0
SCB — 0
SCC — 0
SD Distance between neighboring fingers — 0
VFBSDOFF — 0
W Total width including fingers m 1e-05
XGW Dist from gate contact center to dev edge [m] m 0

Table 2-114. BSIM6 Device Model Parameters

Parameter Description Units Default
Non-saturation effect parameter for strong inversion
A1 — 0
region
A11 Temperature dependence of A1 — 0
Non-saturation effect parameter for moderate
A2 — 0
inversion region
A21 Temperature dependence of A2 — 0
Quantum mechanical effect prefactor cum switch in
ADOS — 0
inversion
AGIDL pre-exponential coeff. for GIDL in mho — 0
AGIDLL Length dependence coefficient of AGIDL — 0
AGIDLW Width dependence coefficient of AGIDL — 0
AGISL pre-exponential coeff. for GISL — 0
AGISLL Length dependence coefficient of AGISL — 0
AGISLW Width dependence coefficient of AGISL — 0
AIGBACC Parameter for Igb — 0.0136
AIGBINV Parameter for Igb — 0.0111
AIGC Parameter for Igc — 0
AIGCL Length dependence coefficient of AIGC — 0
AIGCW Width dependence coefficient of AIGC — 0
AIGD Parameter for Igs d — 0
AIGDL Length dependence coefficient of AIGD — 0
AIGDW Width dependence coefficient of AIGD — 0
AIGS Parameter for Igs d — 0
AIGSL Length dependence coefficient of AIGS — 0
AIGSW Width dependence coefficient of AIGS — 0
ALPHA0 first parameter of Iii, m/V — 0

443
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
ALPHA0L Length dependence coefficient of ALPHA0 — 0
ALPHA0LEXP Length dependence exponent coefficient of ALPHA0 — 1
0: Asymmetry Model turned off - forward mode
ASYMMOD — 0
parameters used, 1: Asymmetry Model turned on
AT Temperature coefficient for saturation velocity — -0.00156
ATL Length Scaling parameter for AT — 0
Charge centroid parameter - slope of CV curve under
BDOS — 1
QME in inversion
BETA0 Vds dependent parameter of Iii, 1/V — 0
BG0SUB Band gap of substrate at 300.15K — 1.17
BGIDL exponential coeff. for GIDL in — 2.3e+09
BGISL exponential coeff. for GISL — 0
BIGBACC Parameter for Igb — 0.00171
BIGBINV Parameter for Igb — 0.000949
BIGC Parameter for Igc — 0
BIGD Parameter for Igs d — 0
BIGS Parameter for Igs d — 0
Unit of L and W for Binning, 1 : micro-meter, 0 :
BINUNIT — 1
default
BVD Drain diode breakdown voltage — 0
BVS Source diode breakdown voltage — 10
CDSCB body-bias sensitivity of sub-threshold slope — 0
CDSCBEDGE — 0
CDSCBL Length dependence coefficient of CDSCB — 0
CDSCBLEXP Length dependence exponent coefficient of CDSCB — 1
CDSCD drain-bias sensitivity of sub-threshold slope — 1e-09
CDSCDEDGE — 0
CDSCDL Length dependence coefficient of CDSCD — 0
CDSCDLEXP Length dependence exponent coefficient of CDSCD — 1
CDSCDLR Length dependence coefficient of CDSCD — 0
CDSCDR drain-bias sensitivity of sub-threshold slope — 0
CF Outer Fringe Cap F 0
CFRCOEFF Coefficient for Outer Fringe Cap — 1
CGBO Gate - Body overlap capacitance — 0
CGDL — 0
CGDO Gate - Drain overlap capacitance — 0
CGIDL exponential coeff. for GIDL in V/m — 0.5

444
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
CGISL exponential coeff. for GISL — 0
CGSL — 0
CGSO Gate - Source overlap capacitance — 0
CIGBACC Parameter for Igb — 0.075
CIGBINV Parameter for Igb — 0.006
CIGC Parameter for Igc — 0
CIGD Parameter for Igs d — 0
CIGS Parameter for Igs d — 0
CIT parameter for interface trap — 0
CITEDGE — 0
CJD Unit area drain-side junction capacitance at zero bias — 0
CJS Unit area source-side junction capacitance at zero bias — 0.0005
Unit length drain-side sidewall junction capacitance at
CJSWD — 0
zero bias
Unit length drain-side gate sidewall junction
CJSWGD — 0
capacitance at zero bias
Unit length source-side gate sidewall junction
CJSWGS — 0
capacitance at zero bias
Unit length source-side sidewall junction capacitance
CJSWS — 5e-10
at zero bias
CKAPPAD — 0.6
CKAPPAS — 0.6
0: Use Bias-independent Overlap Capacitances, 1:
COVMOD — 0
Use Bias-dependent Overlap Capacitances
CTH0 Thermal capacitance — 1e-05
CVMOD 0: Consistent IV-CV, 1: Different IV-CV — 0
DELTA Smoothing function factor for Vdsat — 0.125
DELTAL Length dependence coefficient of DELTA — 0
DELTALEXP Length dependence exponent coefficient of DELTA — 1
DELVT0 — 0
DGAMMAEDGE — 0
DGAMMAEDGEL — 0
DGAMMAEDGELEXP — 1
DLBIN — 0
DLC delta L for CV — 0
DLCIG Delta L for Ig model [m] m 0
DLCIGD Delta L for Ig model [m] m 0

445
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
DMCG Distance of Mid-Contact to Gate edge m 0
DMCGT Dist of Mid-Contact to Gate edge in Test m 0
DMCI Distance of Mid-Contact to Isolation m 0
DMDG Distance of Mid-Diffusion to Gate edge m 0
DSUB Length scaling exponent for DIBL — 1
DTEMP Offset of Device Temperature — 0
DTOX Difference between effective dielectric thickness — 0
DVT0EDGE — 2.2
DVT1EDGE — 0.53
DVT2EDGE Body-bias coefficient for SCE effect for Edge FET — 0
DVTEDGE Vth shift for Edge FET — 0
DVTP0 DITS — 0
DVTP1 DITS — 0
DVTP2 DITS — 0
DVTP3 DITS — 0
DVTP4 DITS — 0
DVTP5 DITS — 0
DWBIN — 0
DWC delta W for CV — 0
DWJ delta W for S/D junctions — 0
EASUB Electron affinity of substrate — 4.05
EDGEFET 0: Edge FET Model Off, 1: Edge FET Model ON — 1
EF Flicker Noise frequency exponent — 1
EGIDL band bending parameter for GIDL — 0.8
EGISL band bending parameter for GISL — 0
EIGBINV Parm for the Si bandgap for Igbinv — 1.1
EM — 4.1e+07
EPSROX Relative dielectric constant of the gate dielectric — 3.9
EPSRSUB Relative dielectric constant of the channel material — 11.9
ETA0 DIBL coefficient — 0.08
ETA0EDGE — 0
ETA0R DIBL coefficient — 0
ETAB Body bias coefficient for subthreshold DIBL effect — -0.07
ETABEDGE — 0
ETABEXP Exponent coefficient of ETAB — 1
ETAMOB Effective field parameter (should be kept close to 1) — 1

446
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
ETAQM Bulk charge coefficient for charge centroid in inversion — 0.54
EU Mobility reduction exponent — 1.5
EUL Length dependence coefficient of EU — 0
EULEXP Length dependence exponent coefficient of EU — 1
EUW Width dependence coefficient of EU — 0
EUWEXP Width dependence exponent coefficient of EU — 1
EUWL Width-Length dependence coefficient of EU — 0
EUWLEXP Width-Length dependence coefficient of EU — 1
FPROUT — 0
FPROUTL Length dependence coefficient of FPROUT — 0
FPROUTLEXP Length dependence exponent coefficient of FPROUT — 1
GBMIN Minimum body conductance — 1e-12
GEOMOD Geo dependent parasitics model — 0
GIDLMOD 0: Turn off GIDL Current, 1: Turn on GIDL Current — 0
IGBMOD 0: Turn off Igb, 1: Turn on Igb — 0
IGCMOD 0: Turn off Igc, Igs and Igd, 1: Turn on Igc, Igs and Igd — 0
IGT Gate Current Temperature Dependence — 2.5
IIT Temperature coefficient for BETA0 — 0
IJTHDFWD Forward drain diode breakdown limiting current — 0
IJTHDREV Reverse drain diode breakdown limiting current — 0
IJTHSFWD Forward source diode breakdown limiting current — 0.1
IJTHSREV Reverse source diode breakdown limiting current — 0.1
Bottom drain junction reverse saturation current
JSD — 0
density
Bottom source junction reverse saturation current
JSS — 0.0001
density
Unit length reverse saturation current for sidewall
JSWD — 0
drain junction
Unit length reverse saturation current for gate-edge
JSWGD — 0
sidewall drain junction
Unit length reverse saturation current for gate-edge
JSWGS — 0
sidewall source junction
Unit length reverse saturation current for sidewall
JSWS — 0
source junction
Bottom drain junction trap-assisted saturation current
JTSD — 0
density
Bottom source junction trap-assisted saturation
JTSS — 0
current density

447
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
Unit length trap-assisted saturation current for
JTSSWD — 0
sidewall drain junction
Unit length trap-assisted saturation current for
JTSSWGD — 0
gate-edge sidewall drain junction
Unit length trap-assisted saturation current for
JTSSWGS — 0
gate-edge sidewall source junction
Unit length trap-assisted saturation current for
JTSSWS — 0
sidewall source junction
JTWEFF Trap assisted tunneling current width dependence — 0
First-order body-bias Vth shift due to Vertical
K1 — 0
Non-uniform doping
K1L Length dependence coefficient of K1 — 0
K1LEXP Length dependence exponent coefficient of K1 — 1
K1W Width dependence coefficient of K1 — 0
K1WEXP Width dependence exponent coefficient of K1 — 1
K1WL Width-Length dependence coefficient of K1 — 0
K1WLEXP Width-Length dependence exponent coefficient of K1 — 1
K2 Vth shift due to Vertical Non-uniform doping — 0
K2L Length dependence coefficient of K2 — 0
K2LEXP Length dependence exponent coefficient of K2 — 1
K2W Width dependence coefficient of K2 — 0
K2WE — 0
K2WEXP Width dependence exponent coefficient of K2 — 1
K2WL Width-Length dependence coefficient of K2 — 0
K2WLEXP Width-Length dependence exponent coefficient of K2 — 1
KT1 Temperature coefficient for Vth — -0.11
KT1EDGE — 0
KT1EXP Temperature coefficient for Vth — 1
KT1EXPEDGE — 0
KT1L Temperature coefficient for Vth — 0
KT1LEDGE — 0
KT2 Temperature coefficient for Vth — 0.022
KT2EDGE — 0
Mobility degradation/enhancement Parameter for
KU0 — 0
Stress Effect
KU0WE — 0
Saturation Velocity degradation/enhancement
KVSAT — 0
Parameter for Stress Effect

448
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
KVTH0 Threshold Shift parameter for stress effect — 0
KVTH0WE — 0
L m 1e-05
LA1 — 0
LA11 — 0
LA2 — 0
LA21 — 0
LAGIDL — 0
LAGISL — 0
LAIGBACC — 0
LAIGBINV — 0
LAIGC — 0
LAIGD — 0
LAIGS — 0
LALPHA0 — 0
LAT — 0
LBETA0 — 0
LBGIDL — 0
LBGISL — 0
LBIGBACC — 0
LBIGBINV — 0
LBIGC — 0
LBIGD — 0
LBIGS — 0
LCDSCB — 0
LCDSCD — 0
LCDSCDR — 0
LCF F 0
LCGDL — 0
LCGIDL — 0
LCGISL — 0
LCGSL — 0
LCIGBACC — 0
LCIGBINV — 0
LCIGC — 0
LCIGD — 0

449
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
LCIGS — 0
LCIT — 0
LCKAPPAD — 0
LCKAPPAS — 0
LDELTA — 0
LDLCIG — 0
LDLCIGD — 0
LDVTP0 — 0
LDVTP1 — 0
LDVTP2 — 0
LDVTP3 — 0
LDVTP4 — 0
LDVTP5 — 0
LEGIDL — 0
LEGISL — 0
LEIGBINV — 0
LETA0 — 0
LETA0R — 0
LETAB — 0
LEU — 0
LFPROUT — 0
LIGT — 0
LIIT — 0
LINT delta L for IV — 0
LINTNOI — 0
LK1 — 0
LK2 — 0
LK2WE — 0
LKT1 — 0
LKT2 — 0
LKU0 Length Dependence of KU0 — 0
LKU0WE — 0
LKVTH0 Length dependence of KVTH0 — 0
LKVTH0WE — 0
LL — 0
LLC — 0

450
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
LLN — 1
LLODKU0 Length Parameter for u0 stress effect — 0
LLODVTH Length Parameter for Vth stress effect — 0
LLONG L of extracted Long channel device m 1e-05
LMAX Maximum length for which this model should be used m 100
LMIN Minimum length for which this model should be used m 0
LMLT Length Shrinking Parameter — 1
LNDEP — 0
LNDEPCV — 0
LNFACTOR — 0
LNGATE — 0
LNIGBACC — 0
LNIGBINV — 0
LNSD — 0
LNTOX — 0
LODETA0 eta0 modification foator for stress effect — 0
LODK2 K2 shift modification factor for stress effect — 0
LP1 Mobility channel length exponential coefficent — 1e-08
LP2 Mobility channel length exponential coefficent — 1e-08
LPCLM — 0
LPCLMCV — 0
LPCLMR — 0
LPDIBLC — 0
LPDIBLCB — 0
LPDIBLCR — 0
LPDITS — 0
LPDITSD — 0
LPHIN — 0
LPOXEDGE — 0
LPRT — 0
LPRWB — 0
LPRWG — 0
LPSAT — 0
LPSATB — 0
LPSATR — 0
LPSCBE1 — 0

451
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
LPSCBE2 — 0
LPTWG — 0
LPTWGR — 0
LPTWGT — 0
LPVAG — 0
LRDSW — 0
LRDSWMIN — 0
LRDW — 0
LRDWMIN — 0
LRSW — 0
LRSWMIN — 0
LTGIDL — 0
LU0 — 0
LU0R — 0
LUA — 0
LUA1 — 0
LUAR — 0
LUC — 0
LUC1 — 0
LUCR — 0
LUCS — 0
LUCSR — 0
LUCSTE — 0
LUD — 0
LUD1 — 0
LUDR — 0
LUTE — 0
LVFB — 0
LVFBCV — 0
LVSAT — 0
LVSATCV — 0
LVSATR — 0
LW — 0
LWC — 0
LWL — 0
LWLC — 0

452
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
LWN — 1
LWR — 0
LXJ — 0
MJD Drain bottom junction capacitance grading coefficient — 0
Source bottom junction capacitance grading
MJS — 0.5
coefficient
Drain sidewall junction capacitance grading
MJSWD — 0
coefficient
Drain-side gate sidewall junction capacitance grading
MJSWGD — 0
coefficient
Source-side gate sidewall junction capacitance
MJSWGS — 0
grading coefficient
Source sidewall junction capacitance grading
MJSWS — 0.33
coefficient
MOBSCALE Mobility scaling model, 0: Old Model, 1: New Model — 0
MULU0 — 1
NDEP Channel Doping Concentration for IV m−3 1e+24
NDEPCV Channel Doping Concentration for CV m−3 0
NDEPCVL1 Length dependence coefficient of NDEPCV — 0
Length dependence coefficient of NDEPCV - For
NDEPCVL2 — 0
Short Channel Devices
NDEPCVLEXP1 Length dependence exponent coefficient of NDEPCV — 0
NDEPCVLEXP2 Length dependence exponent coefficient of NDEPCV — 0
NDEPCVW Width dependence coefficient of NDEPCV — 0
NDEPCVWEXP Width dependence exponent coefficient of NDEPCV — 0
NDEPCVWL Width-Length dependence coefficient of NDEPCV — 0
Width-Length dependence exponent coefficient of
NDEPCVWLEXP — 0
NDEPCV
NDEPL1 Length dependence coefficient of NDEP — 0
Length dependence of NDEP - For Short Channel
NDEPL2 — 0
Devices
NDEPLEXP1 Length dependence exponent coefficient of NDEP — 1
NDEPLEXP2 Length dependence exponent coefficient of NDEP — 2
NDEPW Width dependence coefficient of NDEP — 0
NDEPWEXP Width dependence exponent coefficient of NDEP — 1
NDEPWL Width-Length dependence coefficient of NDEP — 0
Width-Length dependence exponent coefficient of
NDEPWLEXP — 1
NDEP

453
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
NFACTOR Sub-threshold slope factor — 0
NFACTOREDGE — 0
NFACTORL Length dependence coefficient of NFACTOR — 0
Length dependence exponent coefficient of
NFACTORLEXP — 1
NFACTOR
NFACTORW Width dependence coefficient of NFACTOR — 0
NFACTORWEXP Width dependence exponent coefficient of NFACTOR — 1
NFACTORWL Width-Length dependence coefficient of NFACTOR — 0
Width-Length dependence exponent coefficient of
NFACTORWLEXP — 1
NFACTOR
NGATE Gate Doping Concentration m−3 5e+25
NGCON Number of gate contacts — 1
Intrinsic carrier concentration of the substrate at
NI0SUB m−3 1.1e+16
300.15K
NIGBACC Parameter for Igbacc slope — 1
NIGBINV Parameter for Igbinv slope — 3
NJD Drain junction emission coefficient — 0
NJS Source junction emission coefficient — 1
NJTS Non-ideality factor for JTSS — 20
NJTSD Non-ideality factor for JTSD — 0
NJTSSW Non-ideality factor for JTSSWS — 20
NJTSSWD Non-ideality factor for JTSSWD — 0
NJTSSWG Non-ideality factor for JTSSWGS — 20
NJTSSWGD Non-ideality factor for JTSSWGD — 0
NOIA — 6.25e+40
NOIB — 3.125e+25
NOIC — 8.75e+08
NSD S/D Doping Concentration m−3 1e+26
NTNOI — 1
NTOX Exponent for Tox ratio — 1
PA1 — 0
PA11 — 0
PA2 — 0
PA21 — 0
PAGIDL — 0
PAGISL — 0
PAIGBACC — 0

454
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
PAIGBINV — 0
PAIGC — 0
PAIGD — 0
PAIGS — 0
PALPHA0 — 0
PAT — 0
PBD Drain-side bulk junction built-in potential — 0
PBETA0 — 0
PBGIDL — 0
PBGISL — 0
PBIGBACC — 0
PBIGBINV — 0
PBIGC — 0
PBIGD — 0
PBIGS — 0
PBS Source-side bulk junction built-in potential — 1
Built-in potential for Drain-side sidewall junction
PBSWD — 0
capacitance
Built-in potential for Drain-side gate sidewall junction
PBSWGD — 0
capacitance
Built-in potential for Source-side gate sidewall
PBSWGS — 0
junction capacitance
Built-in potential for Source-side sidewall junction
PBSWS — 1
capacitance
PCDSCB — 0
PCDSCD — 0
PCDSCDR — 0
PCF F 0
PCGDL — 0
PCGIDL — 0
PCGISL — 0
PCGSL — 0
PCIGBACC — 0
PCIGBINV — 0
PCIGC — 0
PCIGD — 0
PCIGS — 0

455
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
PCIT — 0
PCKAPPAD — 0
PCKAPPAS — 0
PCLM CLM prefactor — 0
PCLMCV CLM parameter for CV — 0
PCLMCVL — 0
PCLMCVLEXP — 0
PCLMG CLM prefactor gate voltage dependence — 0
PCLML Length dependence coefficient of PCLM — 0
PCLMLEXP Length dependence exponent coefficient of PCLM — 1
PCLMR CLM prefactor — 0
PDELTA — 0
PDIBLC parameter for DIBL effect on Rout — 0
PDIBLCB parameter for DIBL effect on Rout — 0
PDIBLCL Length dependence coefficient of PDIBLC — 0
PDIBLCLEXP Length dependence exponent coefficient of PDIBLC — 1
PDIBLCLEXPR Length dependence exponent coefficient of PDIBLC — 0
PDIBLCLR Length dependence coefficient of PDIBLC — 0
PDIBLCR parameter for DIBL effect on Rout — 0
PDITS Coefficient for drain-induced Vth shifts — 0
PDITSD Vds dep of drain-induced Vth shifts — 0
PDITSL L dep of drain-induced Vth shifts — 0
PDLCIG — 0
PDLCIGD — 0
PDVTP0 — 0
PDVTP1 — 0
PDVTP2 — 0
PDVTP3 — 0
PDVTP4 — 0
PDVTP5 — 0
PEGIDL — 0
PEGISL — 0
PEIGBINV — 0
Whether PS/PD (when given) include gate-edge
PERMOD — 1
perimeter
PETA0 — 0

456
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
PETA0R — 0
PETAB — 0
PEU — 0
PFPROUT — 0
Nonuniform vertical doping effect on surface
PHIN V 0.045
potential, V
PIGCD Igc, S/D partition parameter — 1
PIGCDL Length dependence coefficient of PIGCD — 0
PIGT — 0
PIIT — 0
PK1 — 0
PK2 — 0
PK2WE — 0
PKT1 — 0
PKT2 — 0
PKU0 Cross Term Dependence of KU0 — 0
PKU0WE — 0
PKVTH0 Cross-term dependence of KVTH0 — 0
PKVTH0WE — 0
PNDEP — 0
PNDEPCV — 0
PNFACTOR — 0
PNGATE — 0
PNIGBACC — 0
PNIGBINV — 0
PNSD — 0
PNTOX — 0
POXEDGE Factor for the gate edge Tox — 1
PPCLM — 0
PPCLMCV — 0
PPCLMR — 0
PPDIBLC — 0
PPDIBLCB — 0
PPDIBLCR — 0
PPDITS — 0
PPDITSD — 0

457
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
PPHIN — 0
PPOXEDGE — 0
PPRT — 0
PPRWB — 0
PPRWG — 0
PPSAT — 0
PPSATB — 0
PPSATR — 0
PPSCBE1 — 0
PPSCBE2 — 0
PPTWG — 0
PPTWGR — 0
PPTWGT — 0
PPVAG — 0
PRDSW — 0
PRDSWMIN — 0
PRDW — 0
PRDWMIN — 0
PRSW — 0
PRSWMIN — 0
PRT Temperature coefficient for resistance — 0
PRWB Body bias dependence of resistance — 0
PRWBL Length dependence coefficient of PPRWB — 0
PRWBLEXP Length dependence exponent coefficient of PPRWB — 1
PRWG gate bias dependence of S/D extension resistance — 1
PSAT Gmsat variation with gate bias — 1
PSATB Body bias effect on Idsat — 0
PSATL — 0
PSATLEXP — 1
PSATR Gmsat variation with gate bias — 0
PSATX — 1
PSCBE1 Substrate current body-effect coeff — 4.24e+08
PSCBE2 Substrate current body-effect coeff — 1e-08
PTGIDL — 0
PTWG Idsat variation with gate bias — 0
PTWGL Length dependence coefficient of PTWG — 0

458
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
PTWGLEXP Length dependence exponent coefficient of PTWG — 1
PTWGLEXPR Length dependence exponent coefficient of PTWG — 0
PTWGLR Length dependence coefficient of PTWG — 0
PTWGR Idsat variation with gate bias — 0
PTWGT Temperature coefficient for PTWG — 0
PTWGTL Length Scaling parameter for PTWGT — 0
PU0 — 0
PU0R — 0
PUA — 0
PUA1 — 0
PUAR — 0
PUC — 0
PUC1 — 0
PUCR — 0
PUCS — 0
PUCSR — 0
PUCSTE — 0
PUD — 0
PUD1 — 0
PUDR — 0
PUTE — 0
PVAG Vg dependence of early voltage — 1
PVFB — 0
PVFBCV — 0
PVSAT — 0
PVSATCV — 0
PVSATR — 0
PWR — 0
PXJ — 0
Charge centroid parameter - starting point for QME in
QM0 — 0.001
inversion
RBDB Resistance between bNode and dbNode — 50
RBDBX0 Scaling prefactor for RBDBX — 100
RBDBY0 Scaling prefactor for RBDBY — 100
RBODYMOD Distributed body R model — 0
RBPB Resistance between bNodePrime and bNode — 50

459
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
RBPBX0 — 100
RBPBXL Length Scaling parameter for RBPBX — 0
RBPBXNF Number of fingers Scaling parameter for RBPBX — 0
RBPBXW Width Scaling parameter for RBPBX — 0
RBPBY0 Scaling prefactor for RBPBY — 100
RBPBYL Length Scaling parameter for RBPBY — 0
RBPBYNF Number of fingers Scaling parameter for RBPBY — 0
RBPBYW Width Scaling parameter for RBPBY — 0
RBPD Resistance between bNodePrime and bNode — 50
RBPD0 — 50
RBPDL Length Scaling parameter for RBPD — 0
RBPDNF Number of fingers Scaling parameter for RBPD — 0
RBPDW Width Scaling parameter for RBPD — 0
RBPS Resistance between bNodePrime and sbNode — 50
RBPS0 Scaling prefactor for RBPS 50 Ohms — 50
RBPSL — 0
RBPSNF — 0
RBPSW — 0
RBSB Resistance between bNode and sbNode — 50
RBSBX0 Scaling prefactor for RBSBX — 100
RBSBY0 Scaling prefactor for RBSBY — 100
RBSDBXL Length Scaling parameter for RBSBX and RBDBX — 0
Number of fingers Scaling parameter for RBSBX and
RBSDBXNF — 0
RBDBX
RBSDBXW Width Scaling parameter for RBSBX and RBDBX — 0
RBSDBYL Length Scaling parameter for RBSBY and RBDBY — 0
Number of fingers Scaling parameter for RBSBY and
RBSDBYNF — 0
RBDBY
RBSDBYW Width Scaling parameter for RBSBY and RBDBY — 0
0: Internal bias dependent and external bias
RDSMOD independent s/d resistance model, 1: External s/d — 0
resistance model, 2: Internal s/d resistance model
RDSW zero bias Resistance (RDSMOD=0 and RDSMOD=2) — 20
Geometrical scaling of RDSW (RDSMOD=0 and
RDSWL — 0
RDSMOD=2)
Geometrical scaling of RDSW (RDSMOD=0 and
RDSWLEXP — 1
RDSMOD=2)

460
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
S/D Resistance per unit width at high Vgs
RDSWMIN — 0
(RDSMOD=0 and RDSMOD=2)
RDW zero bias Drain Resistance (RDSMOD=1) — 0
RDWL Geometrical scaling of RDW (RDSMOD=1) — 0
RDWLEXP Geometrical scaling of RDW (RDSMOD=1) — 0
Drain Resistance per unit width at high Vgs
RDWMIN — 0
(RDSMOD=1)
RGATEMOD Gate resistance model selector — 0
Geometry-dependent source/drain resistance, 0:
RGEOMOD — 0
RSH-based, 1: Holistic
RNOIA TNOIMOD = 1 — 0.577
RNOIB TNOIMOD = 1 — 0.5164
RNOIC TNOIMOD = 1 — 0.395
RSH Source-drain sheet resistance — 0
RSHG Gate sheet resistance — 0.1
RSW zero bias Source Resistance (RDSMOD=1) — 10
RSWL Geometrical scaling of RSW (RDSMOD=1) — 0
RSWLEXP Geometrical scaling of RSW (RDSMOD=1) — 1
Source Resistance per unit width at high Vgs
RSWMIN — 0
(RDSMOD=1)
RTH0 Thermal resistance — 0
SA Distance between OD edge from Poly from one side — 0
Reference distance between OD edge from Poly from
SAREF — 1e-06
one side
SB Distance between OD edge from Poly from other side — 0
Reference distance between OD edge from Poly from
SBREF — 1e-06
other side
SC Distance to a single well edge if <=0.0, turn off WPE — 0
SCA — 0
SCB — 0
SCC — 0
SCREF — 1e-06
SD Distance between neighboring fingers — 0
SHMOD 0 : Self heating model OFF, 1 : Self heating model ON — 0
STETA0 eta0 shift related to Vth0 change — 0
STK2 K2 shift factor related to Vth change — 0
TBGASUB Bandgap Temperature Coefficient — 0.000473
TBGBSUB Bandgap Temperature Coefficient — 636

461
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
TCJ Temperature coefficient for CJS/CJD — 0
TCJSW Temperature coefficient for CJSWS/CJSWD — 0
TCJSWG Temperature coefficient for CJSWGS/CJSWGD — 0
TDELTA Temperature coefficient for DELTA — 0
TETA0 Temperature coefficient for ETA0 — 0
TETA0EDGE — 0
TGIDL Temperature coefficient for GIDL/GISL — 0
TKU0 Temperature Coefficient for KU0 — 0
TNFACTOR Temperature exponent for NFACTOR — 0
TNFACTOREDGE — 0
TNJTS Temperature coefficient for NJTS — 0
TNJTSD Temperature coefficient for NJTSD — 0
TNJTSSW Temperature coefficient for NJTSSW — 0
TNJTSSWD Temperature coefficient for NJTSSWD — 0
TNJTSSWG Temperature coefficient for NJTSSWG — 0
TNJTSSWGD Temperature coefficient for NJTSSWGD — 0
TNOIA TNOIMOD = 1 — 0
TNOIB TNOIMOD = 1 — 0
TNOIC Correlation coefficient — 0
TNOIMOD Thermal noise model selector — 0
TNOM Temperature at which the model was extracted — 27
TOXE Effective gate dielectric thickness relative to SiO2, m m 3e-09
Physical gate dielectric thickness,If not given, TOXP
TOXP m 0
is calculated from TOXE and DTOX
TOXREF Target tox value m 3e-09
TPB Temperature coefficient for PBS/PBD — 0
TPBSW Temperature coefficient for PBSWS/PBSWD — 0
TPBSWG Temperature coefficient for PBSWGS/PBSWGD — 0
TYPE — 1
U0 — 0.067
U0L Length dependence coefficient of U0L — 0
U0LEXP Length dependence exponent coefficient of U0L — 1
U0R — 0
UA Mobility reduction coefficient — 0.001
UA1 Temperature coefficient for UA — 0.001
UA1L Length Scaling parameter for UA1 — 0

462
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
UAL Length dependence coefficient of UA — 0
UALEXP Length dependence exponent coefficient of UA — 1
UAR Mobility reduction coefficient — 0
UAW Width dependence coefficient of UA — 0
UAWEXP Width dependence exponent coefficient of UA — 1
UAWL Width-Length dependence coefficient of UA — 0
UAWLEXP Width-Length dependence coefficient of UA — 1
UC Mobility reduction with body bias — 0
UC1 Temperature coefficient for UC — 5.6e-11
UCL Length dependence coefficient of UC — 0
UCLEXP Length dependence exponent coefficient of UC — 1
UCR Mobility reduction with body bias — 0
UCS Coulombic scattering parameter — 2
UCSR Coulombic scattering parameter — 0
UCSTE Temperature coefficient for UCS — -0.004775
UCW Width dependence coefficient of UC — 0
UCWEXP Width dependence exponent coefficient of UC — 1
UCWL Width-Length dependence coefficient of UC — 0
UCWLEXP Width-Length dependence exponent coefficient of UC — 1
UD Coulombic scattering parameter — 0.001
UD1 Temperature coefficient for UD — 0
UD1L Length Scaling parameter for UD1 — 0
UDL Length dependence coefficient of UD — 0
UDLEXP Length dependence exponent coefficient of UD — 1
UDR Coulombic scattering parameter — 0
UP1 Mobility channel length coefficent — 0
UP2 Mobility channel length coefficent — 0
UTE Mobility temperature exponent — -1.5
UTEL Length Scaling parameter for UTE — 0
VFB Flat band voltage V -0.5
VFBCV Flat band voltage for CV — 0
VFBCVL Length dependence coefficient of VFBCV — 0
VFBCVLEXP Length dependence exponent coefficient of VFBCV — 1
VFBCVW Width dependence coefficient of VFBCV — 0
VFBCVWEXP Width dependence exponent coefficient of VFBCV — 1
VFBCVWL Width-Length dependence coefficient of VFBCV — 0

463
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
VFBCVWLEXP Width-Length dependence coefficient of VFBCV — 1
VFBSDOFF — 0
VSAT Saturation Velocity — 100000
VSATCV VSAT parameter for CV — 0
VSATCVL — 0
VSATCVLEXP — 0
VSATCVW — 0
VSATCVWEXP — 0
VSATCVWL — 0
VSATCVWLEXP — 0
VSATL Length dependence coefficient of of VSAT — 0
VSATLEXP Length dependence exponent coefficient of VSAT — 1
VSATR Saturation Velocity — 0
VSATW Width dependence coefficient of of VSAT — 0
VSATWEXP Width dependence exponent coefficient of of VSAT — 1
VSATWL Width-Length dependence coefficient of of VSAT — 0
Width-Length dependence exponent coefficient of of
VSATWLEXP — 1
VSAT
Bottom drain junction trap-assisted current voltage
VTSD — 0
dependent parameter
Bottom source junction trap-assisted current voltage
VTSS — 10
dependent parameter
Unit length trap-assisted current voltage dependent
VTSSWD — 0
parameter for sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGD — 0
parameter for gate-edge sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGS — 10
parameter for gate-edge sidewall source junction
Unit length trap-assisted current voltage dependent
VTSSWS — 10
parameter for sidewall source junction
WA1 — 0
WA11 — 0
WA2 — 0
WA21 — 0
WAGIDL — 0
WAGISL — 0
WAIGBACC — 0
WAIGBINV — 0

464
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
WAIGC — 0
WAIGD — 0
WAIGS — 0
WALPHA0 — 0
WAT — 0
WBETA0 — 0
WBGIDL — 0
WBGISL — 0
WBIGBACC — 0
WBIGBINV — 0
WBIGC — 0
WBIGD — 0
WBIGS — 0
WCDSCB — 0
WCDSCD — 0
WCDSCDR — 0
WCF F 0
WCGDL — 0
WCGIDL — 0
WCGISL — 0
WCGSL — 0
WCIGBACC — 0
WCIGBINV — 0
WCIGC — 0
WCIGD — 0
WCIGS — 0
WCIT — 0
WCKAPPAD — 0
WCKAPPAS — 0
WDELTA — 0
WDLCIG — 0
WDLCIGD — 0
WDVTP0 — 0
WDVTP1 — 0
WDVTP2 — 0
WDVTP3 — 0

465
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
WDVTP4 — 0
WDVTP5 — 0
WEB — 0
WEC — 0
WEDGE — 1e-08
WEGIDL — 0
WEGISL — 0
WEIGBINV — 0
WETA0 — 0
WETA0R — 0
WETAB — 0
WEU — 0
WFPROUT — 0
WIGT — 0
WIIT — 0
WINT delta W for IV — 0
WK1 — 0
WK2 — 0
WK2WE — 0
WKT1 — 0
WKT2 — 0
WKU0 Width Dependence of KU0 — 0
WKU0WE — 0
WKVTH0 Width dependence of KVTH0 — 0
WKVTH0WE — 0
WL — 0
WLC — 0
WLN — 1
WLOD Width Parameter for Stress Effect — 0
WLODKU0 Width Parameter for u0 stress effect — 0
WLODVTH Width Parameter for Vth stress effect — 0
WMAX Maximum width for which this model should be used m 0
WMIN Minimum width for which this model should be used m 0
WMLT Width Shrinking Parameter — 1
WNDEP — 0
WNDEPCV — 0

466
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
WNFACTOR — 0
WNGATE — 0
WNIGBACC — 0
WNIGBINV — 0
WNSD — 0
WNTOX — 0
WPCLM — 0
WPCLMCV — 0
WPCLMR — 0
WPDIBLC — 0
WPDIBLCB — 0
WPDIBLCR — 0
WPDITS — 0
WPDITSD — 0
WPEMOD Model flag — 0
WPHIN — 0
WPOXEDGE — 0
WPRT — 0
WPRWB — 0
WPRWG — 0
WPSAT — 0
WPSATB — 0
WPSATR — 0
WPSCBE1 — 0
WPSCBE2 — 0
WPTWG — 0
WPTWGR — 0
WPTWGT — 0
WPVAG — 0
WR W dependence parameter of S/D extension resistance — 1
WRDSW — 0
WRDSWMIN — 0
WRDW — 0
WRDWMIN — 0
WRSW — 0
WRSWMIN — 0

467
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
WTGIDL — 0
WTH0 Width dependence coefficient for Rth and Cth — 0
WU0 — 0
WU0R — 0
WUA — 0
WUA1 — 0
WUAR — 0
WUC — 0
WUC1 — 0
WUCR — 0
WUCS — 0
WUCSR — 0
WUCSTE — 0
WUD — 0
WUD1 — 0
WUDR — 0
WUTE — 0
WVFB — 0
WVFBCV — 0
WVSAT — 0
WVSATCV — 0
WVSATR — 0
WW — 0
WWC — 0
WWIDE W of extracted Wide channel device m 1e-05
WWL — 0
WWLC — 0
WWN — 1
WWR — 0
WXJ — 0
XGL Variation in Ldrawn m 0
XGW Dist from gate contact center to dev edge [m] m 0
XJ S/D junction depth — 1.5e-07
XJBVD Fitting parameter for drain diode breakdown current — 0
XJBVS Fitting parameter for source diode breakdown current — 1
XL L offset for channel length due to mask/etch effect — 0

468
 Table 2-114. BSIM6 Device Model Parameters
Parameter Description Units Default
1st fitting parm the bias-dependent Rg //make it
XRCRG1 — 12
binnable
2nd fitting parm the bias-dependent Rg //make it
XRCRG2 — 1
binnable
XTID Drain junction current temperature exponent — 0
XTIS Source junction current temperature exponent — 3
XTSD Power dependence of JTSD on temperature — 0
XTSS Power dependence of JTSS on temperature — 0.02
XTSSWD Power dependence of JTSSWD on temperature — 0
XTSSWGD Power dependence of JTSSWGD on temperature — 0
XTSSWGS Power dependence of JTSSWGS on temperature — 0.02
XTSSWS Power dependence of JTSSWS on temperature — 0.02
XW W offset for channel width due to mask/etch effect — 0

Table 2-115. MOSFET level 77 Output Variables

Parameter Description Units Default
QBI Intrinsic Body Charge C none
QSI Intrinsic Source Charge C none
QDI Intrinsic Drain Charge C none
QGI Intrinsic Gate Charge C none
CGGI Intrinsic g-g MOSFET capacitance F none
CGBI Intrinsic g-b MOSFET capacitance F none
CGSI Intrinsic g-s MOSFET capacitance F none
CGDI Intrinsic g-d MOSFET capacitance F none
CSGI Intrinsic s-g MOSFET capacitance F none
CSBI Intrinsic s-b MOSFET capacitance F none
CSSI Intrinsic s-s MOSFET capacitance F none
CSDI Intrinsic s-d MOSFET capacitance F none
CDGI Intrinsic d-g MOSFET capacitance F none
CDBI Intrinsic d-b MOSFET capacitance F none
CDSI Intrinsic d-s MOSFET capacitance F none
CDDI Intrinsic d-d MOSFET capacitance F none
CBGI Intrinsic b-g MOSFET capacitance F none
CBBI Intrinsic b-b MOSFET capacitance F none
CBSI Intrinsic b-s MOSFET capacitance F none

469
 Table 2-115. MOSFET level 77 Output Variables
Parameter Description Units Default
CBDI Intrinsic b-d MOSFET capacitance F none
QB Body Charge C none
QS Source Charge C none
QD Drain Charge C none
QG Gate Charge C none
CGG g-g MOSFET capacitance F none
CGB g-b MOSFET capacitance F none
CGS g-s MOSFET capacitance F none
CGD g-d MOSFET capacitance F none
CSG s-g MOSFET capacitance F none
CSB s-b MOSFET capacitance F none
CSS s-s MOSFET capacitance F none
CSD s-d MOSFET capacitance F none
CDG d-g MOSFET capacitance F none
CDB d-b MOSFET capacitance F none
CDS d-s MOSFET capacitance F none
CDD d-d MOSFET capacitance F none
CBG b-g MOSFET capacitance F none
CBB b-b MOSFET capacitance F none
CBS b-s MOSFET capacitance F none
CBD b-d MOSFET capacitance F none
ISUB Substrate Current A none
IGIDL A none
IGISL A none
IGS A none
IGD A none
IGCS A none
IGCD A none
IGB A none
CGSEXT F none
CGDEXT F none
CGBOV Front Gate Charge F none
CAPBS F none
CAPBD F none
WEFF m none
LEFF m none

470
 Table 2-115. MOSFET level 77 Output Variables
Parameter Description Units Default
WEFFCV m none
LEFFCV m none
IDS Drain-Source current A none
IDEFF Effective drain Current A none
ISEFF Effective Source Current A none
IGEFF Effective Gate Current A none
IBS A none
IBD A none
VDS Drain to Source Voltage V none
VGS Gate to Source Voltage V none
VBS Body to Source Voltage V none
VDSAT V none
GM mho none
GMBS mho none
GDS mho none
TK m none
VTH Threshold Voltage V none

471
2.3.20.11. Level 102 MOSFET Tables (PSP version 102.5)

Xyce includes a legacy version of the PSP MOSFET model, version 102.5. This version is provided because
the more recent 103 versions are not backward compatible with the older 102 versions, and some foundries
provide model cards that use the version 102. Development of new model cards should be done using the
more recent, supported versions of PSP.
The PSP102 device supports output of the internal variables in table 2-118 on the .PRINT line of a netlist.
To access them from a print line, use the syntax N(<instance>:<variable>) where “<instance>”
refers to the name of the specific PSP102 M device in your netlist.

Table 2-116. PSP102VA legacy MOSFET 102.5 Device Instance Parameters

Parameter Description Units Default
ABDRAIN Bottom area of drain junction – 1e-12
ABSOURCE Bottom area of source junction – 1e-12
AD Bottom area of drain junction – 1e-12
AS Bottom area of source junction – 1e-12
DELVTO Threshold voltage shift parameter – 0
FACTUO Zero-field mobility pre-factor – 1
L Design length – 1e-05
LGDRAIN Gate-edge length of drain junction – 1e-06
LGSOURCE Gate-edge length of source junction – 1e-06
LSDRAIN STI-edge length of drain junction – 1e-06
LSSOURCE STI-edge length of source junction – 1e-06
M Alias for MULT – 1
MULT Number of devices in parallel – 1
NF Number of fingers – 1
NGCON Number of gate contacts – 1
PD Perimeter of drain junction – 1e-06
PS Perimeter of source junction – 1e-06
SA Distance between OD-edge and poly from one side – 0
SB Distance between OD-edge and poly from other side – 0
SC Distance between OD-edge and nearest well edge – 0
Integral of the first distribution function for scattered
SCA – 0
well dopants
Integral of the second distribution function for
SCB – 0
scattered well dopants
Integral of the third distribution function for scattered
SCC – 0
well dopants
SD Distance between neighbouring fingers – 0
W Design width – 1e-05

472
 Table 2-116. PSP102VA legacy MOSFET 102.5 Device Instance Parameters
Parameter Description Units Default
XGW Distance from the gate contact to the channel edge – 1e-07

Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters

Parameter Description Units Default
A1L Length dependence of A1 – 0
A1O Geometry independent impact-ionization pre-factor – 1
A1W Width dependence of A1 – 0
A2O Impact-ionization exponent at TR – 10
A3L Length dependence of A3 – 0
Geometry independent saturation-voltage dependence
A3O – 1
of II
A3W Width dependence of A3 – 0
A4L Length dependence of A4 – 0
A4O Geometry independent back-bias dependence of II – 0
A4W Width dependence of A4 – 0
AGIDLDW Width dependence of GIDL pre-factor for drain side – 0
AGIDLW Width dependence of GIDL pre-factor – 0
Length dependence of CLM enhancement factor
ALP1L1 – 0
above threshold
ALP1L2 Second_order length dependence of ALP1 – 0
ALP1LEXP Exponent for length dependence of ALP1 – 0.5
ALP1W Width dependence of ALP1 – 0
Length dependence of CLM enhancement factor
ALP2L1 – 0
below threshold
ALP2L2 Second_order length dependence of ALP2 – 0
ALP2LEXP Exponent for length dependence of ALP2 – 0.5
ALP2W Width dependence of ALP2 – 0
ALPL Length dependence of ALP – 0.0005
ALPLEXP Exponent for length dependence of ALP – 1
ALPNOI Exponent for length offset for flicker noise – 2
ALPW Width dependence of ALP – 0
AXL Length dependence of AX – 0.4
Geometry independent linear/saturation transition
AXO – 18
factor
BETW1 First higher-order width scaling coefficient of BETN – 0

473
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
Second higher-order width scaling coefficient of
BETW2 – 0
BETN
BGIDLDO GIDL probability factor at TR for drain side – 41
BGIDLO GIDL probability factor at TR – 41
Band-to-band tunneling prefactor of bottom
CBBTBOT – 1e-12
component for source-bulk junction
Band-to-band tunneling prefactor of bottom
CBBTBOTD – 1e-12
component for drain-bulk junction
Band-to-band tunneling prefactor of gate-edge
CBBTGAT – 1e-18
component for source-bulk junction
Band-to-band tunneling prefactor of gate-edge
CBBTGATD – 1e-18
component for drain-bulk junction
Band-to-band tunneling prefactor of STI-edge
CBBTSTI – 1e-18
component for source-bulk junction
Band-to-band tunneling prefactor of STI-edge
CBBTSTID – 1e-18
component for drain-bulk junction
CFBO Back-bias dependence of CF – 0
CFL Length dependence of DIBL-parameter – 0
CFLEXP Exponent for length dependence of CF – 2
Outer fringe capacitance for 1 um wide channel for
CFRDW – 0
drain side
CFRW Outer fringe capacitance for 1 um wide channel – 0
CFW Width dependence of CF – 0
Oxide capacitance for gate-bulk overlap for 1 um long
CGBOVL – 0
channel
CGIDLDO Back-bias dependence of GIDL for drain side – 0
CGIDLO Back-bias dependence of GIDL – 0
CHIBO Tunnelling barrier height – 3.1
Zero-bias capacitance per unit-of-area of bottom
CJORBOT – 0.001
component for source-bulk junction
Zero-bias capacitance per unit-of-area of bottom
CJORBOTD – 0.001
component for drain-bulk junction
Zero-bias capacitance per unit-of-length of gate-edge
CJORGAT – 1e-09
component for source-bulk junction
Zero-bias capacitance per unit-of-length of gate-edge
CJORGATD – 1e-09
component for drain-bulk junction
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTI – 1e-09
component for source-bulk junction
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTID – 1e-09
component for drain-bulk junction

474
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
CSL Length dependence of CS – 0
CSLEXP Exponent for length dependence of CS – 0
CSLW Area dependence of CS – 0
Geometry independent coulomb scattering parameter
CSO – 0
at TR
Shockley-Read-Hall prefactor of bottom component
CSRHBOT – 100
for source-bulk junction
Shockley-Read-Hall prefactor of bottom component
CSRHBOTD – 100
for drain-bulk junction
Shockley-Read-Hall prefactor of gate-edge
CSRHGAT – 0.0001
component for source-bulk junction
Shockley-Read-Hall prefactor of gate-edge
CSRHGATD – 0.0001
component for drain-bulk junction
Shockley-Read-Hall prefactor of STI-edge component
CSRHSTI – 0.0001
for source-bulk junction
Shockley-Read-Hall prefactor of STI-edge component
CSRHSTID – 0.0001
for drain-bulk junction
CSW Width dependence of CS – 0
Trap-assisted tunneling prefactor of bottom
CTATBOT – 100
component for source-bulk junction
Trap-assisted tunneling prefactor of bottom
CTATBOTD – 100
component for drain-bulk junction
Trap-assisted tunneling prefactor of gate-edge
CTATGAT – 0.0001
component for source-bulk junction
Trap-assisted tunneling prefactor of gate-edge
CTATGATD – 0.0001
component for drain-bulk junction
Trap-assisted tunneling prefactor of STI-edge
CTATSTI – 0.0001
component for source-bulk junction
Trap-assisted tunneling prefactor of STI-edge
CTATSTID – 0.0001
component for drain-bulk junction
CTL Length dependence of interface states factor – 0
Exponent for length dependence of interface states
CTLEXP – 1
factor
CTLW Area dependence of interface states factor – 0
CTO Geometry-independent interface states factor – 0
CTW Width dependence of interface states factor – 0
DLQ Effective channel length reduction for CV – 0
DLSIL Silicide extension over the physical gate length – 0
DNSUBO Effective doping bias-dependence parameter – 0
DPHIBL Length dependence offset of PHIB – 0

475
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
DPHIBLEXP Exponent for length dependence of offset of PHIB – 1
DPHIBLW Area dependence of offset of PHIB – 0
DPHIBO Geometry independent offset of PHIB – 0
DPHIBW Width dependence of offset of PHIB – 0
DTA Temperature offset w.r.t. ambient circuit temperature – 0
DWQ Effective channel width reduction for CV – 0
EFO Flicker noise frequency exponent – 1
EPSROXO Relative permittivity of gate dielectric – 3.9
Normalization field at the reference temperature for
FBBTRBOT band-to-band tunneling of bottom component for – 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRBOTD band-to-band tunneling of bottom component for – 1e+09
drain-bulk junction
Normalization field at the reference temperature for
FBBTRGAT band-to-band tunneling of gate-edge component for – 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRGATD band-to-band tunneling of gate-edge component for – 1e+09
drain-bulk junction
Normalization field at the reference temperature for
FBBTRSTI band-to-band tunneling of STI-edge component for – 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRSTID band-to-band tunneling of STI-edge component for – 1e+09
drain-bulk junction
FBET1 Relative mobility decrease due to first lateral profile – 0
Width dependence of relative mobility decrease due
FBET1W – 0
to first lateral profile
Relative mobility decrease due to second lateral
FBET2 – 0
profile
FETAO Effective field parameter – 1
Fraction below which source-bulk junction
FJUNQ – 0.03
capacitance components are considered negligible
Fraction below which drain-bulk junction capacitance
FJUNQD – 0.03
components are considered negligible
FNTEXCL Length dependence coefficient of excess noise – 0
FNTO Thermal noise coefficient – 1
First length dependence coefficient for short channel
FOL1 – 0
body effect

476
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
Second length dependence coefficient for short
FOL2 – 0
channel body effect
GC2O Gate current slope factor – 0.375
GC3O Gate current curvature factor – 0.063
GCOO Gate tunnelling energy adjustment – 0
Saturation current density at the reference temperature
IDSATRBOT – 1e-12
of bottom component for source-bulk junction
Saturation current density at the reference temperature
IDSATRBOTD – 1e-12
of bottom component for drain-bulk junction
Saturation current density at the reference temperature
IDSATRGAT – 1e-18
of gate-edge component for source-bulk junction
Saturation current density at the reference temperature
IDSATRGATD – 1e-18
of gate-edge component for drain-bulk junction
Saturation current density at the reference temperature
IDSATRSTI – 1e-18
of STI-edge component for source-bulk junction
Saturation current density at the reference temperature
IDSATRSTID – 1e-18
of STI-edge component for drain-bulk junction
Gate channel current pre-factor for 1 um**2 channel
IGINVLW – 0
area
Gate overlap current pre-factor for 1 um wide channel
IGOVDW – 0
for drain side
IGOVW Gate overlap current pre-factor for 1 um wide channel – 0
Maximum current up to which forward current
IMAX – 1000
behaves exponentially
KUO Mobility degradation/enhancement coefficient – 0
KUOWEL Length dependent mobility degradation factor – 0
KUOWELW Area dependent mobility degradation factor – 0
KUOWEO Geometrical independent mobility degradation factor – 0
KUOWEW Width dependent mobility degradation factor – 0
Saturation velocity degradation/enhancement
KVSAT – 0
coefficient
KVTHO Threshold shift parameter – 0
KVTHOWEL Length dependent threshold shift parameter – 0
KVTHOWELW Area dependent threshold shift parameter – 0
KVTHOWEO Geometrical independent threshold shift parameter – 0
KVTHOWEW Width dependent threshold shift parameter – 0
LAP Effective channel length reduction per side – 0
LEVEL Model level – 1020
LINTNOI Length offset for flicker noise – 0

477
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
LKUO Length dependence of KUO – 0
LKVTHO Length dependence of KVTHO – 0
LLODKUO Length parameter for UO stress effect – 0
LLODVTH Length parameter for VTH-stress effect – 0
LODETAO eta0 shift modification factor for stress effect – 1
Overlap length for gate/drain and gate/source overlap
LOV – 0
capacitance
LOVD Overlap length for gate/drain overlap capacitance – 0
Mobility-related characteristic length of first lateral
LP1 – 1e-08
profile
Width dependence of mobility-related characteristic
LP1W – 0
length of first lateral profile
Mobility-related characteristic length of second lateral
LP2 – 1e-08
profile
LPCK Char. length of lateral doping profile – 1e-08
Width dependence of char. length of lateral doping
LPCKW – 0
profile
LVARL Length dependence of LVAR – 0
Geom. independent difference between actual and
LVARO – 0
programmed gate length
LVARW Width dependence of LVAR – 0
Effective mass (in units of m0) for trap-assisted
MEFFTATBOT tunneling of bottom component for source-bulk – 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATBOTD tunneling of bottom component for drain-bulk – 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATGAT tunneling of gate-edge component for source-bulk – 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATGATD tunneling of gate-edge component for drain-bulk – 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATSTI tunneling of STI-edge component for source-bulk – 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATSTID tunneling of STI-edge component for drain-bulk – 0.25
junction
Geometry independent mobility reduction coefficient
MUEO – 0.5
at TR

478
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
Width dependence of mobility reduction coefficient at
MUEW – 0
TR
First coefficient of flicker noise for 1 um**2 channel
NFALW – 8e+22
area
Second coefficient of flicker noise for 1 um**2
NFBLW – 3e+07
channel area
Third coefficient of flicker noise for 1 um**2 channel
NFCLW – 0
area
NOVDO Effective doping of overlap region for drain side – 5e+25
NOVO Effective doping of overlap region – 5e+25
NPCK Pocket doping level – 1e+24
Width dependence of pocket doping NPCK due to
NPCKW – 0
segregation
NPL Length dependence of gate poly-silicon doping – 0
NPO Geometry-independent gate poly-silicon doping – 1e+26
NSLPO Effective doping bias-dependence parameter – 0.05
NSUBO Geometry independent substrate doping – 3e+23
Width dependence of background doping NSUBO
NSUBW – 0
due to segregation
Grading coefficient of bottom component for
PBOT – 0.5
source-bulk junction
Grading coefficient of bottom component for
PBOTD – 0.5
drain-bulk junction
Breakdown onset tuning parameter of bottom
PBRBOT – 4
component for source-bulk junction
Breakdown onset tuning parameter of bottom
PBRBOTD – 4
component for drain-bulk junction
Breakdown onset tuning parameter of gate-edge
PBRGAT – 4
component for source-bulk junction
Breakdown onset tuning parameter of gate-edge
PBRGATD – 4
component for drain-bulk junction
Breakdown onset tuning parameter of STI-edge
PBRSTI – 4
component for source-bulk junction
Breakdown onset tuning parameter of STI-edge
PBRSTID – 4
component for drain-bulk junction
Grading coefficient of gate-edge component for
PGAT – 0.5
source-bulk junction
Grading coefficient of gate-edge component for
PGATD – 0.5
drain-bulk junction
Zero-temperature bandgap voltage of bottom
PHIGBOT – 1.16
component for source-bulk junction

479
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
Zero-temperature bandgap voltage of bottom
PHIGBOTD – 1.16
component for drain-bulk junction
Zero-temperature bandgap voltage of gate-edge
PHIGGAT – 1.16
component for source-bulk junction
Zero-temperature bandgap voltage of gate-edge
PHIGGATD – 1.16
component for drain-bulk junction
Zero-temperature bandgap voltage of STI-edge
PHIGSTI – 1.16
component for source-bulk junction
Zero-temperature bandgap voltage of STI-edge
PHIGSTID – 1.16
component for drain-bulk junction
PKUO Cross-term dependence of KUO – 0
PKVTHO Cross-term dependence of KVTHO – 0
Grading coefficient of STI-edge component for
PSTI – 0.5
source-bulk junction
Grading coefficient of STI-edge component for
PSTID – 0.5
drain-bulk junction
QMC Quantum-mechanical correction factor – 1
RBULKO Bulk resistance between node BP and BI – 0
RGO Gate resistance – 0
RINT Contact resistance between silicide and ploy – 0
RJUNDO Drain-side bulk resistance between node BI and BD – 0
RJUNSO Source-side bulk resistance between node BI and BS – 0
RSBO Back-bias dependence of series resistance – 0
RSGO Gate-bias dependence of series resistance – 0
RSHG Gate electrode diffusion sheet resistance – 0
Source/drain series resistance for 1 um wide channel
RSW1 – 2500
at TR
RSW2 Higher-order width scaling of RS – 0
RVPOLY Vertical poly resistance – 0
RWELLO Well resistance between node BI and B – 0
Reference distance between OD-edge and poly from
SAREF – 1e-06
one side
Reference distance between OD-edge and poly from
SBREF – 1e-06
other side
Distance between OD-edge and well edge of a
SCREF – 1e-06
reference device
STA2O Temperature dependence of A2 – 0
Length dependence of temperature dependence of
STBETL – 0
BETN

480
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
Area dependence of temperature dependence of
STBETLW – 0
BETN
Geometry independent temperature dependence of
STBETO – 1
BETN
Width dependence of temperature dependence of
STBETW – 0
BETN
STBGIDLDO Temperature dependence of BGIDL for drain side – 0
STBGIDLO Temperature dependence of BGIDL – 0
STCSO Temperature dependence of CS – 0
STETAO eta0 shift factor related to VTHO change – 0
Temperature scaling parameter for band-to-band
STFBBTBOT tunneling of bottom component for source-bulk – -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTBOTD tunneling of bottom component for drain-bulk – -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTGAT tunneling of gate-edge component for source-bulk – -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTGATD tunneling of gate-edge component for drain-bulk – -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTSTI tunneling of STI-edge component for source-bulk – -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTSTID tunneling of STI-edge component for drain-bulk – -0.001
junction
STIGO Temperature dependence of IGINV and IGOV – 2
STMUEO Temperature dependence of MUE – 0
STRSO Temperature dependence of RS – 1
STTHEMUO Temperature dependence of THEMU – 1.5
Length dependence of temperature dependence of
STTHESATL – 0
THESAT
Area dependence of temperature dependence of
STTHESATLW – 0
THESAT
Geometry independent temperature dependence of
STTHESATO – 1
THESAT
Width dependence of temperature dependence of
STTHESATW – 0
THESAT

481
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
Length dependence of temperature dependence of
STVFBL – 0
VFB
STVFBLW Area dependence of temperature dependence of VFB – 0
Geometry-independent temperature dependence of
STVFBO – 0.0005
VFB
STVFBW Width dependence of temperature dependence of VFB – 0
STXCORO Temperature dependence of XCOR – 0
SWGIDL Flag for GIDL current, 0=turn off IGIDL – 0
SWIGATE Flag for gate current, 0=turn off IG – 0
SWIMPACT Flag for impact ionization current, 0=turn off II – 0
Flag for asymmetric junctions; 0=symmetric,
SWJUNASYM – 0
1=asymmetric
SWJUNCAP Flag for juncap, 0=turn off juncap – 0
Flag for JUNCAP-express; 0=full model, 1=express
SWJUNEXP – 0
model
THEMUO Mobility reduction exponent at TR – 1.5
THESATBO Back-bias dependence of velocity saturation – 0
THESATGO Gate-bias dependence of velocity saturation – 0
THESATL Length dependence of THESAT – 0.05
THESATLEXP Exponent for length dependence of THESAT – 1
THESATLW Area dependence of velocity saturation parameter – 0
Geometry independent velocity saturation parameter
THESATO – 0
at TR
THESATW Width dependence of velocity saturation parameter – 0
TKUO Temperature dependence of KUO – 0
TOXO Gate oxide thickness – 2e-09
TOXOVDO Overlap oxide thickness for drain side – 2e-09
TOXOVO Overlap oxide thickness – 2e-09
TR nominal (reference) temperature – 21
TRJ reference temperature – 21
TYPE Channel type parameter, +1=NMOS -1=PMOS – 1
UO Zero-field mobility at TR – 0.05
Built-in voltage at the reference temperature of
VBIRBOT – 1
bottom component for source-bulk junction
Built-in voltage at the reference temperature of
VBIRBOTD – 1
bottom component for drain-bulk junction
Built-in voltage at the reference temperature of
VBIRGAT – 1
gate-edge component for source-bulk junction

482
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
Built-in voltage at the reference temperature of
VBIRGATD – 1
gate-edge component for drain-bulk junction
Built-in voltage at the reference temperature of
VBIRSTI – 1
STI-edge component for source-bulk junction
Built-in voltage at the reference temperature of
VBIRSTID – 1
STI-edge component for drain-bulk junction
Breakdown voltage of bottom component for
VBRBOT – 10
source-bulk junction
Breakdown voltage of bottom component for
VBRBOTD – 10
drain-bulk junction
Breakdown voltage of gate-edge component for
VBRGAT – 10
source-bulk junction
Breakdown voltage of gate-edge component for
VBRGATD – 10
drain-bulk junction
Breakdown voltage of STI-edge component for
VBRSTI – 10
source-bulk junction
Breakdown voltage of STI-edge component for
VBRSTID – 10
drain-bulk junction
VFBL Length dependence of flat-band voltage – 0
VFBLW Area dependence of flat-band voltage – 0
VFBO Geometry-independent flat-band voltage at TR – -1
VFBW Width dependence of flat-band voltage – 0
Typical maximum source-bulk junction voltage;
VJUNREF – 2.5
usually about 2*VSUP
Typical maximum drain-bulk junction voltage; usually
VJUNREFD – 2.5
about 2*VSUP
VNSUBO Effective doping bias-dependence parameter – 0
VPO CLM logarithmic dependence parameter – 0.05
WBET Characteristic width for width scaling of BETN – 1e-09
WEB Coefficient for SCB – 0
WEC Coefficient for SCC – 0
WKUO Width dependence of KUO – 0
WKVTHO Width dependence of KVTHO – 0
WLOD Width parameter – 0
WLODKUO Width parameter for UO stress effect – 0
WLODVTH Width parameter for VTH-stress effect – 0
WOT Effective channel width reduction per side – 0
Char. length of segregation of background doping
WSEG – 1e-08
NSUBO

483
 Table 2-117. PSP102VA legacy MOSFET 102.5 Device Model Parameters
Parameter Description Units Default
WSEGP Char. length of segregation of pocket doping NPCK – 1e-08
WVARL Length dependence of WVAR – 0
Geom. independent difference between actual and
WVARO – 0
programmed field-oxide opening
WVARW Width dependence of WVAR – 0
XCORL Length dependence of non-universality parameter – 0
XCORLW Area dependence of non-universality parameter – 0
XCORO Geometry independent non-universality parameter – 0
XCORW Width dependence of non-universality parameter – 0
Junction depth of gate-edge component for
XJUNGAT – 1e-07
source-bulk junction
Junction depth of gate-edge component for drain-bulk
XJUNGATD – 1e-07
junction
Junction depth of STI-edge component for
XJUNSTI – 1e-07
source-bulk junction
Junction depth of STI-edge component for drain-bulk
XJUNSTID – 1e-07
junction

Table 2-118. MOSFET level 102 Output Variables

Parameter Description Units Default
ctype Flag for channel type none
sdint Flag for source-drain interchange none
ise Total source current A none
ige Total gate current A none
ide Total drain current A none
ibe Total bulk current A none
Drain current, excl. avalanche, tunnel, GISL, GIDL,
ids A none
and junction currents
idb Drain to bulk current A none
isb Source to bulk current A none
igs Gate-source tunneling current A none
igd Gate-drain tunneling current A none
igb Gate-bulk tunneling current A none
igcs Gate-channel tunneling current (source component) A none
igcd Gate-channel tunneling current (drain component) A none
iavl Substrate current due to weak avelanche A none
igisl Gate-induced source leakage current A none

484
 Table 2-118. MOSFET level 102 Output Variables
Parameter Description Units Default
igidl Gate-induced drain leakage current A none
ijs Total source junction current A none
ijsbot Source junction current (bottom component) A none
ijsgat Source junction current (gate-edge component) A none
ijssti Source junction current (STI-edge component) A none
ijd Total drain junction current A none
ijdbot Drain junction current (bottom component) A none
ijdgat Drain junction current (gate-edge component) A none
ijdsti Drain junction current (STI-edge component) A none
vds Drain-source voltage V none
vgs Gate-source voltage V none
vsb Source-bulk voltage V none
vto Zero-bias threshold voltage V none
vts Threshold voltage including back bias effects V none
Threshold voltage including back bias and drain bias
vth V none
effects
Effective gate drive voltage including back bias and
vgt V none
drain bias effects
vdss Drain saturation voltage at actual bias V none
vsat Saturation limit V none
gm Transconductance ˙ −1 none
gmb Substrate transconductance ˙ −1 none
gds Output conductance ˙ −1 none
gjs Source junction conductance ˙ −1 none
gjd Drain junction conductance ˙ −1 none
cdd Drain capacitance F none
cdg Drain-gate capacitance F none
cds Drain-source capacitance F none
cdb Drain-bulk capacitance F none
cgd Gate-drain capacitance F none
cgg Gate capacitance F none
cgs Gate-source capacitance F none
cgb Gate-bulk capacitance F none
csd Source-drain capacitance F none
csg Source-gate capacitance F none
css Source capacitance F none

485
 Table 2-118. MOSFET level 102 Output Variables
Parameter Description Units Default
csb Source-bulk capacitance F none
cbd Bulk-drain capacitance F none
cbg Bulk-gate capacitance F none
cbs Bulk-source capacitance F none
cbb Bulk capacitance F none
cgsol Total gate-source overlap capacitance F none
cgdol Total gate-drain overlap capacitance F none
cjs Total source junction capacitance F none
cjsbot Source junction capacitance (bottom component) F none
cjsgat Source junction capacitance (gate-edge component) F none
cjssti Source junction capacitance (STI-edge component) F none
cjd Total drain junction capacitance F none
cjdbot Drain junction capacitance (bottom component) F none
cjdgat Drain junction capacitance (gate-edge component) F none
cjdsti Drain junction capacitance (STI-edge component) F none
weff Effective channel width for geometrical models m none
leff Effective channel length for geometrical models m none
u Transistor gain none
rout Small-signal output resistance Ohm none
vearly Equivalent Early voltage V none
beff Gain factor A/V2 none
fug Unity gain frequency at actual bias Hz none
rg Gate resistance Ohm none
sfl Flicker noise current spectral density at 1 Hz A2 /Hz none
Input-referred RMS white noise voltage spectral
sqrtsff V/sqrt(Hz) none
density at 1 kHz
Input-referred RMS white noise voltage spectral
sqrtsfw V/sqrt(Hz) none
density
sid White noise current spectral density A2 /Hz none
sig Induced gate noise current spectral density at 1 Hz A2 /Hz none
Imaginary part of correlation coefficient between Sig
cigid none
and Sid
Cross-over frequency above which white noise is
fknee Hz none
dominant
sigs Gate-source current noise spectral density A2 /Hz none
sigd Gate-drain current noise spectral density A2 /Hz none
siavl Impact ionization current noise spectral density A2 /Hz none

486
 Table 2-118. MOSFET level 102 Output Variables
Parameter Description Units Default
ssi Total source junction current noise spectral density A2 /Hz none
sdi Total drain junction current noise spectral density A2 /Hz none
lp_vfb Local parameter VFB after T-scaling and clipping V none
lp_stvfb Local parameter STVFB after clipping V/K none
lp_tox Local parameter TOX after clipping m none
lp_epsrox Local parameter EPSROX after clipping none
lp_neff Local parameter NEFF after clipping m−3 none
lp_vnsub Local parameter VNSUB after clipping V none
lp_nslp Local parameter NSLP after clipping V none
lp_dnsub Local parameter DNSUB after clipping V−1 none
lp_dphib Local parameter DPHIB after clipping V none
lp_np Local parameter NP after clipping m−3 none
lp_ct Local parameter CT after clipping none
lp_toxov Local parameter TOXOV after clipping m none
lp_toxovd Local parameter TOXOVD after clipping m none
lp_nov Local parameter NOV after clipping m−3 none
lp_novd Local parameter NOVD after clipping m−3 none
lp_cf Local parameter CF after clipping none
lp_cfb Local parameter CFB after clipping V−1 none
lp_betn Local parameter BETN after T-scaling and clipping m2 /(V s) none
lp_stbet Local parameter STBET after clipping none
lp_mue Local parameter MUE after T-scaling and clipping m/V none
lp_stmue Local parameter STMUE after clipping none
lp_themu Local parameter THEMU after T-scaling and clipping none
lp_stthemu Local parameter STTHEMU after clipping none
lp_cs Local parameter CS after T-scaling and clipping none
lp_stcs Local parameter STCS after clipping none
lp_xcor Local parameter XCOR after T-scaling and clipping V−1 none
lp_stxcor Local parameter STXCOR after clipping none
lp_feta Local parameter FETA after clipping none
lp_rs Local parameter RS after T-scaling and clipping Ohm none
lp_strs Local parameter STRS after clipping none
lp_rsb Local parameter RSB after clipping V−1 none
lp_rsg Local parameter RSG after clipping V−1 none
lp_thesat Local parameter THESAT after T-scaling and clipping V−1 none
lp_stthesat Local parameter STTHESAT after clipping none

487
 Table 2-118. MOSFET level 102 Output Variables
Parameter Description Units Default
lp_thesatb Local parameter THESATB after clipping V−1 none
lp_thesatg Local parameter THESATG after clipping V−1 none
lp_ax Local parameter AX after clipping none
lp_alp Local parameter ALP after clipping none
lp_alp1 Local parameter ALP1 after clipping V none
lp_alp2 Local parameter ALP2 after clipping V−1 none
lp_vp Local parameter VP after clipping V none
lp_a1 Local parameter A1 after clipping none
lp_a2 Local parameter A2 after T-scaling and clipping V none
lp_sta2 Local parameter STA2 after clipping none
lp_a3 Local parameter A3 after clipping none
lp_a4 Local parameter A4 after clipping 1/sqrt(V) none
lp_gco Local parameter GCO after clipping none
lp_iginv Local parameter IGINV after T-scaling and clipping A none
lp_igov Local parameter IGOV after T-scaling and clipping A none
lp_igovd Local parameter IGOVD after T-scaling and clipping A none
lp_stig Local parameter STIG after clipping none
lp_gc2 Local parameter GC2 after clipping none
lp_gc3 Local parameter GC3 after clipping none
lp_chib Local parameter CHIB after clipping V none
lp_agidl Local parameter AGIDL after clipping A/V3 none
lp_agidld Local parameter AGIDLD after clipping A/V3 none
lp_bgidl Local parameter BGIDL after T-scaling and clipping V none
lp_bgidld Local parameter BGIDLD after T-scaling and clipping V none
lp_stbgidl Local parameter STBGIDL after clipping V/K none
lp_stbgidld Local parameter STBGIDLD after clipping V/K none
lp_cgidl Local parameter CGIDL after clipping none
lp_cgidld Local parameter CGIDLD after clipping none
lp_cox Local parameter COX after clipping F none
lp_cgov Local parameter CGOV after clipping F none
lp_cgovd Local parameter CGOVD after clipping F none
lp_cgbov Local parameter CGBOV after clipping F none
lp_cfr Local parameter CFR after clipping F none
lp_cfrd Local parameter CFRD after clipping F none
lp_fnt Local parameter FNT after clipping none
lp_fntexc Local parameter FNTEXC after clipping none

488
 Table 2-118. MOSFET level 102 Output Variables
Parameter Description Units Default
lp_nfa Local parameter NFA after clipping 1/(V m4 ) none
lp_nfb Local parameter NFB after clipping 1/(V m2 ) none
lp_nfc Local parameter NFC after clipping V−1 none
lp_ef Local parameter EF after clipping none
lp_rg Local parameter RG after clipping Ohm none
lp_rbulk Local parameter RBULK after clipping Ohm none
lp_rwell Local parameter RWELL after clipping Ohm none
lp_rjuns Local parameter RJUNS after clipping Ohm none
lp_rjund Local parameter RJUND after clipping Ohm none
tk Device Temperature K none
Bottom component of total zero-bias source junction
cjosbot F none
capacitance at device temperature
STI-edge component of total zero-bias source junction
cjossti F none
capacitance at device temperature
Gate-edge component of total zero-bias source
cjosgat F none
junction capacitance at device temperature
Built-in voltage of source-side bottom junction at
vbisbot V none
device temperature
Built-in voltage of source-side STI-edge junction at
vbissti V none
device temperature
Built-in voltage of source-side gate-edge junction at
vbisgat V none
device temperature
idsatsbot Total source-side bottom junction saturation current A none
idsatssti Total source-side STI-edge junction saturation current A none
idsatsgat Total source-side gate-edge junction saturation current A none
Bottom component of total zero-bias drain junction
cjosbotd F none
capacitance at device temperature
STI-edge component of total zero-bias drain junction
cjosstid F none
capacitance at device temperature
Gate-edge component of total zero-bias drain junction
cjosgatd F none
capacitance at device temperature
Built-in voltage of drain-side bottom junction at
vbisbotd V none
device temperature
Built-in voltage of drain-side STI-edge junction at
vbisstid V none
device temperature
Built-in voltage of drain-side gate-edge junction at
vbisgatd V none
device temperature
idsatsbotd Total drain-side bottom junction saturation current A none
idsatsstid Total drain-side STI-edge junction saturation current A none

489
 Table 2-118. MOSFET level 102 Output Variables
Parameter Description Units Default
idsatsgatd Total drain-side gate-edge junction saturation current A none

2.3.20.12. Level 103 and 1031 MOSFET Tables (PSP version 103.4)

Xyce includes the PSP MOSFET model, version 103.4 [29]. The version without self-heating is the level
103 MOSFET, and the version with self-heating is the level 1031. Note that the level 1031 MOSFET
requires five nodes on its instance line: drain, gate, source, bulk, and dt. The fifth node will be the
temperature rise of the device due to self-heating.
Full documentation for the PSP model is available on its web site,
http://www.cea.fr/cea-tech/leti/pspsupport. Instance and model parameters for the PSP model
are given in tables 2-119, 2-120, 2-122, and 2-123.
The PSP103 devices support output of the internal variables in table 2-121 and table 2-124 on the .PRINT
line of a netlist. To access them from a print line, use the syntax N(<instance>:<variable>) where
“<instance>” refers to the name of the specific PSP103 M device in your netlist.

Table 2-119. PSP103VA MOSFET Device Instance Parameters

Parameter Description Units Default
ABDRAIN Bottom area of drain junction m2 1e-12
ABSOURCE Bottom area of source junction m2 1e-12
AD Bottom area of drain junction m2 1e-12
AS Bottom area of source junction m2 1e-12
DELVTO Threshold voltage shift parameter V 0
DELVTOEDGE Threshold voltage shift parameter of edge transistor V 0
DTA Temperature offset w.r.t. ambient temperature K 0
FACTUO Zero-field mobility pre-factor — 1
FACTUOEDGE Zero-field mobility pre-factor of edge transistor — 1
JW Gate-edge length of source/drain junction m 1e-06
L Design length m 1e-05
LGDRAIN Gate-edge length of drain junction m 1e-06
LGSOURCE Gate-edge length of source junction m 1e-06
LSDRAIN STI-edge length of drain junction m 1e-06
LSSOURCE STI-edge length of source junction m 1e-06
M Alias for MULT — 1
MULT Number of devices in parallel — 1
NF Number of fingers — 1
NGCON Number of gate contacts — 1
NRD Number of squares of drain diffusion — 0

490
 Table 2-119. PSP103VA MOSFET Device Instance Parameters
Parameter Description Units Default
NRS Number of squares of source diffusion — 0
PD Perimeter of drain junction m 1e-06
PS Perimeter of source junction m 1e-06
SA Distance between OD-edge and poly from one side m 0
SB Distance between OD-edge and poly from other side m 0
SC Distance between OD-edge and nearest well edge m 0
Integral of the first distribution function for scattered
SCA — 0
well dopants
Integral of the second distribution function for
SCB — 0
scattered well dopants
Integral of the third distribution function for scattered
SCC — 0
well dopants
SD Distance between neighbouring fingers m 0
W Design width m 1e-05
XGW Distance from the gate contact to the channel edge m 1e-07

Table 2-120. PSP103VA MOSFET Device Model Parameters

Parameter Description Units Default
A1 Impact-ionization pre-factor — 1
A1L Length dependence of A1 — 0
A1O Geometry independent impact-ionization pre-factor — 1
A1W Width dependence of A1 — 0
A2 Impact-ionization exponent at TR V 10
A2O Impact-ionization exponent at TR V 10
A3 Saturation-voltage dependence of impact-ionization — 1
A3L Length dependence of A3 — 0
Geometry independent saturation-voltage dependence
A3O — 1
of II
A3W Width dependence of A3 — 0
A4 Back-bias dependence of impact-ionization V−1/2 0
A4L Length dependence of A4 — 0
A4O Geometry independent back-bias dependence of II V−1/2 0
A4W Width dependence of A4 — 0
AGIDL GIDL pre-factor A/V3 0
AGIDLD GIDL pre-factor for drain side A/V3 0
AGIDLDW Width dependence of GIDL pre-factor for drain side A/V3 0

491
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
AGIDLW Width dependence of GIDL pre-factor A/V3 0
ALP CLM pre-factor — 0.01
ALP1 CLM enhancement factor above threshold V 0
Length dependence of CLM enhancement factor
ALP1L1 V 0
above threshold
ALP1L2 Second_order length dependence of ALP1 — 0
ALP1LEXP Exponent for length dependence of ALP1 — 0.5
ALP1W Width dependence of ALP1 — 0
ALP2 CLM enhancement factor below threshold V−1 0
Length dependence of CLM enhancement factor
ALP2L1 V−1 0
below threshold
ALP2L2 Second_order length dependence of ALP2 — 0
ALP2LEXP Exponent for length dependence of ALP2 — 0.5
ALP2W Width dependence of ALP2 — 0
ALPL Length dependence of ALP — 0.0005
ALPLEXP Exponent for length dependence of ALP — 1
ALPNOI Exponent for length offset for flicker noise — 2
ALPW Width dependence of ALP — 0
AX Linear/saturation transition factor — 3
AXL Length dependence of AX — 0.4
Geometry independent linear/saturation transition
AXO — 18
factor
BETEDGEW Width scaling coefficient of edge transistor mobility — 0
BETN Channel aspect ratio times zero-field mobility m2 /(Vs) 0.07
Channel aspect ratio times zero-field mobility of edge
BETNEDGE m2 /(Vs) 0.0005
transistor
BETW1 First higher-order width scaling coefficient of BETN — 0
Second higher-order width scaling coefficient of
BETW2 — 0
BETN
BGIDL GIDL probability factor at TR V 41
BGIDLD GIDL probability factor at TR for drain side V 41
BGIDLDO GIDL probability factor at TR for drain side V 41
BGIDLO GIDL probability factor at TR V 41
Band-to-band tunneling prefactor of bottom
CBBTBOT A/V3 1e-12
component for source-bulk junction
Band-to-band tunneling prefactor of bottom
CBBTBOTD A/V3 1e-12
component for drain-bulk junction

492
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Band-to-band tunneling prefactor of gate-edge
CBBTGAT Am/V3 1e-18
component for source-bulk junction
Band-to-band tunneling prefactor of gate-edge
CBBTGATD Am/V3 1e-18
component for drain-bulk junction
Band-to-band tunneling prefactor of STI-edge
CBBTSTI Am/V3 1e-18
component for source-bulk junction
Band-to-band tunneling prefactor of STI-edge
CBBTSTID Am/V3 1e-18
component for drain-bulk junction
CF DIBL-parameter — 0
CFB Back bias dependence of CF V−1 0
Bulk voltage dependence parameter of
CFBEDGE V−1 0
DIBL-parameter of edge transistors
Bulk voltage dependence parameter of
CFBEDGEO V−1 0
DIBL-parameter of edge transistors
CFBO Back-bias dependence of CF V−1 0
CFD Drain voltage dependence of CF V−1 0
Drain voltage dependence parameter of
CFDEDGE V−1 0
DIBL-parameter of edge transistors
Drain voltage dependence parameter of
CFDEDGEO V−1 0
DIBL-parameter of edge transistors
CFDO Drain voltage dependence of CF V−1 0
CFEDGE DIBL parameter of edge transistors — 0
Length dependence of DIBL-parameter of edge
CFEDGEL — 0
transistors
Exponent for length dependence of DIBL-parameter
CFEDGELEXP — 2
of edge transistors
Width dependence of DIBL-parameter of edge
CFEDGEW — 0
transistors
CFL Length dependence of DIBL-parameter — 0
CFLEXP Exponent for length dependence of CF — 2
CFR Outer fringe capacitance F 0
CFRD Outer fringe capacitance for drain side F 0
Outer fringe capacitance for 1 um wide channel for
CFRDW F 0
drain side
CFRW Outer fringe capacitance for 1 um wide channel F 0
CFW Width dependence of CF — 0
CGBOV Oxide capacitance for gate-bulk overlap F 0
Oxide capacitance for gate-bulk overlap for 1 um long
CGBOVL F 0
channel

493
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
CGIDL Back-bias dependence of GIDL — 0
CGIDLD Back-bias dependence of GIDL for drain side — 0
CGIDLDO Back-bias dependence of GIDL for drain side — 0
CGIDLO Back-bias dependence of GIDL — 0
CGOV Oxide capacitance for gate-drain/source overlap F 1e-15
CGOVD Oxide capacitance for gate-drain overlap F 1e-15
CHIB Tunnelling barrier height V 3.1
CHIBO Tunnelling barrier height V 3.1
Zero-bias capacitance per unit-of-area of bottom
CJORBOT F/m2 0.001
component for source-bulk junction
Zero-bias capacitance per unit-of-area of bottom
CJORBOTD F/m2 0.001
component for drain-bulk junction
Zero-bias capacitance per unit-of-length of gate-edge
CJORGAT F/m 1e-09
component for source-bulk junction
Zero-bias capacitance per unit-of-length of gate-edge
CJORGATD F/m 1e-09
component for drain-bulk junction
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTI F/m 1e-09
component for source-bulk junction
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTID F/m 1e-09
component for drain-bulk junction
COX Oxide capacitance for intrinsic channel F 1e-14
CS Coulomb scattering parameter at TR — 0
CSL Length dependence of CS — 0
CSLEXP Exponent for length dependence of CS — 1
CSLW Area dependence of CS — 0
Geometry independent coulomb scattering parameter
CSO — 0
at TR
Shockley-Read-Hall prefactor of bottom component
CSRHBOT A/m3 100
for source-bulk junction
Shockley-Read-Hall prefactor of bottom component
CSRHBOTD A/m3 100
for drain-bulk junction
Shockley-Read-Hall prefactor of gate-edge
CSRHGAT A/m2 0.0001
component for source-bulk junction
Shockley-Read-Hall prefactor of gate-edge
CSRHGATD A/m2 0.0001
component for drain-bulk junction
Shockley-Read-Hall prefactor of STI-edge component
CSRHSTI A/m2 0.0001
for source-bulk junction
Shockley-Read-Hall prefactor of STI-edge component
CSRHSTID A/m2 0.0001
for drain-bulk junction

494
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
CSW Width dependence of CS — 0
CT Interface states factor — 0
Trap-assisted tunneling prefactor of bottom
CTATBOT A/m3 100
component for source-bulk junction
Trap-assisted tunneling prefactor of bottom
CTATBOTD A/m3 100
component for drain-bulk junction
Trap-assisted tunneling prefactor of gate-edge
CTATGAT A/m2 0.0001
component for source-bulk junction
Trap-assisted tunneling prefactor of gate-edge
CTATGATD A/m2 0.0001
component for drain-bulk junction
Trap-assisted tunneling prefactor of STI-edge
CTATSTI A/m2 0.0001
component for source-bulk junction
Trap-assisted tunneling prefactor of STI-edge
CTATSTID A/m2 0.0001
component for drain-bulk junction
CTEDGE Interface states factor of edge transistors — 0
Length dependence of interface states factor of edge
CTEDGEL — 0
transistors
Exponent for length dependence of interface states
CTEDGELEXP — 1
factor of edge transistors
Geometry-independent interface states factor of edge
CTEDGEO — 0
transistors
CTL Length dependence of interface states factor — 0
Exponent for length dependence of interface states
CTLEXP — 1
factor
CTLW Area dependence of interface states factor — 0
CTO Geometry-independent interface states factor — 0
CTW Width dependence of interface states factor — 0
Offset parameter for PHIB in separate charge
DELVTAC V 0
calculation
DELVTACL Length dependence of DELVTAC V 0
Exponent for length dependence of offset of
DELVTACLEXP — 1
DELVTAC
DELVTACLW Area dependence of DELVTAC V 0
Geom. independent offset parameter for PHIB in
DELVTACO V 0
separate charge calculation
DELVTACW Width dependence of DELVTAC V 0
DLQ Effective channel length reduction for CV m 0
DLSIL Silicide extension over the physical gate length m 0
DNSUB Effective doping bias-dependence parameter V−1 0

495
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
DNSUBO Effective doping bias-dependence parameter V−1 0
DPHIB Offset parameter for PHIB V 0
DPHIBEDGE Offset parameter for PHIB of edge transistors V 0
DPHIBEDGEL Length dependence of edge transistor PHIB offset V 0
Exponent for length dependence of edge transistor
DPHIBEDGELEXP — 1
PHIB offset
DPHIBEDGELW Area dependence of edge transistor PHIB offset V 0
DPHIBEDGEO Geometry independent of edge transistor PHIB offset V 0
DPHIBEDGEW Width dependence of edge transistor PHIB offset V 0
DPHIBL Length dependence offset of PHIB V 0
DPHIBLEXP Exponent for length dependence of offset of PHIB — 1
DPHIBLW Area dependence of offset of PHIB V 0
DPHIBO Geometry independent offset of PHIB V 0
DPHIBW Width dependence of offset of PHIB V 0
DTA Temperature offset w.r.t. ambient temperature K 0
DVSBNUD Vsb-range for NUD-effect V 1
DVSBNUDO Vsb range for NUD-effect V 1
DWQ Effective channel width reduction for CV m 0
EF Flicker noise frequency exponent — 1
EFEDGE Flicker noise frequency exponent of edge transistors — 1
EFEDGEO Flicker noise frequency exponent — 1
EFO Flicker noise frequency exponent — 1
EPSROX Relative permittivity of gate dielectric — 3.9
EPSROXO Relative permittivity of gate dielectric — 3.9
Pre-factor for effective substrate doping in separate
FACNEFFAC — 1
charge calculation
FACNEFFACL Length dependence of FACNEFFAC — 0
FACNEFFACLW Area dependence of FACNEFFAC — 0
Geom. independent pre-factor for effective substrate
FACNEFFACO — 1
doping in separate charge calculation
FACNEFFACW Width dependence of FACNEFFAC — 0
Normalization field at the reference temperature for
FBBTRBOT band-to-band tunneling of bottom component for Vm−1 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRBOTD band-to-band tunneling of bottom component for Vm−1 1e+09
drain-bulk junction

496
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Normalization field at the reference temperature for
FBBTRGAT band-to-band tunneling of gate-edge component for Vm−1 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRGATD band-to-band tunneling of gate-edge component for Vm−1 1e+09
drain-bulk junction
Normalization field at the reference temperature for
FBBTRSTI band-to-band tunneling of STI-edge component for Vm−1 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRSTID band-to-band tunneling of STI-edge component for Vm−1 1e+09
drain-bulk junction
FBET1 Relative mobility decrease due to first lateral profile — 0
Width dependence of relative mobility decrease due
FBET1W — 0
to first lateral profile
Relative mobility decrease due to second lateral
FBET2 — 0
profile
FBETEDGE Length dependence of edge transistor mobility — 0
FETA Effective field parameter — 1
FETAO Effective field parameter — 1
Fraction below which source-bulk junction
FJUNQ — 0.03
capacitance components are considered negligible
Fraction below which drain-bulk junction capacitance
FJUNQD — 0.03
components are considered negligible
FNT Thermal noise coefficient — 1
FNTEDGE Thermal noise coefficient of edge transistors — 1
FNTEDGEO Thermal noise coefficient — 1
FNTEXC Excess noise coefficient — 0
FNTEXCL Length dependence coefficient of excess noise — 0
FNTO Thermal noise coefficient — 1
First length dependence coefficient for short channel
FOL1 — 0
body effect
Second length dependence coefficient for short
FOL2 — 0
channel body effect
FREV Coefficient for reverse breakdown current limitation — 1000
GC2 Gate current slope factor — 0.375
GC2O Gate current slope factor — 0.375
GC3 Gate current curvature factor — 0.063
GC3O Gate current curvature factor — 0.063
GCO Gate tunnelling energy adjustment — 0

497
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
GCOO Gate tunnelling energy adjustment — 0
GFACNUD Body-factor change due to NUD-effect — 1
GFACNUDL Length dependence of GFACNUD — 0
GFACNUDLEXP Exponent for length dependence of GFACNUD — 1
GFACNUDLW Area dependence of GFACNUD — 0
Geom. independent body-factor change due to
GFACNUDO — 1
NUD-effect
GFACNUDW Width dependence of GFACNUD — 0
Saturation current density at the reference temperature
IDSATRBOT A/m2 1e-12
of bottom component for source-bulk junction
Saturation current density at the reference temperature
IDSATRBOTD A/m2 1e-12
of bottom component for drain-bulk junction
Saturation current density at the reference temperature
IDSATRGAT A/m 1e-18
of gate-edge component for source-bulk junction
Saturation current density at the reference temperature
IDSATRGATD A/m 1e-18
of gate-edge component for drain-bulk junction
Saturation current density at the reference temperature
IDSATRSTI A/m 1e-18
of STI-edge component for source-bulk junction
Saturation current density at the reference temperature
IDSATRSTID A/m 1e-18
of STI-edge component for drain-bulk junction
IGINV Gate channel current pre-factor A 0
Gate channel current pre-factor for 1 um**2 channel
IGINVLW A 0
area
IGOV Gate overlap current pre-factor A 0
IGOVD Gate overlap current pre-factor for drain side A 0
Gate overlap current pre-factor for 1 um wide channel
IGOVDW A 0
for drain side
IGOVW Gate overlap current pre-factor for 1 um wide channel A 0
Maximum current up to which forward current
IMAX A 1000
behaves exponentially
KUO Mobility degradation/enhancement coefficient m 0
KUOWEL Length dependent mobility degradation factor — 0
KUOWELW Area dependent mobility degradation factor — 0
KUOWEO Geometrical independent mobility degradation factor — 0
KUOWEW Width dependent mobility degradation factor — 0
Saturation velocity degradation/enhancement
KVSAT m 0
coefficient
KVTHO Threshold shift parameter Vm 0
KVTHOWEL Length dependent threshold shift parameter — 0

498
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
KVTHOWELW Area dependent threshold shift parameter — 0
KVTHOWEO Geometrical independent threshold shift parameter — 0
KVTHOWEW Width dependent threshold shift parameter — 0
LAP Effective channel length reduction per side m 0
LEVEL Model level — 103
LINTNOI Length offset for flicker noise m 0
LKUO Length dependence of KUO — 0
LKVTHO Length dependence of KVTHO — 0
LLODKUO Length parameter for UO stress effect — 0
LLODVTH Length parameter for VTH-stress effect — 0
LMAX Dummy parameter to label binning set m 1
LMIN Dummy parameter to label binning set m 0
LODETAO Eta0 shift modification factor for stress effect — 1
Overlap length for gate/drain and gate/source overlap
LOV m 0
capacitance
LOVD Overlap length for gate/drain overlap capacitance m 0
Mobility-related characteristic length of first lateral
LP1 m 1e-08
profile
Width dependence of mobility-related characteristic
LP1W — 0
length of first lateral profile
Mobility-related characteristic length of second lateral
LP2 m 1e-08
profile
LPCK Char. length of lateral doping profile m 1e-08
Width dependence of char. length of lateral doping
LPCKW — 0
profile
Exponent for length dependence of edge transistor
LPEDGE m 1e-08
mobility
LVARL Length dependence of LVAR — 0
Geom. independent difference between actual and
LVARO m 0
programmed gate length
LVARW Width dependence of LVAR — 0
Effective mass (in units of m0) for trap-assisted
MEFFTATBOT tunneling of bottom component for source-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATBOTD tunneling of bottom component for drain-bulk — 0.25
junction

499
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Effective mass (in units of m0) for trap-assisted
MEFFTATGAT tunneling of gate-edge component for source-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATGATD tunneling of gate-edge component for drain-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATSTI tunneling of STI-edge component for source-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATSTID tunneling of STI-edge component for drain-bulk — 0.25
junction
MUE Mobility reduction coefficient at TR m/V 0.5
Geometry independent mobility reduction coefficient
MUEO m/V 0.5
at TR
Width dependence of mobility reduction coefficient at
MUEW — 0
TR
NEFF Effective substrate doping m−3 5e+23
NEFFEDGE Effective substrate doping of edge transistors m−3 5e+23
NFA First coefficient of flicker noise — 8e+22
NFAEDGE First coefficient of flicker noise of edge transistors — 8e+22
First coefficient of flicker noise for 1 um**2 channel
NFAEDGELW — 8e+22
area
First coefficient of flicker noise for 1 um**2 channel
NFALW — 8e+22
area
NFB Second coefficient of flicker noise — 3e+07
NFBEDGE Second coefficient of flicker noise of edge transistors — 3e+07
Second coefficient of flicker noise for 1 um**2
NFBEDGELW — 3e+07
channel area
Second coefficient of flicker noise for 1 um**2
NFBLW — 3e+07
channel area
NFC Third coefficient of flicker noise V−1 0
NFCEDGE Third coefficient of flicker noise of edge transistors V−1 0
Third coefficient of flicker noise for 1 um**2 channel
NFCEDGELW V−1 0
area
Third coefficient of flicker noise for 1 um**2 channel
NFCLW V−1 0
area
NOV Effective doping of overlap region m−3 5e+25
NOVD Effective doping of overlap region for drain side m−3 5e+25
NOVDO Effective doping of overlap region for drain side m−3 5e+25

500
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
NOVO Effective doping of overlap region m−3 5e+25
NP Gate poly-silicon doping m−3 1e+26
NPCK Pocket doping level m−3 1e+24
Width dependence of pocket doping NPCK due to
NPCKW — 0
segregation
NPL Length dependence of gate poly-silicon doping — 0
NPO Geometry-independent gate poly-silicon doping m−3 1e+26
NSLP Effective doping bias-dependence parameter V 0.05
NSLPO Effective doping bias-dependence parameter V 0.05
NSUBEDGEL Length dependence of edge transistor substrate doping — 0
NSUBEDGELW Area dependence of edge transistor substrate doping — 0
Geometry independent substrate doping of edge
NSUBEDGEO m−3 5e+23
transistors
NSUBEDGEW Width dependence of edge transistor substrate doping — 0
NSUBO Geometry independent substrate doping m−3 3e+23
Width dependence of background doping NSUBO
NSUBW — 0
due to segregation
Grading coefficient of bottom component for
PBOT — 0.5
source-bulk junction
Grading coefficient of bottom component for
PBOTD — 0.5
drain-bulk junction
Breakdown onset tuning parameter of bottom
PBRBOT V 4
component for source-bulk junction
Breakdown onset tuning parameter of bottom
PBRBOTD V 4
component for drain-bulk junction
Breakdown onset tuning parameter of gate-edge
PBRGAT V 4
component for source-bulk junction
Breakdown onset tuning parameter of gate-edge
PBRGATD V 4
component for drain-bulk junction
Breakdown onset tuning parameter of STI-edge
PBRSTI V 4
component for source-bulk junction
Breakdown onset tuning parameter of STI-edge
PBRSTID V 4
component for drain-bulk junction
Grading coefficient of gate-edge component for
PGAT — 0.5
source-bulk junction
Grading coefficient of gate-edge component for
PGATD — 0.5
drain-bulk junction
Zero-temperature bandgap voltage of bottom
PHIGBOT V 1.16
component for source-bulk junction

501
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Zero-temperature bandgap voltage of bottom
PHIGBOTD V 1.16
component for drain-bulk junction
Zero-temperature bandgap voltage of gate-edge
PHIGGAT V 1.16
component for source-bulk junction
Zero-temperature bandgap voltage of gate-edge
PHIGGATD V 1.16
component for drain-bulk junction
Zero-temperature bandgap voltage of STI-edge
PHIGSTI V 1.16
component for source-bulk junction
Zero-temperature bandgap voltage of STI-edge
PHIGSTID V 1.16
component for drain-bulk junction
PKUO Cross-term dependence of KUO — 0
PKVTHO Cross-term dependence of KVTHO — 0
PLA1 Coefficient for the length dependence of A1 — 0
PLA3 Coefficient for the length dependence of A3 — 0
PLA4 Coefficient for the length dependence of A4 V−1/2 0
PLAGIDL Coefficient for the length dependence of AGIDL A/V3 0
Coefficient for the length dependence of AGIDL for
PLAGIDLD A/V3 0
drain side
PLALP Coefficient for the length dependence of ALP — 0
PLALP1 Coefficient for the length dependence of ALP1 V 0
PLALP2 Coefficient for the length dependence of ALP2 V−1 0
PLAX Coefficient for the length dependence of AX — 0
PLBETN Coefficient for the length dependence of BETN m2 /(Vs) 0
PLBETNEDGE Coefficient for the length dependence of BETNEDGE m2 /(Vs) 0
PLCF Coefficient for the length dependence of CF — 0
PLCFEDGE Coefficient for the length dependence of CFEDGE — 0
PLCFR Coefficient for the length dependence of CFR F 0
Coefficient for the length dependence of CFR for
PLCFRD F 0
drain side
PLCGBOV Coefficient for the length dependence of CGBOV F 0
PLCGOV Coefficient for the length dependence of CGOV F 0
Coefficient for the length dependence of CGOV for
PLCGOVD F 0
drain side
PLCOX Coefficient for the length dependence of COX F 0
PLCS Coefficient for the length dependence of CS — 0
PLCT Coefficient for the length dependence of CT — 0
PLCTEDGE Coefficient for the length dependence of CTEDGE — 0
PLDELVTAC Coefficient for the length dependence of DELVTAC V 0

502
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
PLDPHIB Coefficient for the length dependence of DPHIB V 0
Coefficient for the length dependence of
PLDPHIBEDGE V 0
DPHIBEDGE
PLFACNEFFAC Coefficient for the length dependence of FACNEFFAC — 0
PLFNTEXC Coefficient for the length dependence of FNTEXC — 0
PLGFACNUD Coefficient for the length dependence of GFACNUD — 0
PLIGINV Coefficient for the length dependence of IGINV A 0
PLIGOV Coefficient for the length dependence of IGOV A 0
Coefficient for the length dependence of IGOV for
PLIGOVD A 0
drain side
PLKUOWE Coefficient for the length dependence part of KUOWE — 0
Coefficient for the length dependence part of
PLKVTHOWE — 0
KVTHOWE
PLMUE Coefficient for the length dependence of MUE m/V 0
PLNEFF Coefficient for the length dependence of NEFF m−3 0
PLNEFFEDGE Coefficient for the length dependence of NEFFEDGE m−3 0
PLNFA Coefficient for the length dependence of NFA — 0
PLNFAEDGE Coefficient for the length dependence of NFAEDGE — 0
PLNFB Coefficient for the length dependence of NFB — 0
PLNFBEDGE Coefficient for the length dependence of NFBEDGE — 0
PLNFC Coefficient for the length dependence of NFC V−1 0
PLNFCEDGE Coefficient for the length dependence of NFCEDGE V−1 0
PLNOV Coefficient for the length dependence of NOV m−3 0
Coefficient for the length dependence of NOV for
PLNOVD m−3 0
drain side
PLNP Coefficient for the length dependence of NP m−3 0
PLPSCE Coefficient for the length dependence of PSCE — 0
PLPSCEEDGE Coefficient for the length dependence of PSCEEDGE — 0
PLRS Coefficient for the length dependence of RS ˙ 0
PLSTBET Coefficient for the length dependence of STBET — 0
Coefficient for the length dependence of
PLSTBETEDGE — 0
STBETEDGE
PLSTTHESAT Coefficient for the length dependence of STTHESAT — 0
PLSTVFB Coefficient for the length dependence of STVFB V/K 0
Coefficient for the length dependence of
PLSTVFBEDGE V/K 0
STVFBEDGE
PLTHESAT Coefficient for the length dependence of THESAT V−1 0
PLTHESATB Coefficient for the length dependence of THESATB V−1 0

503
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
PLTHESATG Coefficient for the length dependence of THESATG V−1 0
PLVFB Coefficient for the length dependence of VFB V 0
Coefficient for the length times width dependence of
PLWA1 — 0
A1
Coefficient for the length times width dependence of
PLWA3 — 0
A3
Coefficient for the length times width dependence of
PLWA4 V−1/2 0
A4
Coefficient for the length times width dependence of
PLWAGIDL A/V3 0
AGIDL
Coefficient for the length times width dependence of
PLWAGIDLD A/V3 0
AGIDL for drain side
Coefficient for the length times width dependence of
PLWALP — 0
ALP
Coefficient for the length times width dependence of
PLWALP1 V 0
ALP1
Coefficient for the length times width dependence of
PLWALP2 V−1 0
ALP2
Coefficient for the length times width dependence of
PLWAX — 0
AX
Coefficient for the length times width dependence of
PLWBETN m2 /(Vs) 0
BETN
Coefficient for the length times width dependence of
PLWBETNEDGE m2 /(Vs) 0
BETNEDGE
Coefficient for the length times width dependence of
PLWCF — 0
CF
Coefficient for the length times width dependence of
PLWCFEDGE — 0
CFEDGE
Coefficient for the length times width dependence of
PLWCFR F 0
CFR
Coefficient for the length times width dependence of
PLWCFRD F 0
CFR for drain side
Coefficient for the length times width dependence of
PLWCGBOV F 0
CGBOV
Coefficient for the length times width dependence of
PLWCGOV F 0
CGOV
Coefficient for the length times width dependence of
PLWCGOVD F 0
CGOV for drain side
Coefficient for the length times width dependence of
PLWCOX F 0
COX

504
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Coefficient for the length times width dependence of
PLWCS — 0
CS
Coefficient for the length times width dependence of
PLWCT — 0
CT
Coefficient for the length times width dependence of
PLWCTEDGE — 0
CTEDGE
Coefficient for the length times width dependence of
PLWDELVTAC V 0
DELVTAC
Coefficient for the length times width dependence of
PLWDPHIB V 0
DPHIB
Coefficient for the length times width dependence of
PLWDPHIBEDGE V 0
DPHIBEDGE
Coefficient for the length times width dependence of
PLWFACNEFFAC — 0
FACNEFFAC
Coefficient for the length times width dependence of
PLWFNTEXC — 0
FNTEXC
Coefficient for the length times width dependence of
PLWGFACNUD — 0
GFACNUD
Coefficient for the length times width dependence of
PLWIGINV A 0
IGINV
Coefficient for the length times width dependence of
PLWIGOV A 0
IGOV
Coefficient for the length times width dependence of
PLWIGOVD A 0
IGOV for drain side
Coefficient for the length times width dependence part
PLWKUOWE — 0
of KUOWE
Coefficient for the length times width dependence part
PLWKVTHOWE — 0
of KVTHOWE
Coefficient for the length times width dependence of
PLWMUE m/V 0
MUE
Coefficient for the length times width dependence of
PLWNEFF m−3 0
NEFF
Coefficient for the length times width dependence of
PLWNEFFEDGE m−3 0
NEFFEDGE
Coefficient for the length times width dependence of
PLWNFA — 0
NFA
Coefficient for the length times width dependence of
PLWNFAEDGE — 0
NFAEDGE
Coefficient for the length times width dependence of
PLWNFB — 0
NFB
Coefficient for the length times width dependence of
PLWNFBEDGE — 0
NFBEDGE

505
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Coefficient for the length times width dependence of
PLWNFC V−1 0
NFC
Coefficient for the length times width dependence of
PLWNFCEDGE V−1 0
NFCEDGE
Coefficient for the length times width dependence of
PLWNOV m−3 0
NOV
Coefficient for the length times width dependence of
PLWNOVD m−3 0
NOV for drain side
Coefficient for the length times width dependence of
PLWNP m−3 0
NP
Coefficient for the length times width dependence of
PLWPSCE — 0
PSCE
Coefficient for the length times width dependence of
PLWPSCEEDGE — 0
PSCEEDGE
Coefficient for the length times width dependence of
PLWRS ˙ 0
RS
Coefficient for the length times width dependence of
PLWSTBET — 0
STBET
Coefficient for the length times width dependence of
PLWSTBETEDGE — 0
STBETEDGE
Coefficient for the length times width dependence of
PLWSTTHESAT — 0
STTHESAT
Coefficient for the length times width dependence of
PLWSTVFB V/K 0
STVFB
Coefficient for the length times width dependence of
PLWSTVFBEDGE V/K 0
STVFBEDGE
Coefficient for the length times width dependence of
PLWTHESAT V−1 0
THESAT
Coefficient for the length times width dependence of
PLWTHESATB V−1 0
THESATB
Coefficient for the length times width dependence of
PLWTHESATG V−1 0
THESATG
Coefficient for the length times width dependence of
PLWVFB V 0
VFB
Coefficient for the length times width dependence of
PLWXCOR V−1 0
XCOR
PLXCOR Coefficient for the length dependence of XCOR V−1 0
POA1 Coefficient for the geometry independent part of A1 — 1
POA2 Coefficient for the geometry independent part of A2 V 10
POA3 Coefficient for the geometry independent part of A3 — 1
POA4 Coefficient for the geometry independent part of A4 V−1/2 0

506
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Coefficient for the geometry independent part of
POAGIDL A/V3 0
AGIDL
Coefficient for the geometry independent part of
POAGIDLD A/V3 0
AGIDL for drain side
POALP Coefficient for the geometry independent part of ALP — 0.01
Coefficient for the geometry independent part of
POALP1 V 0
ALP1
Coefficient for the geometry independent part of
POALP2 V−1 0
ALP2
POAX Coefficient for the geometry independent part of AX — 3
Coefficient for the geometry independent part of
POBETN m2 /(Vs) 0.07
BETN
Coefficient for the geometry independent part of
POBETNEDGE m2 /(Vs) 0.0005
BETNEDGE
Coefficient for the geometry independent part of
POBGIDL V 41
BGIDL
Coefficient for the geometry independent part of
POBGIDLD V 41
BGIDL for drain side
POCF Coefficient for the geometry independent part of CF — 0
POCFB Coefficient for the geometry independent part of CFB V−1 0
Coefficient for the geometry independent part of
POCFBEDGE V−1 0
CFBEDGE
POCFD Coefficient for the geometry independent part of CFD V−1 0
Coefficient for the geometry independent part of
POCFDEDGE V−1 0
CFDEDGE
Coefficient for the geometry independent part of
POCFEDGE — 0
CFEDGE
POCFR Coefficient for the geometry independent part of CFR F 0
Coefficient for the geometry independent part of CFR
POCFRD F 0
for drain side
Coefficient for the geometry independent part of
POCGBOV F 0
CGBOV
Coefficient for the geometry independent part of
POCGIDL — 0
CGIDL
Coefficient for the geometry independent part of
POCGIDLD — 0
CGIDL for drain side
Coefficient for the geometry independent part of
POCGOV F 1e-15
CGOV
Coefficient for the geometry independent part of
POCGOVD F 1e-15
CGOV for drain side

507
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Coefficient for the geometry independent part of
POCHIB V 3.1
CHIB
POCOX Coefficient for the geometry independent part of COX F 1e-14
POCS Coefficient for the geometry independent part of CS — 0
POCT Coefficient for the geometry independent part of CT — 0
Coefficient for the geometry independent part of
POCTEDGE — 0
CTEDGE
Coefficient for the geometry independent part of
PODELVTAC V 0
DELVTAC
Coefficient for the geometry independent part of
PODNSUB V−1 0
DNSUB
Coefficient for the geometry independent part of
PODPHIB V 0
DPHIB
Coefficient for the geometry independent part of
PODPHIBEDGE V 0
DPHIBEDGE
Coefficient for the geometry independent part of
PODVSBNUD V 1
DVSBNUD
POEF Coefficient for the flicker noise frequency exponent — 1
Coefficient for the geometry independent part of
POEFEDGE — 1
EFEDGE
Coefficient for the geometry independent part of
POEPSROX — 3.9
EPSOX
Coefficient for the geometry independent part of
POFACNEFFAC — 1
FACNEFFAC
Coefficient for the geometry independent part of
POFETA — 1
FETA
POFNT Coefficient for the geometry independent part of FNT — 1
Coefficient for the geometry independent part of
POFNTEDGE — 1
FNTEDGE
Coefficient for the geometry independent part of
POFNTEXC — 0
FNTEXC
POGC2 Coefficient for the geometry independent part of GC2 — 0.375
POGC3 Coefficient for the geometry independent part of GC3 — 0.063
POGCO Coefficient for the geometry independent part of GCO — 0
Coefficient for the geometry independent part of
POGFACNUD — 1
GFACNUD
Coefficient for the geometry independent part of
POIGINV A 0
IGINV
Coefficient for the geometry independent part of
POIGOV A 0
IGOV

508
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Coefficient for the geometry independent part of
POIGOVD A 0
IGOV for drain side
Coefficient for the geometry independent part of
POKUOWE — 0
KUOWE
Coefficient for the geometry independent part of
POKVTHOWE — 0
KVTHOWE
POMUE Coefficient for the geometry independent part of MUE m/V 0.5
Coefficient for the geometry independent part of
PONEFF m−3 5e+23
NEFF
Coefficient for the geometry independent part of
PONEFFEDGE m−3 5e+23
NEFFEDGE
PONFA Coefficient for the geometry independent part of NFA — 8e+22
Coefficient for the geometry independent part of
PONFAEDGE — 8e+22
NFAEDGE
PONFB Coefficient for the geometry independent part of NFB — 3e+07
Coefficient for the geometry independent part of
PONFBEDGE — 3e+07
NFBEDGE
PONFC Coefficient for the geometry independent part of NFC V−1 0
Coefficient for the geometry independent part of
PONFCEDGE V−1 0
NFCEDGE
PONOV Coefficient for the geometry independent part of NOV m−3 5e+25
Coefficient for the geometry independent part of NOV
PONOVD m−3 5e+25
for drain side
PONP Coefficient for the geometry independent part of NP m−3 1e+26
Coefficient for the geometry independent part of
PONSLP V 0.05
NSLP
Coefficient for the geometry independent part of
POPSCE — 0
PSCE
Coefficient for the geometry independent part of
POPSCEB V−1 0
PSCEB
Coefficient for the geometry independent part of
POPSCEBEDGE V−1 0
PSCEBEDGE
Coefficient for the geometry independent part of
POPSCED V−1 0
PSCED
Coefficient for the geometry independent part of
POPSCEDEDGE V−1 0
PSCEDEDGE
Coefficient for the geometry independent part of
POPSCEEDGE — 0
PSCEEDGE
PORS Coefficient for the geometry independent part of RS ˙ 30
PORSB Coefficient for the geometry independent part of RSB V−1 0

509
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
PORSG Coefficient for the geometry independent part of RSG V−1 0
POSTA2 Coefficient for the geometry independent part of STA2 V 0
Coefficient for the geometry independent part of
POSTBET — 1
STBET
Coefficient for the geometry independent part of
POSTBETEDGE — 1
STBETEDGE
Coefficient for the geometry independent part of
POSTBGIDL V/K 0
STBGIDL
Coefficient for the geometry independent part of
POSTBGIDLD V/K 0
STBGIDL for drain side
POSTCS Coefficient for the geometry independent part of STCS — 0
POSTIG Coefficient for the geometry independent part of STIG — 2
Coefficient for the geometry independent part of
POSTMUE — 0
STMUE
POSTRS Coefficient for the geometry independent part of STRS — 1
Coefficient for the geometry independent part of
POSTTHEMU — 1.5
STTHEMU
Coefficient for the geometry independent part of
POSTTHESAT — 1
STTHESAT
Coefficient for the geometry independent part of
POSTVFB V/K 0.0005
STVFB
Coefficient for the geometry independent part of
POSTVFBEDGE V/K 0
STVFBEDGE
Coefficient for the geometry independent part of
POSTXCOR — 0
STXCOR
Coefficient for the geometry independent part of
POTHEMU — 1.5
THEMU
Coefficient for the geometry independent part of
POTHESAT V−1 1
THESAT
Coefficient for the geometry independent part of
POTHESATB V−1 0
THESATB
Coefficient for the geometry independent part of
POTHESATG V−1 0
THESATG
POTOX Coefficient for the geometry independent part of TOX m 2e-09
Coefficient for the geometry independent part of
POTOXOV m 2e-09
TOXOV
Coefficient for the geometry independent part of
POTOXOVD m 2e-09
TOXOV for drain side
POVFB Coefficient for the geometry independent part of VFB V -1
Coefficient for the geometry independent part of
POVFBEDGE V -1
VFBEDGE

510
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Coefficient for the geometry independent part of
POVNSUB V 0
VNSUB
POVP Coefficient for the geometry independent part of VP V 0.05
Coefficient for the geometry independent part of
POVSBNUD V 0
VSBNUD
Coefficient for the geometry independent part of
POXCOR V−1 0
XCOR
Subthreshold slope coefficient for short channel
PSCE — 0
transistor
Bulk voltage dependence parameter of subthreshold
PSCEB V−1 0
slope coefficient for short channel transistor
Bulk voltage dependence parameter of subthreshold
PSCEBEDGE V−1 0
slope coefficient for short channel edge transistors
Bulk voltage dependence parameter of subthreshold
PSCEBEDGEO V−1 0
slope coefficient for short channel edge transistors
Bulk voltage dependence parameter of subthreshold
PSCEBO V−1 0
slope coefficient for short channel transistor
Drain voltage dependence parameter of subthreshold
PSCED V−1 0
slope coefficient for short channel transistor
Drain voltage dependence parameter of subthreshold
PSCEDEDGE V−1 0
slope coefficient for short channel edge transistors
Drain voltage dependence parameter of subthreshold
PSCEDEDGEO V−1 0
slope coefficient for short channel edge transistors
Drain voltage dependence parameter of subthreshold
PSCEDO V−1 0
slope coefficient for short channel transistor
Subthreshold slope coefficient for short channel edge
PSCEEDGE — 0
transistors
Length dependence of subthreshold slope coefficient
PSCEEDGEL — 0
for short channel edge transistors
Exponent for length dependence of subthreshold slope
PSCEEDGELEXP — 2
coefficient for short channel edge transistors
Exponent for length dependence of subthreshold slope
PSCEEDGEW — 0
coefficient for short channel edge transistor
Length dependence of subthreshold slope coefficient
PSCEL — 0
for short channel transistor
Exponent for length dependence of subthreshold slope
PSCELEXP — 2
coefficient for short channel transistor
Exponent for length dependence of subthreshold slope
PSCEW — 0
coefficient for short channel transistor
Grading coefficient of STI-edge component for
PSTI — 0.5
source-bulk junction

511
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Grading coefficient of STI-edge component for
PSTID — 0.5
drain-bulk junction
PWA1 Coefficient for the width dependence of A1 — 0
PWA3 Coefficient for the width dependence of A3 — 0
PWA4 Coefficient for the width dependence of A4 V−1/2 0
PWAGIDL Coefficient for the width dependence of AGIDL A/V3 0
Coefficient for the width dependence of AGIDL for
PWAGIDLD A/V3 0
drain side
PWALP Coefficient for the width dependence of ALP — 0
PWALP1 Coefficient for the width dependence of ALP1 V 0
PWALP2 Coefficient for the width dependence of ALP2 V−1 0
PWAX Coefficient for the width dependence of AX — 0
PWBETN Coefficient for the width dependence of BETN m2 /(Vs) 0
PWBETNEDGE Coefficient for the width dependence of BETNEDGE m2 /(Vs) 0
PWCF Coefficient for the width dependence of CF — 0
PWCFEDGE Coefficient for the width dependence of CFEDGE — 0
PWCFR Coefficient for the width dependence of CFR F 0
Coefficient for the width dependence of CFR for drain
PWCFRD F 0
side
PWCGBOV Coefficient for the width dependence of CGBOV F 0
PWCGOV Coefficient for the width dependence of CGOV F 0
Coefficient for the width dependence of CGOV for
PWCGOVD F 0
drain side
PWCOX Coefficient for the width dependence of COX F 0
PWCS Coefficient for the width dependence of CS — 0
PWCT Coefficient for the width dependence of CT — 0
PWCTEDGE Coefficient for the width dependence of CTEDGE — 0
PWDELVTAC Coefficient for the width dependence of DELVTAC V 0
PWDPHIB Coefficient for the width dependence of DPHIB V 0
PWDPHIBEDGE Coefficient for the width dependence of DPHIBEDGE V 0
PWFACNEFFAC Coefficient for the width dependence of FACNEFFAC — 0
PWFNTEXC Coefficient for the width dependence of FNTEXC — 0
PWGFACNUD Coefficient for the width dependence of GFACNUD — 0
PWIGINV Coefficient for the width dependence of IGINV A 0
PWIGOV Coefficient for the width dependence of IGOV A 0
Coefficient for the width dependence of IGOV for
PWIGOVD A 0
drain side

512
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
PWKUOWE Coefficient for the width dependence part of KUOWE — 0
Coefficient for the width dependence part of
PWKVTHOWE — 0
KVTHOWE
PWMUE Coefficient for the width dependence of MUE m/V 0
PWNEFF Coefficient for the width dependence of NEFF m−3 0
PWNEFFEDGE Coefficient for the width dependence of NEFFEDGE m−3 0
PWNFA Coefficient for the width dependence of NFA — 0
PWNFAEDGE Coefficient for the width dependence of NFAEDGE — 0
PWNFB Coefficient for the width dependence of NFB — 0
PWNFBEDGE Coefficient for the width dependence of NFBEDGE — 0
PWNFC Coefficient for the width dependence of NFC V−1 0
PWNFCEDGE Coefficient for the width dependence of NFCEDGE V−1 0
PWNOV Coefficient for the width dependence of NOV m−3 0
Coefficient for the width dependence of NOV for drain
PWNOVD m−3 0
side
PWNP Coefficient for the width dependence of NP m−3 0
PWPSCE Coefficient for the width dependence of PSCE — 0
PWPSCEEDGE Coefficient for the width dependence of PSCEEDGE — 0
PWRS Coefficient for the width dependence of RS ˙ 0
PWSTBET Coefficient for the width dependence of STBET — 0
PWSTBETEDGE Coefficient for the width dependence of STBETEDGE — 0
PWSTTHESAT Coefficient for the width dependence of STTHESAT — 0
PWSTVFB Coefficient for the width dependence of STVFB V/K 0
PWSTVFBEDGE Coefficient for the width dependence of STVFBEDGE V/K 0
PWTHESAT Coefficient for the width dependence of THESAT V−1 0
PWTHESATB Coefficient for the width dependence of THESATB V−1 0
PWTHESATG Coefficient for the width dependence of THESATG V−1 0
PWVFB Coefficient for the width dependence of VFB V 0
PWXCOR Coefficient for the width dependence of XCOR V−1 0
QMC Quantum-mechanical correction factor — 1
RBULK Bulk resistance between node BP and BI ˙ 0
RBULKO Bulk resistance between node BP and BI ˙ 0
RDE External drain resistance ˙ 0
RG Gate resistance ˙ 0
RGO Gate resistance ˙ 0
RINT Contact resistance between silicide and ploy ˙ m2 0

513
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
RJUND Drain-side bulk resistance between node BI and BD ˙ 0
RJUNDO Drain-side bulk resistance between node BI and BD ˙ 0
RJUNS Source-side bulk resistance between node BI and BS ˙ 0
RJUNSO Source-side bulk resistance between node BI and BS ˙ 0
RS Series resistance at TR ˙ 30
RSB Back-bias dependence of series resistance V−1 0
RSBO Back-bias dependence of series resistance V−1 0
RSE External source resistance ˙ 0
RSG Gate-bias dependence of series resistance V−1 0
RSGO Gate-bias dependence of series resistance V−1 0
RSH Sheet resistance of source diffusion ˙/□ 0
RSHD Sheet resistance of drain diffusion ˙/□ 0
RSHG Gate electrode diffusion sheet resistance ˙/□ 0
Source/drain series resistance for 1 um wide channel
RSW1 ˙ 50
at TR
RSW2 Higher-order width scaling of RS — 0
RVPOLY Vertical poly resistance ˙ m2 0
RWELL Well resistance between node BI and B ˙ 0
RWELLO Well resistance between node BI and B ˙ 0
Reference distance between OD-edge and poly from
SAREF m 1e-06
one side
Reference distance between OD-edge and poly from
SBREF m 1e-06
other side
Distance between OD-edge and well edge of a
SCREF m 1e-06
reference device
STA2 Temperature dependence of A2 V 0
STA2O Temperature dependence of A2 V 0
STBET Temperature dependence of BETN — 1
STBETEDGE Temperature dependence of BETNEDGE — 1
Length dependence of temperature dependence of
STBETEDGEL — 0
BETNEDGE
Area dependence of temperature dependence of
STBETEDGELW — 0
BETNEDGE
Geometry independent temperature dependence of
STBETEDGEO — 1
BETNEDGE
Width dependence of temperature dependence of
STBETEDGEW — 0
BETNEDGE

514
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Length dependence of temperature dependence of
STBETL — 0
BETN
Area dependence of temperature dependence of
STBETLW — 0
BETN
Geometry independent temperature dependence of
STBETO — 1
BETN
Width dependence of temperature dependence of
STBETW — 0
BETN
STBGIDL Temperature dependence of BGIDL V/K 0
STBGIDLD Temperature dependence of BGIDL for drain side V/K 0
STBGIDLDO Temperature dependence of BGIDL for drain side V/K 0
STBGIDLO Temperature dependence of BGIDL V/K 0
STCS Temperature dependence of CS — 0
STCSO Temperature dependence of CS — 0
STETAO Eta0 shift factor related to VTHO change m 0
Temperature scaling parameter for band-to-band
STFBBTBOT tunneling of bottom component for source-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTBOTD tunneling of bottom component for drain-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTGAT tunneling of gate-edge component for source-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTGATD tunneling of gate-edge component for drain-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTSTI tunneling of STI-edge component for source-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTSTID tunneling of STI-edge component for drain-bulk 1/K -0.001
junction
STIG Temperature dependence of IGINV and IGOV — 2
STIGO Temperature dependence of IGINV and IGOV — 2
STMUE Temperature dependence of MUE — 0
STMUEO Temperature dependence of MUE — 0
STRS Temperature dependence of RS — 1
STRSO Temperature dependence of RS — 1
STTHEMU Temperature dependence of THEMU — 1.5

515
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
STTHEMUO Temperature dependence of THEMU — 1.5
STTHESAT Temperature dependence of THESAT — 1
Length dependence of temperature dependence of
STTHESATL — 0
THESAT
Area dependence of temperature dependence of
STTHESATLW — 0
THESAT
Geometry independent temperature dependence of
STTHESATO — 1
THESAT
Width dependence of temperature dependence of
STTHESATW — 0
THESAT
STVFB Temperature dependence of VFB V/K 0.0005
STVFBEDGE Temperature dependence of VFBEDGE V/K 0.0005
Length dependence of temperature dependence of
STVFBEDGEL V/K 0
VFBEDGE
Area dependence of temperature dependence of
STVFBEDGELW V/K 0
VFBEDGE
Geometry-independent temperature dependence of
STVFBEDGEO V/K 0.0005
VFBEDGE
Width dependence of temperature dependence of
STVFBEDGEW V/K 0
VFBEDGE
Length dependence of temperature dependence of
STVFBL V/K 0
VFB
STVFBLW Area dependence of temperature dependence of VFB V/K 0
Geometry-independent temperature dependence of
STVFBO V/K 0.0005
VFB
STVFBW Width dependence of temperature dependence of VFB V/K 0
STXCOR Temperature dependence of XCOR — 0
STXCORO Temperature dependence of XCOR — 0
SWDELVTAC Flag for separate capacitance calculation; 0=off, 1=on — 0
SWEDGE Flag for drain current of edge transistors; 0=off, 1=on — 0
Flag for geometrical model, 0=local, 1=global,
SWGEO — 1
2=binning
SWGIDL Flag for GIDL current, 0=turn off IGIDL — 0
SWIGATE Flag for gate current, 0=turn off IG — 0
SWIGN Flag for induced gate noise; 0=off, 1=on — 1
SWIMPACT Flag for impact ionization current, 0=turn off II — 0
Flag for asymmetric junctions; 0=symmetric,
SWJUNASYM — 0
1=asymmetric
SWJUNCAP Flag for juncap, 0=turn off juncap — 0

516
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Flag for JUNCAP-express; 0=full model, 1=express
SWJUNEXP — 0
model
Flag for NUD-effect; 0=off, 1=on,
SWNUD — 0
2=on+CV-correction
THEMU Mobility reduction exponent at TR — 1.5
THEMUO Mobility reduction exponent at TR — 1.5
THESAT Velocity saturation parameter at TR V−1 1
THESATB Back-bias dependence of velocity saturation V−1 0
THESATBO Back-bias dependence of velocity saturation V−1 0
THESATG Gate-bias dependence of velocity saturation V−1 0
THESATGO Gate-bias dependence of velocity saturation V−1 0
THESATL Length dependence of THESAT V−1 0.05
THESATLEXP Exponent for length dependence of THESAT — 1
THESATLW Area dependence of velocity saturation parameter — 0
Geometry independent velocity saturation parameter
THESATO V−1 0
at TR
THESATW Width dependence of velocity saturation parameter — 0
TKUO Temperature dependence of KUO — 0
TOX Gate oxide thickness m 2e-09
TOXO Gate oxide thickness m 2e-09
TOXOV Overlap oxide thickness m 2e-09
TOXOVD Overlap oxide thickness for drain side m 2e-09
TOXOVDO Overlap oxide thickness for drain side m 2e-09
TOXOVO Overlap oxide thickness m 2e-09
TR nominal (reference) temperature ◦C 21
TRJ Reference temperature ◦C 21
TYPE Channel type parameter, +1=NMOS -1=PMOS — 1
UO Zero-field mobility at TR m2 /(Vs) 0.05
Built-in voltage at the reference temperature of
VBIRBOT V 1
bottom component for source-bulk junction
Built-in voltage at the reference temperature of
VBIRBOTD V 1
bottom component for drain-bulk junction
Built-in voltage at the reference temperature of
VBIRGAT V 1
gate-edge component for source-bulk junction
Built-in voltage at the reference temperature of
VBIRGATD V 1
gate-edge component for drain-bulk junction
Built-in voltage at the reference temperature of
VBIRSTI V 1
STI-edge component for source-bulk junction

517
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
Built-in voltage at the reference temperature of
VBIRSTID V 1
STI-edge component for drain-bulk junction
Breakdown voltage of bottom component for
VBRBOT V 10
source-bulk junction
Breakdown voltage of bottom component for
VBRBOTD V 10
drain-bulk junction
Breakdown voltage of gate-edge component for
VBRGAT V 10
source-bulk junction
Breakdown voltage of gate-edge component for
VBRGATD V 10
drain-bulk junction
Breakdown voltage of STI-edge component for
VBRSTI V 10
source-bulk junction
Breakdown voltage of STI-edge component for
VBRSTID V 10
drain-bulk junction
VFB Flat band voltage at TR V -1
VFBEDGE Flat band voltage of edge transistors at TR V -1
Geometry-independent flat-band voltage of edge
VFBEDGEO V -1
transistors at TR
VFBL Length dependence of flat-band voltage V 0
VFBLW Area dependence of flat-band voltage V 0
VFBO Geometry-independent flat-band voltage at TR V -1
VFBW Width dependence of flat-band voltage V 0
Typical maximum source-bulk junction voltage;
VJUNREF V 2.5
usually about 2*VSUP
Typical maximum drain-bulk junction voltage; usually
VJUNREFD V 2.5
about 2*VSUP
VNSUB Effective doping bias-dependence parameter V 0
VNSUBO Effective doping bias-dependence parameter V 0
VP CLM logarithm dependence factor V 0.05
VPO CLM logarithmic dependence parameter V 0.05
VSBNUD Lower Vsb value for NUD-effect V 0
VSBNUDO Lower Vsb value for NUD-effect V 0
WBET Characteristic width for width scaling of BETN m 1e-09
WEB Coefficient for SCB — 0
WEC Coefficient for SCC — 0
WEDGE Electrical width of edge transistor per side m 1e-08
WEDGEW Width dependence of edge WEDGE — 0
WKUO Width dependence of KUO — 0
WKVTHO Width dependence of KVTHO — 0

518
 Table 2-120. PSP103VA MOSFET Device Model Parameters
Parameter Description Units Default
WLOD Width parameter m 0
WLODKUO Width parameter for UO stress effect — 0
WLODVTH Width parameter for VTH-stress effect — 0
WMAX Dummy parameter to label binning set m 1
WMIN Dummy parameter to label binning set m 0
WOT Effective channel width reduction per side m 0
Char. length of segregation of background doping
WSEG m 1e-08
NSUBO
WSEGP Char. length of segregation of pocket doping NPCK m 1e-08
WVARL Length dependence of WVAR — 0
Geom. independent difference between actual and
WVARO m 0
programmed field-oxide opening
WVARW Width dependence of WVAR — 0
XCOR Non-universality factor V−1 0
XCORL Length dependence of non-universality parameter — 0
XCORLW Area dependence of non-universality parameter — 0
XCORO Geometry independent non-universality parameter V−1 0
XCORW Width dependence of non-universality parameter — 0
Junction depth of gate-edge component for
XJUNGAT m 1e-07
source-bulk junction
Junction depth of gate-edge component for drain-bulk
XJUNGATD m 1e-07
junction
Junction depth of STI-edge component for
XJUNSTI m 1e-07
source-bulk junction
Junction depth of STI-edge component for drain-bulk
XJUNSTID m 1e-07
junction

Table 2-121. MOSFET level 103 Output Variables

Parameter Description Units Default
ctype Flag for channel type none
sdint Flag for source-drain interchange none
ise Total source current A none
ige Total gate current A none
ide Total drain current A none
ibe Total bulk current A none
Drain current, excl. edge transistor currents,
ids A none
avalanche, tunnel, GISL, GIDL, and junction currents

519
 Table 2-121. MOSFET level 103 Output Variables
Parameter Description Units Default
idb Drain to bulk current A none
isb Source to bulk current A none
igs Gate-source tunneling current A none
igd Gate-drain tunneling current A none
igb Gate-bulk tunneling current A none
idedge Drain current of edge transistors A none
igcs Gate-channel tunneling current (source component) A none
igcd Gate-channel tunneling current (drain component) A none
iavl Substrate current due to weak avelanche A none
igisl Gate-induced source leakage current A none
igidl Gate-induced drain leakage current A none
ijs Total source junction current A none
ijsbot Source junction current (bottom component) A none
ijsgat Source junction current (gate-edge component) A none
ijssti Source junction current (STI-edge component) A none
ijd Total drain junction current A none
ijdbot Drain junction current (bottom component) A none
ijdgat Drain junction current (gate-edge component) A none
ijdsti Drain junction current (STI-edge component) A none
vds Drain-source voltage V none
vgs Gate-source voltage V none
vsb Source-bulk voltage V none
vto Zero-bias threshold voltage V none
vts Threshold voltage including back bias effects V none
Threshold voltage including back bias and drain bias
vth V none
effects
Effective gate drive voltage including back bias and
vgt V none
drain bias effects
vdss Drain saturation voltage at actual bias V none
vsat Saturation limit none
gm Transconductance ˙ −1 none
gmb Substrate transconductance ˙ −1 none
gds Output conductance ˙ −1 none
gjs Source junction conductance ˙ −1 none
gjd Drain junction conductance ˙ −1 none
cdd Drain capacitance F none

520
 Table 2-121. MOSFET level 103 Output Variables
Parameter Description Units Default
cdg Drain-gate capacitance F none
cds Drain-source capacitance F none
cdb Drain-bulk capacitance F none
cgd Gate-drain capacitance F none
cgg Gate capacitance F none
cgs Gate-source capacitance F none
cgb Gate-bulk capacitance F none
csd Source-drain capacitance F none
csg Source-gate capacitance F none
css Source capacitance F none
csb Source-bulk capacitance F none
cbd Bulk-drain capacitance F none
cbg Bulk-gate capacitance F none
cbs Bulk-source capacitance F none
cbb Bulk capacitance F none
cgsol Total gate-source overlap capacitance F none
cgdol Total gate-drain overlap capacitance F none
cjs Total source junction capacitance F none
cjsbot Source junction capacitance (bottom component) F none
cjsgat Source junction capacitance (gate-edge component) F none
cjssti Source junction capacitance (STI-edge component) F none
cjd Total drain junction capacitance F none
cjdbot Drain junction capacitance (bottom component) F none
cjdgat Drain junction capacitance (gate-edge component) F none
cjdsti Drain junction capacitance (STI-edge component) F none
weff Effective channel width for geometrical models m none
leff Effective channel length for geometrical models m none
u Transistor gain none
rout Small-signal output resistance Ohm none
vearly Equivalent Early voltage V none
beff Gain factor A/V2 none
fug Unity gain frequency at actual bias Hz none
rg Gate resistance Ohm none
sfl Flicker noise current spectral density at 1 Hz A2 /Hz none
Input-referred RMS white noise voltage spectral
sqrtsff V/sqrt(Hz) none
density at 1 kHz

521
 Table 2-121. MOSFET level 103 Output Variables
Parameter Description Units Default
Input-referred RMS white noise voltage spectral
sqrtsfw V/sqrt(Hz) none
density
sid White noise current spectral density A2 /Hz none
sig Induced gate noise current spectral density at 1 Hz A2 /Hz none
Imaginary part of correlation coefficient between Sig
cigid none
and Sid
Cross-over frequency above which white noise is
fknee Hz none
dominant
sigs Gate-source current noise spectral density A2 /Hz none
sigd Gate-drain current noise spectral density A2 /Hz none
siavl Impact ionization current noise spectral density A2 /Hz none
ssi Total source junction current noise spectral density A2 /Hz none
sdi Total drain junction current noise spectral density A2 /Hz none
Flicker noise current spectral density at 1 Hz of edge
sfledge A2 /Hz none
transistors
White noise current spectral density of edge
sidedge A2 /Hz none
transistors
lp_vfb Local parameter VFB after T-scaling and clipping V none
lp_stvfb Local parameter STVFB after clipping V/K none
lp_tox Local parameter TOX after clipping m none
lp_epsrox Local parameter EPSROX after clipping none
lp_neff Local parameter NEFF after clipping m−3 none
lp_facneffac Local parameter FACNEFFAC after clipping none
lp_gfacnud Local parameter GFACNUD after clipping none
lp_vsbnud Local parameter VSBNUD after clipping V none
lp_dvsbnud Local parameter DVSBNUD after clipping V none
lp_vnsub Local parameter VNSUB after clipping V none
lp_nslp Local parameter NSLP after clipping V none
lp_dnsub Local parameter DNSUB after clipping V−1 none
lp_dphib Local parameter DPHIB after clipping V none
lp_delvtac Local parameter DELVTAC after clipping V none
lp_np Local parameter NP after clipping m−3 none
lp_ct Local parameter CT after clipping none
lp_toxov Local parameter TOXOV after clipping m none
lp_toxovd Local parameter TOXOVD after clipping m none
lp_nov Local parameter NOV after clipping m−3 none
lp_novd Local parameter NOVD after clipping m−3 none

522
 Table 2-121. MOSFET level 103 Output Variables
Parameter Description Units Default
lp_cf Local parameter CF after clipping none
lp_cfd Local parameter CFD after clipping V−1 none
lp_cfb Local parameter CFB after clipping V−1 none
lp_psce Local parameter PSCE after clipping none
lp_psceb Local parameter PSCEB after clipping V−1 none
lp_psced Local parameter PSCED after clipping V−1 none
lp_betn Local parameter BETN after T-scaling and clipping m2 /(V s) none
lp_stbet Local parameter STBET after clipping none
lp_mue Local parameter MUE after T-scaling and clipping m/V none
lp_stmue Local parameter STMUE after clipping none
lp_themu Local parameter THEMU after T-scaling and clipping none
lp_stthemu Local parameter STTHEMU after clipping none
lp_cs Local parameter CS after T-scaling and clipping none
lp_stcs Local parameter STCS after clipping none
lp_xcor Local parameter XCOR after T-scaling and clipping V−1 none
lp_stxcor Local parameter STXCOR after clipping none
lp_feta Local parameter FETA after clipping none
lp_rs Local parameter RS after T-scaling and clipping Ohm none
lp_strs Local parameter STRS after clipping none
lp_rsb Local parameter RSB after clipping V−1 none
lp_rsg Local parameter RSG after clipping V−1 none
lp_thesat Local parameter THESAT after T-scaling and clipping V−1 none
lp_stthesat Local parameter STTHESAT after clipping none
lp_thesatb Local parameter THESATB after clipping V−1 none
lp_thesatg Local parameter THESATG after clipping V−1 none
lp_ax Local parameter AX after clipping none
lp_alp Local parameter ALP after clipping none
lp_alp1 Local parameter ALP1 after clipping V none
lp_alp2 Local parameter ALP2 after clipping V−1 none
lp_vp Local parameter VP after clipping V none
lp_a1 Local parameter A1 after clipping none
lp_a2 Local parameter A2 after T-scaling and clipping V none
lp_sta2 Local parameter STA2 after clipping none
lp_a3 Local parameter A3 after clipping none
lp_a4 Local parameter A4 after clipping 1/sqrt(V) none
lp_gco Local parameter GCO after clipping none

523
 Table 2-121. MOSFET level 103 Output Variables
Parameter Description Units Default
lp_iginv Local parameter IGINV after T-scaling and clipping A none
lp_igov Local parameter IGOV after T-scaling and clipping A none
lp_igovd Local parameter IGOVD after T-scaling and clipping A none
lp_stig Local parameter STIG after clipping none
lp_gc2 Local parameter GC2 after clipping none
lp_gc3 Local parameter GC3 after clipping none
lp_chib Local parameter CHIB after clipping V none
lp_agidl Local parameter AGIDL after clipping A/V3 none
lp_agidld Local parameter AGIDLD after clipping A/V3 none
lp_bgidl Local parameter BGIDL after T-scaling and clipping V none
lp_bgidld Local parameter BGIDLD after T-scaling and clipping V none
lp_stbgidl Local parameter STBGIDL after clipping V/K none
lp_stbgidld Local parameter STBGIDLD after clipping V/K none
lp_cgidl Local parameter CGIDL after clipping none
lp_cgidld Local parameter CGIDLD after clipping none
lp_cox Local parameter COX after clipping F none
lp_cgov Local parameter CGOV after clipping F none
lp_cgovd Local parameter CGOVD after clipping F none
lp_cgbov Local parameter CGBOV after clipping F none
lp_cfr Local parameter CFR after clipping F none
lp_cfrd Local parameter CFRD after clipping F none
lp_fnt Local parameter FNT after clipping none
lp_fntexc Local parameter FNTEXC after clipping none
lp_nfa Local parameter NFA after clipping 1/(V m4 ) none
lp_nfb Local parameter NFB after clipping 1/(V m4 ) none
lp_nfc Local parameter NFC after clipping V−1 none
lp_ef Local parameter EF after clipping none
Local parameter VFBEDGE after T-scaling and
lp_vfbedge V none
clipping
lp_stvfbedge Local parameter STVFBEDGE after clipping V/K none
lp_dphibedge Local parameter DPHIBEDGE after clipping V none
lp_neffedge Local parameter NEFFEDGE after clipping m−3 none
lp_ctedge Local parameter CTEDGE after clipping none
Local parameter BETNEDGE after T-scaling and
lp_betnedge m2 /(Vs) none
clipping
lp_stbetedge Local parameter STBETEDGE after clipping none

524
 Table 2-121. MOSFET level 103 Output Variables
Parameter Description Units Default
lp_psceedge Local parameter PSCEEDGE after clipping none
lp_pscebedge Local parameter PSCEBEDGE after clipping V−1 none
lp_pscededge Local parameter PSCEDEDGE after clipping V−1 none
lp_cfedge Local parameter CFEDGE after clipping V none
lp_cfdedge Local parameter CFDEDGE after clipping V−1 none
lp_cfbedge Local parameter CFBEDGE after clipping V−1 none
lp_fntedge Local parameter FNTEDGE after clipping none
lp_nfaedge Local parameter NFAEDGE after clipping 1/(V m4 ) none
lp_nfbedge Local parameter NFBEDGE after clipping 1/(V m4 ) none
lp_nfcedge Local parameter NFCEDGE after clipping V−1 none
lp_efedge Local parameter EFEDGE after clipping none
lp_rg Local parameter RG after clipping Ohm none
lp_rse Local parameter RSE after clipping Ohm none
lp_rde Local parameter RDE after clipping Ohm none
lp_rbulk Local parameter RBULK after clipping Ohm none
lp_rwell Local parameter RWELL after clipping Ohm none
lp_rjuns Local parameter RJUNS after clipping Ohm none
lp_rjund Local parameter RJUND after clipping Ohm none
tk Device Temperature K none
Bottom component of total zero-bias source junction
cjosbot F none
capacitance at device temperature
STI-edge component of total zero-bias source junction
cjossti F none
capacitance at device temperature
Gate-edge component of total zero-bias source
cjosgat F none
junction capacitance at device temperature
Built-in voltage of source-side bottom junction at
vbisbot V none
device temperature
Built-in voltage of source-side STI-edge junction at
vbissti V none
device temperature
Built-in voltage of source-side gate-edge junction at
vbisgat V none
device temperature
idsatsbot Total source-side bottom junction saturation current A none
idsatssti Total source-side STI-edge junction saturation current A none
idsatsgat Total source-side gate-edge junction saturation current A none
Bottom component of total zero-bias drain junction
cjosbotd F none
capacitance at device temperature
STI-edge component of total zero-bias drain junction
cjosstid F none
capacitance at device temperature

525
 Table 2-121. MOSFET level 103 Output Variables
Parameter Description Units Default
Gate-edge component of total zero-bias drain junction
cjosgatd F none
capacitance at device temperature
Built-in voltage of drain-side bottom junction at
vbisbotd V none
device temperature
Built-in voltage of drain-side STI-edge junction at
vbisstid V none
device temperature
Built-in voltage of drain-side gate-edge junction at
vbisgatd V none
device temperature
idsatsbotd Total drain-side bottom junction saturation current A none
idsatsstid Total drain-side STI-edge junction saturation current A none
idsatsgatd Total drain-side gate-edge junction saturation current A none

Table 2-122. PSP103VA MOSFET with self-heating Device Instance Parameters

Parameter Description Units Default
ABDRAIN Bottom area of drain junction m2 1e-12
ABSOURCE Bottom area of source junction m2 1e-12
AD Bottom area of drain junction m2 1e-12
AS Bottom area of source junction m2 1e-12
DELVTO Threshold voltage shift parameter V 0
DELVTOEDGE Threshold voltage shift parameter of edge transistor V 0
DTA Temperature offset w.r.t. ambient temperature K 0
FACTUO Zero-field mobility pre-factor — 1
FACTUOEDGE Zero-field mobility pre-factor of edge transistor — 1
JW Gate-edge length of source/drain junction m 1e-06
L Design length m 1e-05
LGDRAIN Gate-edge length of drain junction m 1e-06
LGSOURCE Gate-edge length of source junction m 1e-06
LSDRAIN STI-edge length of drain junction m 1e-06
LSSOURCE STI-edge length of source junction m 1e-06
M Alias for MULT — 1
MULT Number of devices in parallel — 1
NF Number of fingers — 1
NGCON Number of gate contacts — 1
NRD Number of squares of drain diffusion — 0
NRS Number of squares of source diffusion — 0
PD Perimeter of drain junction m 1e-06

526
 Table 2-122. PSP103VA MOSFET with self-heating Device Instance Parameters
Parameter Description Units Default
PS Perimeter of source junction m 1e-06
SA Distance between OD-edge and poly from one side m 0
SB Distance between OD-edge and poly from other side m 0
SC Distance between OD-edge and nearest well edge m 0
Integral of the first distribution function for scattered
SCA — 0
well dopants
Integral of the second distribution function for
SCB — 0
scattered well dopants
Integral of the third distribution function for scattered
SCC — 0
well dopants
SD Distance between neighbouring fingers m 0
W Design width m 1e-05
XGW Distance from the gate contact to the channel edge m 1e-07

Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters

Parameter Description Units Default
A1 Impact-ionization pre-factor — 1
A1L Length dependence of A1 — 0
A1O Geometry independent impact-ionization pre-factor — 1
A1W Width dependence of A1 — 0
A2 Impact-ionization exponent at TR V 10
A2O Impact-ionization exponent at TR V 10
A3 Saturation-voltage dependence of impact-ionization — 1
A3L Length dependence of A3 — 0
Geometry independent saturation-voltage dependence
A3O — 1
of II
A3W Width dependence of A3 — 0
A4 Back-bias dependence of impact-ionization V−1/2 0
A4L Length dependence of A4 — 0
A4O Geometry independent back-bias dependence of II V−1/2 0
A4W Width dependence of A4 — 0
AGIDL GIDL pre-factor A/V3 0
AGIDLD GIDL pre-factor for drain side A/V3 0
AGIDLDW Width dependence of GIDL pre-factor for drain side A/V3 0
AGIDLW Width dependence of GIDL pre-factor A/V3 0
ALP CLM pre-factor — 0.01

527
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
ALP1 CLM enhancement factor above threshold V 0
Length dependence of CLM enhancement factor
ALP1L1 V 0
above threshold
ALP1L2 Second_order length dependence of ALP1 — 0
ALP1LEXP Exponent for length dependence of ALP1 — 0.5
ALP1W Width dependence of ALP1 — 0
ALP2 CLM enhancement factor below threshold V−1 0
Length dependence of CLM enhancement factor
ALP2L1 V−1 0
below threshold
ALP2L2 Second_order length dependence of ALP2 — 0
ALP2LEXP Exponent for length dependence of ALP2 — 0.5
ALP2W Width dependence of ALP2 — 0
ALPL Length dependence of ALP — 0.0005
ALPLEXP Exponent for length dependence of ALP — 1
ALPNOI Exponent for length offset for flicker noise — 2
ALPW Width dependence of ALP — 0
AX Linear/saturation transition factor — 3
AXL Length dependence of AX — 0.4
Geometry independent linear/saturation transition
AXO — 18
factor
BETEDGEW Width scaling coefficient of edge transistor mobility — 0
BETN Channel aspect ratio times zero-field mobility m2 /(Vs) 0.07
Channel aspect ratio times zero-field mobility of edge
BETNEDGE m2 /(Vs) 0.0005
transistor
BETW1 First higher-order width scaling coefficient of BETN — 0
Second higher-order width scaling coefficient of
BETW2 — 0
BETN
BGIDL GIDL probability factor at TR V 41
BGIDLD GIDL probability factor at TR for drain side V 41
BGIDLDO GIDL probability factor at TR for drain side V 41
BGIDLO GIDL probability factor at TR V 41
Band-to-band tunneling prefactor of bottom
CBBTBOT A/V3 1e-12
component for source-bulk junction
Band-to-band tunneling prefactor of bottom
CBBTBOTD A/V3 1e-12
component for drain-bulk junction
Band-to-band tunneling prefactor of gate-edge
CBBTGAT Am/V3 1e-18
component for source-bulk junction

528
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Band-to-band tunneling prefactor of gate-edge
CBBTGATD Am/V3 1e-18
component for drain-bulk junction
Band-to-band tunneling prefactor of STI-edge
CBBTSTI Am/V3 1e-18
component for source-bulk junction
Band-to-band tunneling prefactor of STI-edge
CBBTSTID Am/V3 1e-18
component for drain-bulk junction
CF DIBL-parameter — 0
CFB Back bias dependence of CF V−1 0
Bulk voltage dependence parameter of
CFBEDGE V−1 0
DIBL-parameter of edge transistors
Bulk voltage dependence parameter of
CFBEDGEO V−1 0
DIBL-parameter of edge transistors
CFBO Back-bias dependence of CF V−1 0
CFD Drain voltage dependence of CF V−1 0
Drain voltage dependence parameter of
CFDEDGE V−1 0
DIBL-parameter of edge transistors
Drain voltage dependence parameter of
CFDEDGEO V−1 0
DIBL-parameter of edge transistors
CFDO Drain voltage dependence of CF V−1 0
CFEDGE DIBL parameter of edge transistors — 0
Length dependence of DIBL-parameter of edge
CFEDGEL — 0
transistors
Exponent for length dependence of DIBL-parameter
CFEDGELEXP — 2
of edge transistors
Width dependence of DIBL-parameter of edge
CFEDGEW — 0
transistors
CFL Length dependence of DIBL-parameter — 0
CFLEXP Exponent for length dependence of CF — 2
CFR Outer fringe capacitance F 0
CFRD Outer fringe capacitance for drain side F 0
Outer fringe capacitance for 1 um wide channel for
CFRDW F 0
drain side
CFRW Outer fringe capacitance for 1 um wide channel F 0
CFW Width dependence of CF — 0
CGBOV Oxide capacitance for gate-bulk overlap F 0
Oxide capacitance for gate-bulk overlap for 1 um long
CGBOVL F 0
channel
CGIDL Back-bias dependence of GIDL — 0
CGIDLD Back-bias dependence of GIDL for drain side — 0

529
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
CGIDLDO Back-bias dependence of GIDL for drain side — 0
CGIDLO Back-bias dependence of GIDL — 0
CGOV Oxide capacitance for gate-drain/source overlap F 1e-15
CGOVD Oxide capacitance for gate-drain overlap F 1e-15
CHIB Tunnelling barrier height V 3.1
CHIBO Tunnelling barrier height V 3.1
Zero-bias capacitance per unit-of-area of bottom
CJORBOT F/m2 0.001
component for source-bulk junction
Zero-bias capacitance per unit-of-area of bottom
CJORBOTD F/m2 0.001
component for drain-bulk junction
Zero-bias capacitance per unit-of-length of gate-edge
CJORGAT F/m 1e-09
component for source-bulk junction
Zero-bias capacitance per unit-of-length of gate-edge
CJORGATD F/m 1e-09
component for drain-bulk junction
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTI F/m 1e-09
component for source-bulk junction
Zero-bias capacitance per unit-of-length of STI-edge
CJORSTID F/m 1e-09
component for drain-bulk junction
COX Oxide capacitance for intrinsic channel F 1e-14
CS Coulomb scattering parameter at TR — 0
CSL Length dependence of CS — 0
CSLEXP Exponent for length dependence of CS — 1
CSLW Area dependence of CS — 0
Geometry independent coulomb scattering parameter
CSO — 0
at TR
Shockley-Read-Hall prefactor of bottom component
CSRHBOT A/m3 100
for source-bulk junction
Shockley-Read-Hall prefactor of bottom component
CSRHBOTD A/m3 100
for drain-bulk junction
Shockley-Read-Hall prefactor of gate-edge
CSRHGAT A/m2 0.0001
component for source-bulk junction
Shockley-Read-Hall prefactor of gate-edge
CSRHGATD A/m2 0.0001
component for drain-bulk junction
Shockley-Read-Hall prefactor of STI-edge component
CSRHSTI A/m2 0.0001
for source-bulk junction
Shockley-Read-Hall prefactor of STI-edge component
CSRHSTID A/m2 0.0001
for drain-bulk junction
CSW Width dependence of CS — 0
CT Interface states factor — 0

530
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Trap-assisted tunneling prefactor of bottom
CTATBOT A/m3 100
component for source-bulk junction
Trap-assisted tunneling prefactor of bottom
CTATBOTD A/m3 100
component for drain-bulk junction
Trap-assisted tunneling prefactor of gate-edge
CTATGAT A/m2 0.0001
component for source-bulk junction
Trap-assisted tunneling prefactor of gate-edge
CTATGATD A/m2 0.0001
component for drain-bulk junction
Trap-assisted tunneling prefactor of STI-edge
CTATSTI A/m2 0.0001
component for source-bulk junction
Trap-assisted tunneling prefactor of STI-edge
CTATSTID A/m2 0.0001
component for drain-bulk junction
CTEDGE Interface states factor of edge transistors — 0
Length dependence of interface states factor of edge
CTEDGEL — 0
transistors
Exponent for length dependence of interface states
CTEDGELEXP — 1
factor of edge transistors
Geometry-independent interface states factor of edge
CTEDGEO — 0
transistors
CTH Thermal capacitance — 0
Length-correction to width dependence of thermal
CTHLW — 0
capacitance
CTHO Geometry independent part of thermal capacitance — 0
CTHW1 Width dependence of thermal capacitance — 0
CTHW2 Offset in width dependence of thermal capacitance — 0
CTL Length dependence of interface states factor — 0
Exponent for length dependence of interface states
CTLEXP — 1
factor
CTLW Area dependence of interface states factor — 0
CTO Geometry-independent interface states factor — 0
CTW Width dependence of interface states factor — 0
Offset parameter for PHIB in separate charge
DELVTAC V 0
calculation
DELVTACL Length dependence of DELVTAC V 0
Exponent for length dependence of offset of
DELVTACLEXP — 1
DELVTAC
DELVTACLW Area dependence of DELVTAC V 0
Geom. independent offset parameter for PHIB in
DELVTACO V 0
separate charge calculation
DELVTACW Width dependence of DELVTAC V 0

531
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
DLQ Effective channel length reduction for CV m 0
DLSIL Silicide extension over the physical gate length m 0
DNSUB Effective doping bias-dependence parameter V−1 0
DNSUBO Effective doping bias-dependence parameter V−1 0
DPHIB Offset parameter for PHIB V 0
DPHIBEDGE Offset parameter for PHIB of edge transistors V 0
DPHIBEDGEL Length dependence of edge transistor PHIB offset V 0
Exponent for length dependence of edge transistor
DPHIBEDGELEXP — 1
PHIB offset
DPHIBEDGELW Area dependence of edge transistor PHIB offset V 0
DPHIBEDGEO Geometry independent of edge transistor PHIB offset V 0
DPHIBEDGEW Width dependence of edge transistor PHIB offset V 0
DPHIBL Length dependence offset of PHIB V 0
DPHIBLEXP Exponent for length dependence of offset of PHIB — 1
DPHIBLW Area dependence of offset of PHIB V 0
DPHIBO Geometry independent offset of PHIB V 0
DPHIBW Width dependence of offset of PHIB V 0
DTA Temperature offset w.r.t. ambient temperature K 0
DVSBNUD Vsb-range for NUD-effect V 1
DVSBNUDO Vsb range for NUD-effect V 1
DWQ Effective channel width reduction for CV m 0
EF Flicker noise frequency exponent — 1
EFEDGE Flicker noise frequency exponent of edge transistors — 1
EFEDGEO Flicker noise frequency exponent — 1
EFO Flicker noise frequency exponent — 1
EPSROX Relative permittivity of gate dielectric — 3.9
EPSROXO Relative permittivity of gate dielectric — 3.9
Pre-factor for effective substrate doping in separate
FACNEFFAC — 1
charge calculation
FACNEFFACL Length dependence of FACNEFFAC — 0
FACNEFFACLW Area dependence of FACNEFFAC — 0
Geom. independent pre-factor for effective substrate
FACNEFFACO — 1
doping in separate charge calculation
FACNEFFACW Width dependence of FACNEFFAC — 0
Normalization field at the reference temperature for
FBBTRBOT band-to-band tunneling of bottom component for Vm−1 1e+09
source-bulk junction

532
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Normalization field at the reference temperature for
FBBTRBOTD band-to-band tunneling of bottom component for Vm−1 1e+09
drain-bulk junction
Normalization field at the reference temperature for
FBBTRGAT band-to-band tunneling of gate-edge component for Vm−1 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRGATD band-to-band tunneling of gate-edge component for Vm−1 1e+09
drain-bulk junction
Normalization field at the reference temperature for
FBBTRSTI band-to-band tunneling of STI-edge component for Vm−1 1e+09
source-bulk junction
Normalization field at the reference temperature for
FBBTRSTID band-to-band tunneling of STI-edge component for Vm−1 1e+09
drain-bulk junction
FBET1 Relative mobility decrease due to first lateral profile — 0
Width dependence of relative mobility decrease due
FBET1W — 0
to first lateral profile
Relative mobility decrease due to second lateral
FBET2 — 0
profile
FBETEDGE Length dependence of edge transistor mobility — 0
FETA Effective field parameter — 1
FETAO Effective field parameter — 1
Fraction below which source-bulk junction
FJUNQ — 0.03
capacitance components are considered negligible
Fraction below which drain-bulk junction capacitance
FJUNQD — 0.03
components are considered negligible
FNT Thermal noise coefficient — 1
FNTEDGE Thermal noise coefficient of edge transistors — 1
FNTEDGEO Thermal noise coefficient — 1
FNTEXC Excess noise coefficient — 0
FNTEXCL Length dependence coefficient of excess noise — 0
FNTO Thermal noise coefficient — 1
First length dependence coefficient for short channel
FOL1 — 0
body effect
Second length dependence coefficient for short
FOL2 — 0
channel body effect
FREV Coefficient for reverse breakdown current limitation — 1000
GC2 Gate current slope factor — 0.375
GC2O Gate current slope factor — 0.375

533
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
GC3 Gate current curvature factor — 0.063
GC3O Gate current curvature factor — 0.063
GCO Gate tunnelling energy adjustment — 0
GCOO Gate tunnelling energy adjustment — 0
GFACNUD Body-factor change due to NUD-effect — 1
GFACNUDL Length dependence of GFACNUD — 0
GFACNUDLEXP Exponent for length dependence of GFACNUD — 1
GFACNUDLW Area dependence of GFACNUD — 0
Geom. independent body-factor change due to
GFACNUDO — 1
NUD-effect
GFACNUDW Width dependence of GFACNUD — 0
Saturation current density at the reference temperature
IDSATRBOT A/m2 1e-12
of bottom component for source-bulk junction
Saturation current density at the reference temperature
IDSATRBOTD A/m2 1e-12
of bottom component for drain-bulk junction
Saturation current density at the reference temperature
IDSATRGAT A/m 1e-18
of gate-edge component for source-bulk junction
Saturation current density at the reference temperature
IDSATRGATD A/m 1e-18
of gate-edge component for drain-bulk junction
Saturation current density at the reference temperature
IDSATRSTI A/m 1e-18
of STI-edge component for source-bulk junction
Saturation current density at the reference temperature
IDSATRSTID A/m 1e-18
of STI-edge component for drain-bulk junction
IGINV Gate channel current pre-factor A 0
Gate channel current pre-factor for 1 um**2 channel
IGINVLW A 0
area
IGOV Gate overlap current pre-factor A 0
IGOVD Gate overlap current pre-factor for drain side A 0
Gate overlap current pre-factor for 1 um wide channel
IGOVDW A 0
for drain side
IGOVW Gate overlap current pre-factor for 1 um wide channel A 0
Maximum current up to which forward current
IMAX A 1000
behaves exponentially
KUO Mobility degradation/enhancement coefficient m 0
KUOWEL Length dependent mobility degradation factor — 0
KUOWELW Area dependent mobility degradation factor — 0
KUOWEO Geometrical independent mobility degradation factor — 0
KUOWEW Width dependent mobility degradation factor — 0

534
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Saturation velocity degradation/enhancement
KVSAT m 0
coefficient
KVTHO Threshold shift parameter Vm 0
KVTHOWEL Length dependent threshold shift parameter — 0
KVTHOWELW Area dependent threshold shift parameter — 0
KVTHOWEO Geometrical independent threshold shift parameter — 0
KVTHOWEW Width dependent threshold shift parameter — 0
LAP Effective channel length reduction per side m 0
LEVEL Model level — 103
LINTNOI Length offset for flicker noise m 0
LKUO Length dependence of KUO — 0
LKVTHO Length dependence of KVTHO — 0
LLODKUO Length parameter for UO stress effect — 0
LLODVTH Length parameter for VTH-stress effect — 0
LMAX Dummy parameter to label binning set m 1
LMIN Dummy parameter to label binning set m 0
LODETAO Eta0 shift modification factor for stress effect — 1
Overlap length for gate/drain and gate/source overlap
LOV m 0
capacitance
LOVD Overlap length for gate/drain overlap capacitance m 0
Mobility-related characteristic length of first lateral
LP1 m 1e-08
profile
Width dependence of mobility-related characteristic
LP1W — 0
length of first lateral profile
Mobility-related characteristic length of second lateral
LP2 m 1e-08
profile
LPCK Char. length of lateral doping profile m 1e-08
Width dependence of char. length of lateral doping
LPCKW — 0
profile
Exponent for length dependence of edge transistor
LPEDGE m 1e-08
mobility
LVARL Length dependence of LVAR — 0
Geom. independent difference between actual and
LVARO m 0
programmed gate length
LVARW Width dependence of LVAR — 0
Effective mass (in units of m0) for trap-assisted
MEFFTATBOT tunneling of bottom component for source-bulk — 0.25
junction

535
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Effective mass (in units of m0) for trap-assisted
MEFFTATBOTD tunneling of bottom component for drain-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATGAT tunneling of gate-edge component for source-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATGATD tunneling of gate-edge component for drain-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATSTI tunneling of STI-edge component for source-bulk — 0.25
junction
Effective mass (in units of m0) for trap-assisted
MEFFTATSTID tunneling of STI-edge component for drain-bulk — 0.25
junction
MUE Mobility reduction coefficient at TR m/V 0.5
Geometry independent mobility reduction coefficient
MUEO m/V 0.5
at TR
Width dependence of mobility reduction coefficient at
MUEW — 0
TR
NEFF Effective substrate doping m−3 5e+23
NEFFEDGE Effective substrate doping of edge transistors m−3 5e+23
NFA First coefficient of flicker noise — 8e+22
NFAEDGE First coefficient of flicker noise of edge transistors — 8e+22
First coefficient of flicker noise for 1 um**2 channel
NFAEDGELW — 8e+22
area
First coefficient of flicker noise for 1 um**2 channel
NFALW — 8e+22
area
NFB Second coefficient of flicker noise — 3e+07
NFBEDGE Second coefficient of flicker noise of edge transistors — 3e+07
Second coefficient of flicker noise for 1 um**2
NFBEDGELW — 3e+07
channel area
Second coefficient of flicker noise for 1 um**2
NFBLW — 3e+07
channel area
NFC Third coefficient of flicker noise V−1 0
NFCEDGE Third coefficient of flicker noise of edge transistors V−1 0
Third coefficient of flicker noise for 1 um**2 channel
NFCEDGELW V−1 0
area
Third coefficient of flicker noise for 1 um**2 channel
NFCLW V−1 0
area
NOV Effective doping of overlap region m−3 5e+25

536
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
NOVD Effective doping of overlap region for drain side m−3 5e+25
NOVDO Effective doping of overlap region for drain side m−3 5e+25
NOVO Effective doping of overlap region m−3 5e+25
NP Gate poly-silicon doping m−3 1e+26
NPCK Pocket doping level m−3 1e+24
Width dependence of pocket doping NPCK due to
NPCKW — 0
segregation
NPL Length dependence of gate poly-silicon doping — 0
NPO Geometry-independent gate poly-silicon doping m−3 1e+26
NSLP Effective doping bias-dependence parameter V 0.05
NSLPO Effective doping bias-dependence parameter V 0.05
NSUBEDGEL Length dependence of edge transistor substrate doping — 0
NSUBEDGELW Area dependence of edge transistor substrate doping — 0
Geometry independent substrate doping of edge
NSUBEDGEO m−3 5e+23
transistors
NSUBEDGEW Width dependence of edge transistor substrate doping — 0
NSUBO Geometry independent substrate doping m−3 3e+23
Width dependence of background doping NSUBO
NSUBW — 0
due to segregation
Grading coefficient of bottom component for
PBOT — 0.5
source-bulk junction
Grading coefficient of bottom component for
PBOTD — 0.5
drain-bulk junction
Breakdown onset tuning parameter of bottom
PBRBOT V 4
component for source-bulk junction
Breakdown onset tuning parameter of bottom
PBRBOTD V 4
component for drain-bulk junction
Breakdown onset tuning parameter of gate-edge
PBRGAT V 4
component for source-bulk junction
Breakdown onset tuning parameter of gate-edge
PBRGATD V 4
component for drain-bulk junction
Breakdown onset tuning parameter of STI-edge
PBRSTI V 4
component for source-bulk junction
Breakdown onset tuning parameter of STI-edge
PBRSTID V 4
component for drain-bulk junction
Grading coefficient of gate-edge component for
PGAT — 0.5
source-bulk junction
Grading coefficient of gate-edge component for
PGATD — 0.5
drain-bulk junction

537
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Zero-temperature bandgap voltage of bottom
PHIGBOT V 1.16
component for source-bulk junction
Zero-temperature bandgap voltage of bottom
PHIGBOTD V 1.16
component for drain-bulk junction
Zero-temperature bandgap voltage of gate-edge
PHIGGAT V 1.16
component for source-bulk junction
Zero-temperature bandgap voltage of gate-edge
PHIGGATD V 1.16
component for drain-bulk junction
Zero-temperature bandgap voltage of STI-edge
PHIGSTI V 1.16
component for source-bulk junction
Zero-temperature bandgap voltage of STI-edge
PHIGSTID V 1.16
component for drain-bulk junction
PKUO Cross-term dependence of KUO — 0
PKVTHO Cross-term dependence of KVTHO — 0
PLA1 Coefficient for the length dependence of A1 — 0
PLA3 Coefficient for the length dependence of A3 — 0
PLA4 Coefficient for the length dependence of A4 V−1/2 0
PLAGIDL Coefficient for the length dependence of AGIDL A/V3 0
Coefficient for the length dependence of AGIDL for
PLAGIDLD A/V3 0
drain side
PLALP Coefficient for the length dependence of ALP — 0
PLALP1 Coefficient for the length dependence of ALP1 V 0
PLALP2 Coefficient for the length dependence of ALP2 V−1 0
PLAX Coefficient for the length dependence of AX — 0
PLBETN Coefficient for the length dependence of BETN m2 /(Vs) 0
PLBETNEDGE Coefficient for the length dependence of BETNEDGE m2 /(Vs) 0
PLCF Coefficient for the length dependence of CF — 0
PLCFEDGE Coefficient for the length dependence of CFEDGE — 0
PLCFR Coefficient for the length dependence of CFR F 0
Coefficient for the length dependence of CFR for
PLCFRD F 0
drain side
PLCGBOV Coefficient for the length dependence of CGBOV F 0
PLCGOV Coefficient for the length dependence of CGOV F 0
Coefficient for the length dependence of CGOV for
PLCGOVD F 0
drain side
PLCOX Coefficient for the length dependence of COX F 0
PLCS Coefficient for the length dependence of CS — 0
PLCT Coefficient for the length dependence of CT — 0

538
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
PLCTEDGE Coefficient for the length dependence of CTEDGE — 0
PLDELVTAC Coefficient for the length dependence of DELVTAC V 0
PLDPHIB Coefficient for the length dependence of DPHIB V 0
Coefficient for the length dependence of
PLDPHIBEDGE V 0
DPHIBEDGE
PLFACNEFFAC Coefficient for the length dependence of FACNEFFAC — 0
PLFNTEXC Coefficient for the length dependence of FNTEXC — 0
PLGFACNUD Coefficient for the length dependence of GFACNUD — 0
PLIGINV Coefficient for the length dependence of IGINV A 0
PLIGOV Coefficient for the length dependence of IGOV A 0
Coefficient for the length dependence of IGOV for
PLIGOVD A 0
drain side
PLKUOWE Coefficient for the length dependence part of KUOWE — 0
Coefficient for the length dependence part of
PLKVTHOWE — 0
KVTHOWE
PLMUE Coefficient for the length dependence of MUE m/V 0
PLNEFF Coefficient for the length dependence of NEFF m−3 0
PLNEFFEDGE Coefficient for the length dependence of NEFFEDGE m−3 0
PLNFA Coefficient for the length dependence of NFA — 0
PLNFAEDGE Coefficient for the length dependence of NFAEDGE — 0
PLNFB Coefficient for the length dependence of NFB — 0
PLNFBEDGE Coefficient for the length dependence of NFBEDGE — 0
PLNFC Coefficient for the length dependence of NFC V−1 0
PLNFCEDGE Coefficient for the length dependence of NFCEDGE V−1 0
PLNOV Coefficient for the length dependence of NOV m−3 0
Coefficient for the length dependence of NOV for
PLNOVD m−3 0
drain side
PLNP Coefficient for the length dependence of NP m−3 0
PLPSCE Coefficient for the length dependence of PSCE — 0
PLPSCEEDGE Coefficient for the length dependence of PSCEEDGE — 0
PLRS Coefficient for the length dependence of RS ˙ 0
PLSTBET Coefficient for the length dependence of STBET — 0
Coefficient for the length dependence of
PLSTBETEDGE — 0
STBETEDGE
PLSTTHESAT Coefficient for the length dependence of STTHESAT — 0
PLSTVFB Coefficient for the length dependence of STVFB V/K 0
Coefficient for the length dependence of
PLSTVFBEDGE V/K 0
STVFBEDGE

539
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
PLTHESAT Coefficient for the length dependence of THESAT V−1 0
PLTHESATB Coefficient for the length dependence of THESATB V−1 0
PLTHESATG Coefficient for the length dependence of THESATG V−1 0
PLVFB Coefficient for the length dependence of VFB V 0
Coefficient for the length times width dependence of
PLWA1 — 0
A1
Coefficient for the length times width dependence of
PLWA3 — 0
A3
Coefficient for the length times width dependence of
PLWA4 V−1/2 0
A4
Coefficient for the length times width dependence of
PLWAGIDL A/V3 0
AGIDL
Coefficient for the length times width dependence of
PLWAGIDLD A/V3 0
AGIDL for drain side
Coefficient for the length times width dependence of
PLWALP — 0
ALP
Coefficient for the length times width dependence of
PLWALP1 V 0
ALP1
Coefficient for the length times width dependence of
PLWALP2 V−1 0
ALP2
Coefficient for the length times width dependence of
PLWAX — 0
AX
Coefficient for the length times width dependence of
PLWBETN m2 /(Vs) 0
BETN
Coefficient for the length times width dependence of
PLWBETNEDGE m2 /(Vs) 0
BETNEDGE
Coefficient for the length times width dependence of
PLWCF — 0
CF
Coefficient for the length times width dependence of
PLWCFEDGE — 0
CFEDGE
Coefficient for the length times width dependence of
PLWCFR F 0
CFR
Coefficient for the length times width dependence of
PLWCFRD F 0
CFR for drain side
Coefficient for the length times width dependence of
PLWCGBOV F 0
CGBOV
Coefficient for the length times width dependence of
PLWCGOV F 0
CGOV
Coefficient for the length times width dependence of
PLWCGOVD F 0
CGOV for drain side

540
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Coefficient for the length times width dependence of
PLWCOX F 0
COX
Coefficient for the length times width dependence of
PLWCS — 0
CS
Coefficient for the length times width dependence of
PLWCT — 0
CT
Coefficient for the length times width dependence of
PLWCTEDGE — 0
CTEDGE
Coefficient for the length times width dependence of
PLWDELVTAC V 0
DELVTAC
Coefficient for the length times width dependence of
PLWDPHIB V 0
DPHIB
Coefficient for the length times width dependence of
PLWDPHIBEDGE V 0
DPHIBEDGE
Coefficient for the length times width dependence of
PLWFACNEFFAC — 0
FACNEFFAC
Coefficient for the length times width dependence of
PLWFNTEXC — 0
FNTEXC
Coefficient for the length times width dependence of
PLWGFACNUD — 0
GFACNUD
Coefficient for the length times width dependence of
PLWIGINV A 0
IGINV
Coefficient for the length times width dependence of
PLWIGOV A 0
IGOV
Coefficient for the length times width dependence of
PLWIGOVD A 0
IGOV for drain side
Coefficient for the length times width dependence part
PLWKUOWE — 0
of KUOWE
Coefficient for the length times width dependence part
PLWKVTHOWE — 0
of KVTHOWE
Coefficient for the length times width dependence of
PLWMUE m/V 0
MUE
Coefficient for the length times width dependence of
PLWNEFF m−3 0
NEFF
Coefficient for the length times width dependence of
PLWNEFFEDGE m−3 0
NEFFEDGE
Coefficient for the length times width dependence of
PLWNFA — 0
NFA
Coefficient for the length times width dependence of
PLWNFAEDGE — 0
NFAEDGE
Coefficient for the length times width dependence of
PLWNFB — 0
NFB

541
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Coefficient for the length times width dependence of
PLWNFBEDGE — 0
NFBEDGE
Coefficient for the length times width dependence of
PLWNFC V−1 0
NFC
Coefficient for the length times width dependence of
PLWNFCEDGE V−1 0
NFCEDGE
Coefficient for the length times width dependence of
PLWNOV m−3 0
NOV
Coefficient for the length times width dependence of
PLWNOVD m−3 0
NOV for drain side
Coefficient for the length times width dependence of
PLWNP m−3 0
NP
Coefficient for the length times width dependence of
PLWPSCE — 0
PSCE
Coefficient for the length times width dependence of
PLWPSCEEDGE — 0
PSCEEDGE
Coefficient for the length times width dependence of
PLWRS ˙ 0
RS
Coefficient for the length times width dependence of
PLWSTBET — 0
STBET
Coefficient for the length times width dependence of
PLWSTBETEDGE — 0
STBETEDGE
Coefficient for the length times width dependence of
PLWSTTHESAT — 0
STTHESAT
Coefficient for the length times width dependence of
PLWSTVFB V/K 0
STVFB
Coefficient for the length times width dependence of
PLWSTVFBEDGE V/K 0
STVFBEDGE
Coefficient for the length times width dependence of
PLWTHESAT V−1 0
THESAT
Coefficient for the length times width dependence of
PLWTHESATB V−1 0
THESATB
Coefficient for the length times width dependence of
PLWTHESATG V−1 0
THESATG
Coefficient for the length times width dependence of
PLWVFB V 0
VFB
Coefficient for the length times width dependence of
PLWXCOR V−1 0
XCOR
PLXCOR Coefficient for the length dependence of XCOR V−1 0
POA1 Coefficient for the geometry independent part of A1 — 1
POA2 Coefficient for the geometry independent part of A2 V 10

542
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
POA3 Coefficient for the geometry independent part of A3 — 1
POA4 Coefficient for the geometry independent part of A4 V−1/2 0
Coefficient for the geometry independent part of
POAGIDL A/V3 0
AGIDL
Coefficient for the geometry independent part of
POAGIDLD A/V3 0
AGIDL for drain side
POALP Coefficient for the geometry independent part of ALP — 0.01
Coefficient for the geometry independent part of
POALP1 V 0
ALP1
Coefficient for the geometry independent part of
POALP2 V−1 0
ALP2
POAX Coefficient for the geometry independent part of AX — 3
Coefficient for the geometry independent part of
POBETN m2 /(Vs) 0.07
BETN
Coefficient for the geometry independent part of
POBETNEDGE m2 /(Vs) 0.0005
BETNEDGE
Coefficient for the geometry independent part of
POBGIDL V 41
BGIDL
Coefficient for the geometry independent part of
POBGIDLD V 41
BGIDL for drain side
POCF Coefficient for the geometry independent part of CF — 0
POCFB Coefficient for the geometry independent part of CFB V−1 0
Coefficient for the geometry independent part of
POCFBEDGE V−1 0
CFBEDGE
POCFD Coefficient for the geometry independent part of CFD V−1 0
Coefficient for the geometry independent part of
POCFDEDGE V−1 0
CFDEDGE
Coefficient for the geometry independent part of
POCFEDGE — 0
CFEDGE
POCFR Coefficient for the geometry independent part of CFR F 0
Coefficient for the geometry independent part of CFR
POCFRD F 0
for drain side
Coefficient for the geometry independent part of
POCGBOV F 0
CGBOV
Coefficient for the geometry independent part of
POCGIDL — 0
CGIDL
Coefficient for the geometry independent part of
POCGIDLD — 0
CGIDL for drain side
Coefficient for the geometry independent part of
POCGOV F 1e-15
CGOV

543
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Coefficient for the geometry independent part of
POCGOVD F 1e-15
CGOV for drain side
Coefficient for the geometry independent part of
POCHIB V 3.1
CHIB
POCOX Coefficient for the geometry independent part of COX F 1e-14
POCS Coefficient for the geometry independent part of CS — 0
POCT Coefficient for the geometry independent part of CT — 0
Coefficient for the geometry independent part of
POCTEDGE — 0
CTEDGE
Coefficient for the geometry independent part of
PODELVTAC V 0
DELVTAC
Coefficient for the geometry independent part of
PODNSUB V−1 0
DNSUB
Coefficient for the geometry independent part of
PODPHIB V 0
DPHIB
Coefficient for the geometry independent part of
PODPHIBEDGE V 0
DPHIBEDGE
Coefficient for the geometry independent part of
PODVSBNUD V 1
DVSBNUD
POEF Coefficient for the flicker noise frequency exponent — 1
Coefficient for the geometry independent part of
POEFEDGE — 1
EFEDGE
Coefficient for the geometry independent part of
POEPSROX — 3.9
EPSOX
Coefficient for the geometry independent part of
POFACNEFFAC — 1
FACNEFFAC
Coefficient for the geometry independent part of
POFETA — 1
FETA
POFNT Coefficient for the geometry independent part of FNT — 1
Coefficient for the geometry independent part of
POFNTEDGE — 1
FNTEDGE
Coefficient for the geometry independent part of
POFNTEXC — 0
FNTEXC
POGC2 Coefficient for the geometry independent part of GC2 — 0.375
POGC3 Coefficient for the geometry independent part of GC3 — 0.063
POGCO Coefficient for the geometry independent part of GCO — 0
Coefficient for the geometry independent part of
POGFACNUD — 1
GFACNUD
Coefficient for the geometry independent part of
POIGINV A 0
IGINV

544
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Coefficient for the geometry independent part of
POIGOV A 0
IGOV
Coefficient for the geometry independent part of
POIGOVD A 0
IGOV for drain side
Coefficient for the geometry independent part of
POKUOWE — 0
KUOWE
Coefficient for the geometry independent part of
POKVTHOWE — 0
KVTHOWE
POMUE Coefficient for the geometry independent part of MUE m/V 0.5
Coefficient for the geometry independent part of
PONEFF m−3 5e+23
NEFF
Coefficient for the geometry independent part of
PONEFFEDGE m−3 5e+23
NEFFEDGE
PONFA Coefficient for the geometry independent part of NFA — 8e+22
Coefficient for the geometry independent part of
PONFAEDGE — 8e+22
NFAEDGE
PONFB Coefficient for the geometry independent part of NFB — 3e+07
Coefficient for the geometry independent part of
PONFBEDGE — 3e+07
NFBEDGE
PONFC Coefficient for the geometry independent part of NFC V−1 0
Coefficient for the geometry independent part of
PONFCEDGE V−1 0
NFCEDGE
PONOV Coefficient for the geometry independent part of NOV m−3 5e+25
Coefficient for the geometry independent part of NOV
PONOVD m−3 5e+25
for drain side
PONP Coefficient for the geometry independent part of NP m−3 1e+26
Coefficient for the geometry independent part of
PONSLP V 0.05
NSLP
Coefficient for the geometry independent part of
POPSCE — 0
PSCE
Coefficient for the geometry independent part of
POPSCEB V−1 0
PSCEB
Coefficient for the geometry independent part of
POPSCEBEDGE V−1 0
PSCEBEDGE
Coefficient for the geometry independent part of
POPSCED V−1 0
PSCED
Coefficient for the geometry independent part of
POPSCEDEDGE V−1 0
PSCEDEDGE
Coefficient for the geometry independent part of
POPSCEEDGE — 0
PSCEEDGE
PORS Coefficient for the geometry independent part of RS ˙ 30

545
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
PORSB Coefficient for the geometry independent part of RSB V−1 0
PORSG Coefficient for the geometry independent part of RSG V−1 0
POSTA2 Coefficient for the geometry independent part of STA2 V 0
Coefficient for the geometry independent part of
POSTBET — 1
STBET
Coefficient for the geometry independent part of
POSTBETEDGE — 1
STBETEDGE
Coefficient for the geometry independent part of
POSTBGIDL V/K 0
STBGIDL
Coefficient for the geometry independent part of
POSTBGIDLD V/K 0
STBGIDL for drain side
POSTCS Coefficient for the geometry independent part of STCS — 0
POSTIG Coefficient for the geometry independent part of STIG — 2
Coefficient for the geometry independent part of
POSTMUE — 0
STMUE
POSTRS Coefficient for the geometry independent part of STRS — 1
Coefficient for the geometry independent part of
POSTTHEMU — 1.5
STTHEMU
Coefficient for the geometry independent part of
POSTTHESAT — 1
STTHESAT
Coefficient for the geometry independent part of
POSTVFB V/K 0.0005
STVFB
Coefficient for the geometry independent part of
POSTVFBEDGE V/K 0
STVFBEDGE
Coefficient for the geometry independent part of
POSTXCOR — 0
STXCOR
Coefficient for the geometry independent part of
POTHEMU — 1.5
THEMU
Coefficient for the geometry independent part of
POTHESAT V−1 1
THESAT
Coefficient for the geometry independent part of
POTHESATB V−1 0
THESATB
Coefficient for the geometry independent part of
POTHESATG V−1 0
THESATG
POTOX Coefficient for the geometry independent part of TOX m 2e-09
Coefficient for the geometry independent part of
POTOXOV m 2e-09
TOXOV
Coefficient for the geometry independent part of
POTOXOVD m 2e-09
TOXOV for drain side
POVFB Coefficient for the geometry independent part of VFB V -1

546
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Coefficient for the geometry independent part of
POVFBEDGE V -1
VFBEDGE
Coefficient for the geometry independent part of
POVNSUB V 0
VNSUB
POVP Coefficient for the geometry independent part of VP V 0.05
Coefficient for the geometry independent part of
POVSBNUD V 0
VSBNUD
Coefficient for the geometry independent part of
POXCOR V−1 0
XCOR
Subthreshold slope coefficient for short channel
PSCE — 0
transistor
Bulk voltage dependence parameter of subthreshold
PSCEB V−1 0
slope coefficient for short channel transistor
Bulk voltage dependence parameter of subthreshold
PSCEBEDGE V−1 0
slope coefficient for short channel edge transistors
Bulk voltage dependence parameter of subthreshold
PSCEBEDGEO V−1 0
slope coefficient for short channel edge transistors
Bulk voltage dependence parameter of subthreshold
PSCEBO V−1 0
slope coefficient for short channel transistor
Drain voltage dependence parameter of subthreshold
PSCED V−1 0
slope coefficient for short channel transistor
Drain voltage dependence parameter of subthreshold
PSCEDEDGE V−1 0
slope coefficient for short channel edge transistors
Drain voltage dependence parameter of subthreshold
PSCEDEDGEO V−1 0
slope coefficient for short channel edge transistors
Drain voltage dependence parameter of subthreshold
PSCEDO V−1 0
slope coefficient for short channel transistor
Subthreshold slope coefficient for short channel edge
PSCEEDGE — 0
transistors
Length dependence of subthreshold slope coefficient
PSCEEDGEL — 0
for short channel edge transistors
Exponent for length dependence of subthreshold slope
PSCEEDGELEXP — 2
coefficient for short channel edge transistors
Exponent for length dependence of subthreshold slope
PSCEEDGEW — 0
coefficient for short channel edge transistor
Length dependence of subthreshold slope coefficient
PSCEL — 0
for short channel transistor
Exponent for length dependence of subthreshold slope
PSCELEXP — 2
coefficient for short channel transistor
Exponent for length dependence of subthreshold slope
PSCEW — 0
coefficient for short channel transistor

547
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Grading coefficient of STI-edge component for
PSTI — 0.5
source-bulk junction
Grading coefficient of STI-edge component for
PSTID — 0.5
drain-bulk junction
PWA1 Coefficient for the width dependence of A1 — 0
PWA3 Coefficient for the width dependence of A3 — 0
PWA4 Coefficient for the width dependence of A4 V−1/2 0
PWAGIDL Coefficient for the width dependence of AGIDL A/V3 0
Coefficient for the width dependence of AGIDL for
PWAGIDLD A/V3 0
drain side
PWALP Coefficient for the width dependence of ALP — 0
PWALP1 Coefficient for the width dependence of ALP1 V 0
PWALP2 Coefficient for the width dependence of ALP2 V−1 0
PWAX Coefficient for the width dependence of AX — 0
PWBETN Coefficient for the width dependence of BETN m2 /(Vs) 0
PWBETNEDGE Coefficient for the width dependence of BETNEDGE m2 /(Vs) 0
PWCF Coefficient for the width dependence of CF — 0
PWCFEDGE Coefficient for the width dependence of CFEDGE — 0
PWCFR Coefficient for the width dependence of CFR F 0
Coefficient for the width dependence of CFR for drain
PWCFRD F 0
side
PWCGBOV Coefficient for the width dependence of CGBOV F 0
PWCGOV Coefficient for the width dependence of CGOV F 0
Coefficient for the width dependence of CGOV for
PWCGOVD F 0
drain side
PWCOX Coefficient for the width dependence of COX F 0
PWCS Coefficient for the width dependence of CS — 0
PWCT Coefficient for the width dependence of CT — 0
PWCTEDGE Coefficient for the width dependence of CTEDGE — 0
PWDELVTAC Coefficient for the width dependence of DELVTAC V 0
PWDPHIB Coefficient for the width dependence of DPHIB V 0
PWDPHIBEDGE Coefficient for the width dependence of DPHIBEDGE V 0
PWFACNEFFAC Coefficient for the width dependence of FACNEFFAC — 0
PWFNTEXC Coefficient for the width dependence of FNTEXC — 0
PWGFACNUD Coefficient for the width dependence of GFACNUD — 0
PWIGINV Coefficient for the width dependence of IGINV A 0
PWIGOV Coefficient for the width dependence of IGOV A 0

548
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Coefficient for the width dependence of IGOV for
PWIGOVD A 0
drain side
PWKUOWE Coefficient for the width dependence part of KUOWE — 0
Coefficient for the width dependence part of
PWKVTHOWE — 0
KVTHOWE
PWMUE Coefficient for the width dependence of MUE m/V 0
PWNEFF Coefficient for the width dependence of NEFF m−3 0
PWNEFFEDGE Coefficient for the width dependence of NEFFEDGE m−3 0
PWNFA Coefficient for the width dependence of NFA — 0
PWNFAEDGE Coefficient for the width dependence of NFAEDGE — 0
PWNFB Coefficient for the width dependence of NFB — 0
PWNFBEDGE Coefficient for the width dependence of NFBEDGE — 0
PWNFC Coefficient for the width dependence of NFC V−1 0
PWNFCEDGE Coefficient for the width dependence of NFCEDGE V−1 0
PWNOV Coefficient for the width dependence of NOV m−3 0
Coefficient for the width dependence of NOV for drain
PWNOVD m−3 0
side
PWNP Coefficient for the width dependence of NP m−3 0
PWPSCE Coefficient for the width dependence of PSCE — 0
PWPSCEEDGE Coefficient for the width dependence of PSCEEDGE — 0
PWRS Coefficient for the width dependence of RS ˙ 0
PWSTBET Coefficient for the width dependence of STBET — 0
PWSTBETEDGE Coefficient for the width dependence of STBETEDGE — 0
PWSTTHESAT Coefficient for the width dependence of STTHESAT — 0
PWSTVFB Coefficient for the width dependence of STVFB V/K 0
PWSTVFBEDGE Coefficient for the width dependence of STVFBEDGE V/K 0
PWTHESAT Coefficient for the width dependence of THESAT V−1 0
PWTHESATB Coefficient for the width dependence of THESATB V−1 0
PWTHESATG Coefficient for the width dependence of THESATG V−1 0
PWVFB Coefficient for the width dependence of VFB V 0
PWXCOR Coefficient for the width dependence of XCOR V−1 0
QMC Quantum-mechanical correction factor — 1
RBULK Bulk resistance between node BP and BI ˙ 0
RBULKO Bulk resistance between node BP and BI ˙ 0
RDE External drain resistance ˙ 0
RG Gate resistance ˙ 0
RGO Gate resistance ˙ 0

549
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
RINT Contact resistance between silicide and ploy ˙ m2 0
RJUND Drain-side bulk resistance between node BI and BD ˙ 0
RJUNDO Drain-side bulk resistance between node BI and BD ˙ 0
RJUNS Source-side bulk resistance between node BI and BS ˙ 0
RJUNSO Source-side bulk resistance between node BI and BS ˙ 0
RS Series resistance at TR ˙ 30
RSB Back-bias dependence of series resistance V−1 0
RSBO Back-bias dependence of series resistance V−1 0
RSE External source resistance ˙ 0
RSG Gate-bias dependence of series resistance V−1 0
RSGO Gate-bias dependence of series resistance V−1 0
RSH Sheet resistance of source diffusion ˙/□ 0
RSHD Sheet resistance of drain diffusion ˙/□ 0
RSHG Gate electrode diffusion sheet resistance ˙/□ 0
Source/drain series resistance for 1 um wide channel
RSW1 ˙ 50
at TR
RSW2 Higher-order width scaling of RS — 0
RTH Thermal resistance — 0
Length-correction to width dependence of thermal
RTHLW — 0
resistance
RTHO Geometry independent part of thermal resistance — 0
RTHW1 Width dependence of thermal resistance — 0
RTHW2 Offset in width dependence of thermal resistance — 0
RVPOLY Vertical poly resistance ˙ m2 0
RWELL Well resistance between node BI and B ˙ 0
RWELLO Well resistance between node BI and B ˙ 0
Reference distance between OD-edge and poly from
SAREF m 1e-06
one side
Reference distance between OD-edge and poly from
SBREF m 1e-06
other side
Distance between OD-edge and well edge of a
SCREF m 1e-06
reference device
STA2 Temperature dependence of A2 V 0
STA2O Temperature dependence of A2 V 0
STBET Temperature dependence of BETN — 1
STBETEDGE Temperature dependence of BETNEDGE — 1
Length dependence of temperature dependence of
STBETEDGEL — 0
BETNEDGE

550
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Area dependence of temperature dependence of
STBETEDGELW — 0
BETNEDGE
Geometry independent temperature dependence of
STBETEDGEO — 1
BETNEDGE
Width dependence of temperature dependence of
STBETEDGEW — 0
BETNEDGE
Length dependence of temperature dependence of
STBETL — 0
BETN
Area dependence of temperature dependence of
STBETLW — 0
BETN
Geometry independent temperature dependence of
STBETO — 1
BETN
Width dependence of temperature dependence of
STBETW — 0
BETN
STBGIDL Temperature dependence of BGIDL V/K 0
STBGIDLD Temperature dependence of BGIDL for drain side V/K 0
STBGIDLDO Temperature dependence of BGIDL for drain side V/K 0
STBGIDLO Temperature dependence of BGIDL V/K 0
STCS Temperature dependence of CS — 0
STCSO Temperature dependence of CS — 0
STETAO Eta0 shift factor related to VTHO change m 0
Temperature scaling parameter for band-to-band
STFBBTBOT tunneling of bottom component for source-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTBOTD tunneling of bottom component for drain-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTGAT tunneling of gate-edge component for source-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTGATD tunneling of gate-edge component for drain-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTSTI tunneling of STI-edge component for source-bulk 1/K -0.001
junction
Temperature scaling parameter for band-to-band
STFBBTSTID tunneling of STI-edge component for drain-bulk 1/K -0.001
junction
STIG Temperature dependence of IGINV and IGOV — 2
STIGO Temperature dependence of IGINV and IGOV — 2

551
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
STMUE Temperature dependence of MUE — 0
STMUEO Temperature dependence of MUE — 0
STRS Temperature dependence of RS — 1
STRSO Temperature dependence of RS — 1
STRTH Temperature sensitivity of RTH — 0
STRTHO Temperature sensitivity of RTH — 0
STTHEMU Temperature dependence of THEMU — 1.5
STTHEMUO Temperature dependence of THEMU — 1.5
STTHESAT Temperature dependence of THESAT — 1
Length dependence of temperature dependence of
STTHESATL — 0
THESAT
Area dependence of temperature dependence of
STTHESATLW — 0
THESAT
Geometry independent temperature dependence of
STTHESATO — 1
THESAT
Width dependence of temperature dependence of
STTHESATW — 0
THESAT
STVFB Temperature dependence of VFB V/K 0.0005
STVFBEDGE Temperature dependence of VFBEDGE V/K 0.0005
Length dependence of temperature dependence of
STVFBEDGEL V/K 0
VFBEDGE
Area dependence of temperature dependence of
STVFBEDGELW V/K 0
VFBEDGE
Geometry-independent temperature dependence of
STVFBEDGEO V/K 0.0005
VFBEDGE
Width dependence of temperature dependence of
STVFBEDGEW V/K 0
VFBEDGE
Length dependence of temperature dependence of
STVFBL V/K 0
VFB
STVFBLW Area dependence of temperature dependence of VFB V/K 0
Geometry-independent temperature dependence of
STVFBO V/K 0.0005
VFB
STVFBW Width dependence of temperature dependence of VFB V/K 0
STXCOR Temperature dependence of XCOR — 0
STXCORO Temperature dependence of XCOR — 0
SWDELVTAC Flag for separate capacitance calculation; 0=off, 1=on — 0
SWEDGE Flag for drain current of edge transistors; 0=off, 1=on — 0
Flag for geometrical model, 0=local, 1=global,
SWGEO — 1
2=binning

552
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
SWGIDL Flag for GIDL current, 0=turn off IGIDL — 0
SWIGATE Flag for gate current, 0=turn off IG — 0
SWIGN Flag for induced gate noise; 0=off, 1=on — 1
SWIMPACT Flag for impact ionization current, 0=turn off II — 0
Flag for asymmetric junctions; 0=symmetric,
SWJUNASYM — 0
1=asymmetric
SWJUNCAP Flag for juncap, 0=turn off juncap — 0
Flag for JUNCAP-express; 0=full model, 1=express
SWJUNEXP — 0
model
Flag for NUD-effect; 0=off, 1=on,
SWNUD — 0
2=on+CV-correction
THEMU Mobility reduction exponent at TR — 1.5
THEMUO Mobility reduction exponent at TR — 1.5
THESAT Velocity saturation parameter at TR V−1 1
THESATB Back-bias dependence of velocity saturation V−1 0
THESATBO Back-bias dependence of velocity saturation V−1 0
THESATG Gate-bias dependence of velocity saturation V−1 0
THESATGO Gate-bias dependence of velocity saturation V−1 0
THESATL Length dependence of THESAT V−1 0.05
THESATLEXP Exponent for length dependence of THESAT — 1
THESATLW Area dependence of velocity saturation parameter — 0
Geometry independent velocity saturation parameter
THESATO V−1 0
at TR
THESATW Width dependence of velocity saturation parameter — 0
TKUO Temperature dependence of KUO — 0
TOX Gate oxide thickness m 2e-09
TOXO Gate oxide thickness m 2e-09
TOXOV Overlap oxide thickness m 2e-09
TOXOVD Overlap oxide thickness for drain side m 2e-09
TOXOVDO Overlap oxide thickness for drain side m 2e-09
TOXOVO Overlap oxide thickness m 2e-09
TR nominal (reference) temperature ◦C 21
TRJ Reference temperature ◦C 21
TYPE Channel type parameter, +1=NMOS -1=PMOS — 1
UO Zero-field mobility at TR m2 /(Vs) 0.05
Built-in voltage at the reference temperature of
VBIRBOT V 1
bottom component for source-bulk junction

553
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
Built-in voltage at the reference temperature of
VBIRBOTD V 1
bottom component for drain-bulk junction
Built-in voltage at the reference temperature of
VBIRGAT V 1
gate-edge component for source-bulk junction
Built-in voltage at the reference temperature of
VBIRGATD V 1
gate-edge component for drain-bulk junction
Built-in voltage at the reference temperature of
VBIRSTI V 1
STI-edge component for source-bulk junction
Built-in voltage at the reference temperature of
VBIRSTID V 1
STI-edge component for drain-bulk junction
Breakdown voltage of bottom component for
VBRBOT V 10
source-bulk junction
Breakdown voltage of bottom component for
VBRBOTD V 10
drain-bulk junction
Breakdown voltage of gate-edge component for
VBRGAT V 10
source-bulk junction
Breakdown voltage of gate-edge component for
VBRGATD V 10
drain-bulk junction
Breakdown voltage of STI-edge component for
VBRSTI V 10
source-bulk junction
Breakdown voltage of STI-edge component for
VBRSTID V 10
drain-bulk junction
VFB Flat band voltage at TR V -1
VFBEDGE Flat band voltage of edge transistors at TR V -1
Geometry-independent flat-band voltage of edge
VFBEDGEO V -1
transistors at TR
VFBL Length dependence of flat-band voltage V 0
VFBLW Area dependence of flat-band voltage V 0
VFBO Geometry-independent flat-band voltage at TR V -1
VFBW Width dependence of flat-band voltage V 0
Typical maximum source-bulk junction voltage;
VJUNREF V 2.5
usually about 2*VSUP
Typical maximum drain-bulk junction voltage; usually
VJUNREFD V 2.5
about 2*VSUP
VNSUB Effective doping bias-dependence parameter V 0
VNSUBO Effective doping bias-dependence parameter V 0
VP CLM logarithm dependence factor V 0.05
VPO CLM logarithmic dependence parameter V 0.05
VSBNUD Lower Vsb value for NUD-effect V 0
VSBNUDO Lower Vsb value for NUD-effect V 0

554
 Table 2-123. PSP103VA MOSFET with self-heating Device Model Parameters
Parameter Description Units Default
WBET Characteristic width for width scaling of BETN m 1e-09
WEB Coefficient for SCB — 0
WEC Coefficient for SCC — 0
WEDGE Electrical width of edge transistor per side m 1e-08
WEDGEW Width dependence of edge WEDGE — 0
WKUO Width dependence of KUO — 0
WKVTHO Width dependence of KVTHO — 0
WLOD Width parameter m 0
WLODKUO Width parameter for UO stress effect — 0
WLODVTH Width parameter for VTH-stress effect — 0
WMAX Dummy parameter to label binning set m 1
WMIN Dummy parameter to label binning set m 0
WOT Effective channel width reduction per side m 0
Char. length of segregation of background doping
WSEG m 1e-08
NSUBO
WSEGP Char. length of segregation of pocket doping NPCK m 1e-08
WVARL Length dependence of WVAR — 0
Geom. independent difference between actual and
WVARO m 0
programmed field-oxide opening
WVARW Width dependence of WVAR — 0
XCOR Non-universality factor V−1 0
XCORL Length dependence of non-universality parameter — 0
XCORLW Area dependence of non-universality parameter — 0
XCORO Geometry independent non-universality parameter V−1 0
XCORW Width dependence of non-universality parameter — 0
Junction depth of gate-edge component for
XJUNGAT m 1e-07
source-bulk junction
Junction depth of gate-edge component for drain-bulk
XJUNGATD m 1e-07
junction
Junction depth of STI-edge component for
XJUNSTI m 1e-07
source-bulk junction
Junction depth of STI-edge component for drain-bulk
XJUNSTID m 1e-07
junction

555
 Table 2-124. MOSFET level 1031 Output Variables
Parameter Description Units Default
ctype Flag for channel type none
sdint Flag for source-drain interchange none
ise Total source current A none
ige Total gate current A none
ide Total drain current A none
ibe Total bulk current A none
Drain current, excl. edge transistor currents,
ids A none
avalanche, tunnel, GISL, GIDL, and junction currents
idb Drain to bulk current A none
isb Source to bulk current A none
igs Gate-source tunneling current A none
igd Gate-drain tunneling current A none
igb Gate-bulk tunneling current A none
idedge Drain current of edge transistors A none
igcs Gate-channel tunneling current (source component) A none
igcd Gate-channel tunneling current (drain component) A none
iavl Substrate current due to weak avelanche A none
igisl Gate-induced source leakage current A none
igidl Gate-induced drain leakage current A none
ijs Total source junction current A none
ijsbot Source junction current (bottom component) A none
ijsgat Source junction current (gate-edge component) A none
ijssti Source junction current (STI-edge component) A none
ijd Total drain junction current A none
ijdbot Drain junction current (bottom component) A none
ijdgat Drain junction current (gate-edge component) A none
ijdsti Drain junction current (STI-edge component) A none
vds Drain-source voltage V none
vgs Gate-source voltage V none
vsb Source-bulk voltage V none
vto Zero-bias threshold voltage V none
vts Threshold voltage including back bias effects V none
Threshold voltage including back bias and drain bias
vth V none
effects
Effective gate drive voltage including back bias and
vgt V none
drain bias effects
vdss Drain saturation voltage at actual bias V none

556
 Table 2-124. MOSFET level 1031 Output Variables
Parameter Description Units Default
vsat Saturation limit none
gm Transconductance ˙ −1 none
gmb Substrate transconductance ˙ −1 none
gds Output conductance ˙ −1 none
gjs Source junction conductance ˙ −1 none
gjd Drain junction conductance ˙ −1 none
cdd Drain capacitance F none
cdg Drain-gate capacitance F none
cds Drain-source capacitance F none
cdb Drain-bulk capacitance F none
cgd Gate-drain capacitance F none
cgg Gate capacitance F none
cgs Gate-source capacitance F none
cgb Gate-bulk capacitance F none
csd Source-drain capacitance F none
csg Source-gate capacitance F none
css Source capacitance F none
csb Source-bulk capacitance F none
cbd Bulk-drain capacitance F none
cbg Bulk-gate capacitance F none
cbs Bulk-source capacitance F none
cbb Bulk capacitance F none
cgsol Total gate-source overlap capacitance F none
cgdol Total gate-drain overlap capacitance F none
cjs Total source junction capacitance F none
cjsbot Source junction capacitance (bottom component) F none
cjsgat Source junction capacitance (gate-edge component) F none
cjssti Source junction capacitance (STI-edge component) F none
cjd Total drain junction capacitance F none
cjdbot Drain junction capacitance (bottom component) F none
cjdgat Drain junction capacitance (gate-edge component) F none
cjdsti Drain junction capacitance (STI-edge component) F none
weff Effective channel width for geometrical models m none
leff Effective channel length for geometrical models m none
u Transistor gain none
rout Small-signal output resistance Ohm none

557
 Table 2-124. MOSFET level 1031 Output Variables
Parameter Description Units Default
vearly Equivalent Early voltage V none
beff Gain factor A/V2 none
fug Unity gain frequency at actual bias Hz none
rg Gate resistance Ohm none
sfl Flicker noise current spectral density at 1 Hz A2 /Hz none
Input-referred RMS white noise voltage spectral
sqrtsff V/sqrt(Hz) none
density at 1 kHz
Input-referred RMS white noise voltage spectral
sqrtsfw V/sqrt(Hz) none
density
sid White noise current spectral density A2 /Hz none
sig Induced gate noise current spectral density at 1 Hz A2 /Hz none
Imaginary part of correlation coefficient between Sig
cigid none
and Sid
Cross-over frequency above which white noise is
fknee Hz none
dominant
sigs Gate-source current noise spectral density A2 /Hz none
sigd Gate-drain current noise spectral density A2 /Hz none
siavl Impact ionization current noise spectral density A2 /Hz none
ssi Total source junction current noise spectral density A2 /Hz none
sdi Total drain junction current noise spectral density A2 /Hz none
Flicker noise current spectral density at 1 Hz of edge
sfledge A2 /Hz none
transistors
White noise current spectral density of edge
sidedge A2 /Hz none
transistors
lp_vfb Local parameter VFB after T-scaling and clipping V none
lp_stvfb Local parameter STVFB after clipping V/K none
lp_tox Local parameter TOX after clipping m none
lp_epsrox Local parameter EPSROX after clipping none
lp_neff Local parameter NEFF after clipping m−3 none
lp_facneffac Local parameter FACNEFFAC after clipping none
lp_gfacnud Local parameter GFACNUD after clipping none
lp_vsbnud Local parameter VSBNUD after clipping V none
lp_dvsbnud Local parameter DVSBNUD after clipping V none
lp_vnsub Local parameter VNSUB after clipping V none
lp_nslp Local parameter NSLP after clipping V none
lp_dnsub Local parameter DNSUB after clipping V−1 none
lp_dphib Local parameter DPHIB after clipping V none
lp_delvtac Local parameter DELVTAC after clipping V none

558
 Table 2-124. MOSFET level 1031 Output Variables
Parameter Description Units Default
lp_np Local parameter NP after clipping m−3 none
lp_ct Local parameter CT after clipping none
lp_toxov Local parameter TOXOV after clipping m none
lp_toxovd Local parameter TOXOVD after clipping m none
lp_nov Local parameter NOV after clipping m−3 none
lp_novd Local parameter NOVD after clipping m−3 none
lp_cf Local parameter CF after clipping none
lp_cfd Local parameter CFD after clipping V−1 none
lp_cfb Local parameter CFB after clipping V−1 none
lp_psce Local parameter PSCE after clipping none
lp_psceb Local parameter PSCEB after clipping V−1 none
lp_psced Local parameter PSCED after clipping V−1 none
lp_betn Local parameter BETN after T-scaling and clipping m2 /(V s) none
lp_stbet Local parameter STBET after clipping none
lp_mue Local parameter MUE after T-scaling and clipping m/V none
lp_stmue Local parameter STMUE after clipping none
lp_themu Local parameter THEMU after T-scaling and clipping none
lp_stthemu Local parameter STTHEMU after clipping none
lp_cs Local parameter CS after T-scaling and clipping none
lp_stcs Local parameter STCS after clipping none
lp_xcor Local parameter XCOR after T-scaling and clipping V−1 none
lp_stxcor Local parameter STXCOR after clipping none
lp_feta Local parameter FETA after clipping none
lp_rs Local parameter RS after T-scaling and clipping Ohm none
lp_strs Local parameter STRS after clipping none
lp_rsb Local parameter RSB after clipping V−1 none
lp_rsg Local parameter RSG after clipping V−1 none
lp_thesat Local parameter THESAT after T-scaling and clipping V−1 none
lp_stthesat Local parameter STTHESAT after clipping none
lp_thesatb Local parameter THESATB after clipping V−1 none
lp_thesatg Local parameter THESATG after clipping V−1 none
lp_ax Local parameter AX after clipping none
lp_alp Local parameter ALP after clipping none
lp_alp1 Local parameter ALP1 after clipping V none
lp_alp2 Local parameter ALP2 after clipping V−1 none
lp_vp Local parameter VP after clipping V none

559
 Table 2-124. MOSFET level 1031 Output Variables
Parameter Description Units Default
lp_a1 Local parameter A1 after clipping none
lp_a2 Local parameter A2 after T-scaling and clipping V none
lp_sta2 Local parameter STA2 after clipping none
lp_a3 Local parameter A3 after clipping none
lp_a4 Local parameter A4 after clipping 1/sqrt(V) none
lp_gco Local parameter GCO after clipping none
lp_iginv Local parameter IGINV after T-scaling and clipping A none
lp_igov Local parameter IGOV after T-scaling and clipping A none
lp_igovd Local parameter IGOVD after T-scaling and clipping A none
lp_stig Local parameter STIG after clipping none
lp_gc2 Local parameter GC2 after clipping none
lp_gc3 Local parameter GC3 after clipping none
lp_chib Local parameter CHIB after clipping V none
lp_agidl Local parameter AGIDL after clipping A/V3 none
lp_agidld Local parameter AGIDLD after clipping A/V3 none
lp_bgidl Local parameter BGIDL after T-scaling and clipping V none
lp_bgidld Local parameter BGIDLD after T-scaling and clipping V none
lp_stbgidl Local parameter STBGIDL after clipping V/K none
lp_stbgidld Local parameter STBGIDLD after clipping V/K none
lp_cgidl Local parameter CGIDL after clipping none
lp_cgidld Local parameter CGIDLD after clipping none
lp_cox Local parameter COX after clipping F none
lp_cgov Local parameter CGOV after clipping F none
lp_cgovd Local parameter CGOVD after clipping F none
lp_cgbov Local parameter CGBOV after clipping F none
lp_cfr Local parameter CFR after clipping F none
lp_cfrd Local parameter CFRD after clipping F none
lp_fnt Local parameter FNT after clipping none
lp_fntexc Local parameter FNTEXC after clipping none
lp_nfa Local parameter NFA after clipping 1/(V m4 ) none
lp_nfb Local parameter NFB after clipping 1/(V m4 ) none
lp_nfc Local parameter NFC after clipping V−1 none
lp_ef Local parameter EF after clipping none
Local parameter VFBEDGE after T-scaling and
lp_vfbedge V none
clipping
lp_stvfbedge Local parameter STVFBEDGE after clipping V/K none

560
 Table 2-124. MOSFET level 1031 Output Variables
Parameter Description Units Default
lp_dphibedge Local parameter DPHIBEDGE after clipping V none
lp_neffedge Local parameter NEFFEDGE after clipping m−3 none
lp_ctedge Local parameter CTEDGE after clipping none
Local parameter BETNEDGE after T-scaling and
lp_betnedge m2 /(Vs) none
clipping
lp_stbetedge Local parameter STBETEDGE after clipping none
lp_psceedge Local parameter PSCEEDGE after clipping none
lp_pscebedge Local parameter PSCEBEDGE after clipping V−1 none
lp_pscededge Local parameter PSCEDEDGE after clipping V−1 none
lp_cfedge Local parameter CFEDGE after clipping V none
lp_cfdedge Local parameter CFDEDGE after clipping V−1 none
lp_cfbedge Local parameter CFBEDGE after clipping V−1 none
lp_fntedge Local parameter FNTEDGE after clipping none
lp_nfaedge Local parameter NFAEDGE after clipping 1/(V m4 ) none
lp_nfbedge Local parameter NFBEDGE after clipping 1/(V m4 ) none
lp_nfcedge Local parameter NFCEDGE after clipping V−1 none
lp_efedge Local parameter EFEDGE after clipping none
lp_rg Local parameter RG after clipping Ohm none
lp_rse Local parameter RSE after clipping Ohm none
lp_rde Local parameter RDE after clipping Ohm none
lp_rbulk Local parameter RBULK after clipping Ohm none
lp_rwell Local parameter RWELL after clipping Ohm none
lp_rjuns Local parameter RJUNS after clipping Ohm none
lp_rjund Local parameter RJUND after clipping Ohm none
lp_rth Local parameter RTH after T-scaling and clipping K/W none
lp_cth Local parameter CTH after clipping J/K none
lp_strth Local parameter STRTH after clipping none
pdiss Power dissipation W none
dtsh Temperature rise due to self heating K none
tk Device Temperature K none
Bottom component of total zero-bias source junction
cjosbot F none
capacitance at device temperature
STI-edge component of total zero-bias source junction
cjossti F none
capacitance at device temperature
Gate-edge component of total zero-bias source
cjosgat F none
junction capacitance at device temperature

561
 Table 2-124. MOSFET level 1031 Output Variables
Parameter Description Units Default
Built-in voltage of source-side bottom junction at
vbisbot V none
device temperature
Built-in voltage of source-side STI-edge junction at
vbissti V none
device temperature
Built-in voltage of source-side gate-edge junction at
vbisgat V none
device temperature
idsatsbot Total source-side bottom junction saturation current A none
idsatssti Total source-side STI-edge junction saturation current A none
idsatsgat Total source-side gate-edge junction saturation current A none
Bottom component of total zero-bias drain junction
cjosbotd F none
capacitance at device temperature
STI-edge component of total zero-bias drain junction
cjosstid F none
capacitance at device temperature
Gate-edge component of total zero-bias drain junction
cjosgatd F none
capacitance at device temperature
Built-in voltage of drain-side bottom junction at
vbisbotd V none
device temperature
Built-in voltage of drain-side STI-edge junction at
vbisstid V none
device temperature
Built-in voltage of drain-side gate-edge junction at
vbisgatd V none
device temperature
idsatsbotd Total drain-side bottom junction saturation current A none
idsatsstid Total drain-side STI-edge junction saturation current A none
idsatsgatd Total drain-side gate-edge junction saturation current A none

562
2.3.20.13. Level 110 MOSFET Tables (BSIM CMG version 110.0.0)

Xyce includes the BSIM CMG Common Multi-gate model version 110. The code in Xyce was generated
from the BSIM group’s Verilog-A input using the default “ifdef” lines provided, and therefore supports
only the subset of BSIM CMG features those defaults enable. Instance and model parameters for the BSIM
CMG model are given in tables 2-125 and 2-126. Details of the model are documented in the BSIM-CMG
technical report[30], available from the BSIM web site at
http://bsim.berkeley.edu/models/bsimcmg/.
The BSIM CMG devices support output of the internal variables in tables 2-130, 2-133, and 2-127 on the
.PRINT line of a netlist. To access them from a print line, use the syntax N(<instance>:<variable>)
where “<instance>” refers to the name of the specific level 107 or 108 M device in your netlist.

Table 2-125. BSIM-CMG FINFET v110.0.0 Device Instance Parameters

Parameter Description Units Default
ADEJ Drain junction area (BULKMOD=1 or 2) m2 0
ADEO Drain-to-substrate overlap area through oxide m2 0
ASEJ Source junction area (BULKMOD=1 or 2) m2 0
ASEO Source-to-substrate overlap area through oxide m2 0
Constant drain-to-source fringe capacitance (all
CDSP F 0
CGEOMOD)
Constant gate-to-drain fringe capacitance
CGDP — 0
(CGEOMOD=1)
Constant gate-to-source fringe capacitance
CGSP — 0
(CGEOMOD=1)
Constant gate-to-drain overlap capacitance
COVD — 0
(CGEOMOD=1)
Constant gate-to-source overlap capacitance
COVS — 0
(CGEOMOD=1)
D Diameter of the cylinder (GEOMOD=3) m 4e-08
FPITCH Fin pitch m 8e-08
L Designed gate length m 3e-08
LRSD Length of the source/drain m 0
M multiplicity factor — 1
NF Number of fingers — 1
Number of fins per finger (real number enables
NFIN — 1
optimization)
NFINNOM Nominal number of fins per finger — 1
NGCON Number of gate contact (1 or 2 sided) — 1
NRD Number of source diffusion squares — 0
NRS Number of source diffusion squares — 0
Drain-to-substrate PN junction perimeter
PDEJ m 0
(BULKMOD=1 or 2)

563
 Table 2-125. BSIM-CMG FINFET v110.0.0 Device Instance Parameters
Parameter Description Units Default
Perimeter of drain-to-substrate overlap region through
PDEO m 0
oxide
Source-to-substrate PN junction perimeter
PSEJ m 0
(BULKMOD=1 or 2)
Perimeter of source-to-substrate overlap region
PSEO m 0
through oxide
TFIN Body (fin) thickness m 1.5e-08

Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters

Parameter Description Units Default
Non-saturation effect parameter for strong inversion
A1 — 0
Region
A11 Temperature dependence of A1 — 0
Non-saturation effect parameter for moderate
A2 — 0
Inversion Region
A21 Temperature dependence of A2 — 0
ACH_UFCM Area of the channel for the unified Model m2 1
ADEJ Drain junction area (BULKMOD=1 or 2) m2 0
ADEO Drain-to-substrate overlap area through oxide m2 0
ADVTP0 Pre-exponential coefficient for DITS — 0
ADVTP1 Pre-exponential coefficient for DVTP1 — 0
AEU Pre-exponential coefficient for EU — 0
AEUR Reverse-mode pre-exponential coefficient for EU — 0
AGIDL Pre-exponential coefficient for GIDL — 0
AGISL Pre-exponential coefficient for GISL — 6.055e-12
AIGBACC Parameter for Igb in accumulation — 0.0136
AIGBACC1 Parameter for Igb in accumulation — 0
AIGBINV Parameter for Igb in inversion — 0.0111
AIGBINV1 Parameter for Igb in inversion — 0
AIGC Parameter for Igc in inversion — 0.0136
AIGC1 Parameter for Igc in inversion — 0
AIGD Parameter for Igd in inversion — 0
AIGD1 Parameter for Igd in inversion — 0
AIGEN Thermal generation current parameter — 0
AIGS Parameter for Igs in inversion — 0.0136
AIGS1 Parameter for Igs in inversion — 0

564
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
ALPHA0 First parameter of Iii m/V 0
ALPHA01 Temperature dependence of ALPHA0 — 0
ALPHA1 L scaling parameter of Iii V−1 0
ALPHA11 Temperature dependence ALPHA1 — 0
Mobile charge scaling term taking QM effects into
ALPHA_UFCM — 0.5556
account
ALPHAII0 First parameter of Iii for IIMOD=2 m/V 0
ALPHAII01 Temperature dependence of ALPHAII0 — 0
ALPHAII1 L scaling parameter of Iii for IIMOD=2 V−1 0
ALPHAII11 Temperature dependence of ALPHAII1 — 0
AMEXP Pre-exponential coefficient for MEXP — 0
AMEXPR Pre-exponential coefficient for MEXPR — 0
APCLM Pre-exponential coefficient for PCLM — 0
APCLMR Reverse-mode pre-exponential coefficient for PCLM — 0
APSAT Pre-exponential coefficient for PSAT — 0
APSATCV Pre-exponential coefficient for PSATCV — 0
APTWG Pre-exponential coefficient for PTWG — 0
Parameter for geometric dependence of Tcen on
AQMTCEN — 0
R/TFIN/HFIN
ARDSW Pre-exponential coefficient for RDSW — 0
ARDW Pre-exponential coefficient for RDW — 0
Extra raised source/drain cross sectional areaat the
ARSDEND m2 0
two ends of the FinFET
ARSW Pre-exponential coefficient for RSW — 0
ASEJ Source junction area (BULKMOD=1 or 2) m2 0
ASEO Source-to-substrate overlap area through oxide m2 0
ASHEXP Exponent to tune RTH dependence of NFINTOTAL — 1
Extra silicide cross sectional area at the two ends of
ASILIEND m2 0
the FinFET
0: Turn off asymmetry model - forward mode
ASYMMOD — 0
parameters used; 1: Turn on asymmetry model
AT Saturation velocity temperature coefficient — -0.00156
ATCV Saturation velocity temperature coefficient for CV — 0
Reverse-mode saturation velocity temperature
ATR — 0
coefficient
AUA Pre-exponential coefficient for UA — 0
AUAR Reverse-mode pre-exponential coefficient for UA — 0
AUD Pre-exponential coefficient for UD — 0

565
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
AUDR Reverse-mode pre-exponential coefficient for UD — 0
AVSAT Pre-exponential coefficient for VSAT — 0
AVSAT1 Pre-exponential coefficient for VSAT1 — 0
AVSATCV Pre-exponential coefficient for VSATCV — 0
BDVTP0 Exponential coefficient for DITS — 1e-07
BDVTP1 Exponential coefficient for DVTP1 — 1e-07
BETA0 Vds dependence parameter of Iii V−1 0
BETAII0 Vds dependence parameter of Iii V−1 0
BETAII1 Vds dependence parameter of Iii — 0
BETAII2 Vds dependence parameter of Iii V 0.1
BEU Exponential coefficient for EU — 1e-07
BEUR Reverse-mode exponential coefficient for EU — 0
BG0SUB Bandgap of substrate at 300.15K — 1.12
BGIDL Exponential coefficient for GIDL — 0
BGISL Exponential coefficient for GISL — 3e+08
BIGBACC Parameter for Igb in accumulation — 0.00171
BIGBINV Parameter for Igb in inversion — 0.000949
BIGC Parameter for Igc in inversion — 0.00171
BIGD Parameter for Igd in inversion — 0
BIGEN Thermal generation current parameter — 0
BIGS Parameter for Igs in inversion — 0.00171
BMEXP Exponential coefficient for MEXP — 1
BMEXPR Exponential coefficient for MEXPR — 0
BPCLM Exponential coefficient for PCLM — 1e-07
BPCLMR Reverse-mode exponential coefficient for PCLM — 0
BPSAT Exponential coefficient for PSAT — 1
BPSATCV Exponential coefficient for PSATCV — 0
BPTWG Exponential coefficient for PTWG — 1e-07
Parameter for geometric dependence of Tcen on
BQMTCEN — 1.2e-08
R/TFIN/HFIN
BRDSW exponential coefficient for RDSW — 1e-07
BRDW Exponential coefficient for RDW — 1e-07
BRSW Exponential coefficient for RSW — 1e-07
BSHEXP Exponent to tune RTH dependence of NF — 1
BUA Exponential coefficient for UA — 1e-07
BUAR Reverse-mode exponential coefficient for UAR — 0

566
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
BUD Exponential coefficient for UD — 5e-08
BUDR Reverse-mode exponential coefficient for UD — 0
0: SOI multi-gate; 1: Bulk multi-gate; 2: for
BULKMOD — 0
decoupled bulk multi-gate
BVD Drain diode breakdown voltage V 0
BVS Source diode breakdown voltage V 10
BVSAT Exponential coefficient for VSAT — 1e-07
BVSAT1 Exponential coefficient for VSAT1 — 0
BVSATCV Exponential coefficient for VSATCV — 0
CDSC Coupling capacitance between S/D and channel — 0.007
CDSCD Drain-bias sensitivity of CDSC — 0.007
CDSCDN1 NFIN dependence of CDSCD — 0
CDSCDN2 NFIN dependence of CDSCD — 100000
CDSCDR Reverse-mode drain-bias sensitivity of CDSC — 0
CDSCDRN1 NFIN dependence of CDSCD — 0
CDSCDRN2 NFIN dependence of CDSCD — 0
CDSCN1 NFIN dependence of CDSC — 0
CDSCN2 NFIN dependence of CDSC — 100000
Constant drain-to-source fringe capacitance (all
CDSP F 0
CGEOMOD)
CFD Outer fringe capacitance at drain side — 0
CFS Outer fringe capacitance at source side — 2.5e-11
Bias dependent component of gate-to-substrate
CGBL overlap capacitance per unit channel length per fin per — 0
finger
Gate-to-substrate overlap capacitance per unit channel
CGBN — 0
length per fin per finger
Gate-to-substrate overlap capacitance per unit channel
CGBO — 0
length per finger per NGCON
Overlap capacitance between gate and lightly-doped
CGDL — 0
drain region (for CGEOMOD = 0, 2)
Non LDD region drain-gate overlap capacitance per
CGDO — 0
unit channel width
Constant gate-to-drain fringe capacitance
CGDP — 0
(CGEOMOD=1)
For CGEOMOD=1 only, this switch enables the
CGEO1SW parameters COVS, COVD, CGSP, and CGDP to be in — 0
F per fin, per gate-finger, per unit channel width
CGEOA Fitting parameter for CGEOMOD=2 — 1

567
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
CGEOB Fitting parameter for CGEOMOD=2 — 0
CGEOC Fitting parameter for CGEOMOD=2 — 0
CGEOD Fitting parameter for CGEOMOD=2 — 0
CGEOE Fitting parameter for CGEOMOD=2 — 1
Geometry-dependent parasitic capacitance model
CGEOMOD — 0
selector
CGIDL Parameter for body-effect of GIDL — 0
CGISL Parameter for body-effect of GISL — 0.5
Overlap capacitance between gate and lightly-doped
CGSL — 0
source region (for CGEOMOD = 0, 2)
Non LDD region source-gate overlap capacitance per
CGSO — 0
unit channel width
Constant gate-to-source fringe capacitance
CGSP — 0
(CGEOMOD=1)
Average channel charge weighting factor, +1:
CHARGEWF — 0
source-side, 0: middle, -1: drain-side
CIGBACC Parameter for Igb in accumulation V−1 0.075
CIGBINV Parameter for Igb in inversion V−1 0.006
CIGC Parameter for Igc in inversion V−1 0.075
CIGD Parameter for Igd in inversion V−1 0
CIGS Parameter for Igs in inversion V−1 0.075
CINS_UFCM Insulator capacitance for the unified Model — 1
CIT Parameter for interface trap — 0
Parameter for interface trap in reverse mode for
CITR — 0
asymmetric model
CJD Unit area drain-side junction capacitance at zero bias — 0
CJS Unit area source-side junction capacitance at zero bias — 0.0005
Unit length drain-side sidewall junction capacitance at
CJSWD — 0
zero bias
Unit length drain-side gate sidewall junction
CJSWGD — 0
capacitance at zero bias
Unit length source-side gate sidewall junction
CJSWGS — 0
capacitance at zero bias
Unit length source-side sidewall junction capacitance
CJSWS — 5e-10
at zero bias
CKAPPAB Bias dependent gate-to-substrate parasitic capacitance — 0.6
Coefficient of bias-dependent overlap capacitance for
CKAPPAD V 0
the drain side (for CGEOMOD = 0, 2)

568
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
Coefficient of bias-dependent overlap capacitance for
CKAPPAS V 0.6
the source side (for CGEOMOD = 0, 2)
Constant gate-to-drain overlap capacitance
COVD — 0
(CGEOMOD=1)
Constant gate-to-source overlap capacitance
COVS — 0
(CGEOMOD=1)
Ratio of the corner area filled with silicon to the total
CRATIO — 0.5
corner area
Coefficient for source/drain-to-substrate sidewall
CSDESW — 0
capacitance
CTH0 Thermal capacitance — 1e-05
D Diameter of the cylinder (GEOMOD=3) m 4e-08
Change in silicon/silicide interface length due to
DELTAPRSD m 0
non-rectangular epi
DELTAVSAT velocity saturation parameter in the linear region — 1
Velocity saturation parameter in the linear region for
DELTAVSATCV — 0
the capacitance model
DELTAW Change of effective width due to shape of fin/cylinder m 0
CV change of effective width due to shape of
DELTAWCV m 0
fin/cylinder
Change in flatband voltage: Vfb_accumulation -
DELVFBACC V 0
Vfb_inversion
DELVTRAND Variability in Vth V 0
DEVTYPE 0: PMOS; 1: NMOS — 1
DLBIN Delta L for binning m 0
DLC Delta L for C-V model m 0
Delta L for C-V model in accumulation region
DLCACC m 0
(BULKMOD=1 or 2)
DLCIGD Delta L for Igd model m 0
DLCIGS Delta L for Igs model m 0
DROUT L dependence of DIBL effect on Rout — 1.06
DSUB DIBL exponent coefficient — 1.06
DTEMP Variability in device temperature ◦C 0
DVT0 SCE coefficient — 0
SCE exponent coefficient. After binning it should be
DVT1 — 0.6
within (0:inf)
Subthreshold swing exponent coefficient. After
DVT1SS — 0
binning it should be within (0:inf)
DVTP0 Coefficient for drain-induced Vth shift (DITS) — 0

569
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
DVTP1 DITS exponent coefficient — 0
DVTP2 DITS model parameter — 0
DVTSHIFT Vth shift handle V 0
DVTSHIFTR Vth shift handle for asymmetric mode — 0
EASUB Electron affinity of substrate — 4.05
EF Flicker noise frequency exponent — 1
EGIDL Band bending parameter for GIDL V 0
EGISL Band bending parameter for GISL V 0.2
EIGBINV Parameter for Igb in inversion V 1.1
EM Flicker noise parameter — 4.1e+07
EMOBT Temperature coefficient of ETAMOB — 0
EOT Equivalent oxide thickness m 1e-09
EOTACC Equivalent oxide thickness for accumulation region m 0
Equivalent oxide thickness of the buried oxide (SOI
EOTBOX m 1.4e-07
FinFET)
EPSROX Relative dielectric constant of the gate dielectric — 3.9
EPSRSP Relative dielectric constant of the spacer — 3.9
EPSRSUB Relative dielectric constant of the channel material — 11.9
ESATII Saturation channel E-field for Iii — 1e+07
ETA0 DIBL coefficient — 0.6
ETA0LT Coupled NFIN and length dependence of ETA0 — 0
ETA0N1 NFIN dependence of ETA0 — 0
ETA0N2 NFIN dependence of ETA0 — 100000
ETA0R Reverse-mode DIBL coefficient — 0
ETAMOB Effective field parameter — 2
ETAQM Bulk charge coefficient for Tcen — 0.54
EU Phonon/surface roughness scattering parameter — 2.5
Reverse-mode phonon/surface roughness scattering
EUR — 0
parameter
FECH End-channel factor for different orientation/shape — 1
FECHCV CV end-channel factor for different orientation/shape — 1
FPITCH Fin pitch m 8e-08
0: Double gate; 1: Triple gate; 2: Quadruple gate; 3:
GEOMOD — 0
Cylindrical gate; 4: Unified fin Shape
0: Turn off GIDL/GISL current; 1: Turn on
GIDLMOD — 0
GIDL/GISL current
HEPI Height of the raised source/drain on top of the fin m 1e-08

570
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
HFIN Fin height m 3e-08
IDS0MULT Variability in drain current for miscellaneous reasons — 1
IGBMOD 0: Turn off Igb; 1: Turn on Igb — 0
0: Disable gate current clamps; 1: Enable gate current
IGCLAMP — 1
clamps
IGCMOD 0: Turn off Igc, Igs and Igd; 1: Turn on Igc, Igs and Igd — 0
IGT Gate current temperature dependence — 2.5
0: Turn off impact ionization current; 1: BSIM4-based
IIMOD — 0
model; 2: BSIMSOI-based model
IIMOD2CLAMP1 Clamp1 of SII1*Vg term in IIMOD=2 model V 0.1
IIMOD2CLAMP2 Clamp2 of SII0*Vg term in IIMOD=2 model V 0.1
IIMOD2CLAMP3 Clamp3 of Ratio term in IIMOD=2 model V 0.1
Impact ionization temperature dependence for IIMOD
IIT — -0.5
=1
IJTHDFWD Forward drain diode breakdown limiting current A 0
IJTHDREV Reverse drain diode breakdown limiting current A 0
IJTHSFWD Forward source diode breakdown limiting current A 0.1
IJTHSREV Reverse source diode breakdown limiting current A 0.1
Parameter for Vgs clamping for inversion region
IMIN — 1e-15
calculation in accumulation
Bottom drain junction reverse saturation current
JSD — 0
density
Bottom source junction reverse saturation current
JSS — 0.0001
density
Unit length reverse saturation current for sidewall
JSWD — 0
drain junction
Unit length reverse saturation current for gate-edge
JSWGD — 0
sidewall drain junction
Unit length reverse saturation current for gate-edge
JSWGS — 0
sidewall source junction
Unit length reverse saturation current for sidewall
JSWS — 0
source junction
Bottom drain junction trap-assisted saturation current
JTSD — 0
density
Bottom source junction trap-assisted saturation
JTSS — 0
current density
Unit length trap-assisted saturation current for
JTSSWD — 0
sidewall drain junction
Unit length trap-assisted saturation current for
JTSSWGD — 0
gate-edge sidewall drain junction

571
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
Unit length trap-assisted saturation current for
JTSSWGS — 0
gate-edge sidewall source junction
Unit length trap-assisted saturation current for
JTSSWS — 0
sidewall source junction
JTWEFF Trap-assisted tunneling current width dependence m 0
K0 Lateral NUD voltage parameter V 0
Temperature dependence of lateral NUD voltage
K01 V/K 0
parameter
Correction factor for strong inversion used in Mnud.
K0SI — 1
After binning it should be within (0:inf)
K0SI1 Temperature dependence of K0SI — 0
K0SISAT Correction factor for strong inversion used in Mnud — 0
K0SISAT1 Temperature dependence of K0SISAT — 0
K1 Body effect coefficient for subthreshold region — 1e-06
K11 Temperature dependence of K1 — 0
K1RSCE K1 for reverse short channel effect calculation — 0
K2 Body effect coefficient for BULKMOD==2 — 0
K21 Temperature dependence of K2 — 0
K2SAT Correction factor for K2 in saturation (high Vds) — 0
K2SAT1 Temperature dependence of K2SAT — 0
K2SI Correction factor for strong inversion used in Mob — 0
K2SI1 Temperature dependence of K2SI — 0
K2SISAT Correction factor for strong inversion used in Mob — 0
K2SISAT1 Temperature dependence of K2SISAT — 0
KSATIV Parameter for long channel Vdsat — 1
KSATIVR KSATIV in asymmetric mode — 0
KT1 Vth temperature coefficient V 0
KT1L Vth temperature L coefficient — 0
L Designed gate length m 3e-08
LA1 — 0
LA11 — 0
LA2 — 0
LA21 — 0
LAGIDL — 0
LAGISL — 0
LAIGBACC — 0
LAIGBACC1 — 0

572
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
LAIGBINV — 0
LAIGBINV1 — 0
LAIGC — 0
LAIGC1 — 0
LAIGD — 0
LAIGD1 — 0
LAIGEN — 0
LAIGS — 0
LAIGS1 — 0
LALPHA0 — 0
LALPHA1 m/V 0
LALPHAII0 — 0
LALPHAII1 m/V 0
LAT — 0
LATCV — 0
LATR — 0
LBETA0 m/V 0
LBETAII0 m/V 0
LBETAII1 — 0
LBETAII2 — 0
LBGIDL V 0
LBGISL V 0
LBIGBACC — 0
LBIGBINV — 0
LBIGC — 0
LBIGD — 0
LBIGEN — 0
LBIGS — 0
LCDSC — 0
LCDSCD — 0
LCDSCDR — 0
LCFD F 0
LCFS F 0
LCGBL F 0
LCGDL F 0
LCGIDL — 0

573
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
LCGISL — 0
LCGSL F 0
LCIGBACC m/V 0
LCIGBINV m/V 0
LCIGC m/V 0
LCIGD m/V 0
LCIGS m/V 0
LCIT — 0
LCITR — 0
LCKAPPAB — 0
LCKAPPAD — 0
LCKAPPAS — 0
LCOVD F 0
LCOVS F 0
LDELTAVSAT — 0
LDELTAVSATCV — 0
LDROUT — 0
LDSUB — 0
LDVT0 — 0
LDVT1 — 0
LDVT1SS — 0
LDVTB — 0
LDVTSHIFT — 0
LDVTSHIFTR — 0
LEGIDL — 0
LEGISL — 0
LEIGBINV — 0
LEMOBT — 0
LESATII V 0
LETA0 — 0
LETA0R — 0
LETAMOB — 0
LEU — 0
LEUR — 0
LIGT — 0
LII Channel length dependence parameter of Iii — 5e-10

574
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
LIIT — 0
LINT Length reduction parameter (dopant diffusion effect) m 0
LINTIGEN Lint for thermal generation current m 0
LINTNOI L offset for flicker noise calculation m2 0
LK0 — 0
LK01 — 0
LK0SI — 0
LK0SI1 — 0
LK0SISAT — 0
LK0SISAT1 — 0
LK1 — 0
LK11 — 0
LK1RSCE — 0
LK2 — 0
LK21 — 0
LK2SAT — 0
LK2SAT1 — 0
LK2SI — 0
LK2SI1 — 0
LK2SISAT — 0
LK2SISAT1 — 0
LKSATIV — 0
LKSATIVR — 0
LKT1 — 0
LL Length reduction parameter (dopant diffusion effect) — 0
LLC Length reduction parameter (dopant diffusion effect) — 0
LLII — 0
LLN Length reduction parameter (dopant diffusion effect) — 1
LLPE0 m2 0
LLPEB — 0
LMEXP — 0
LMEXPR — 0
LNBODY — 0
LNGATE — 0
LNIGBACC — 0
LNIGBINV — 0

575
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
LNTGEN — 0
LNTOX — 0
LPA Mobility L power coefficient — 1
LPAR Reverse-mode mobility L power coefficient — 0
LPCLM — 0
LPCLMCV — 0
LPCLMG — 0
LPCLMR — 0
LPDIBL1 — 0
LPDIBL1R — 0
LPDIBL2 — 0
LPDIBL2R — 0
LPE0 Equivalent length of pocket region at zero bias m 5e-09
LPGIDL — 0
LPGISL — 0
LPHIBE — 0
LPHIG — 0
LPHIN — 0
LPIGCD — 0
LPOXEDGE — 0
LPRT — 0
LPRWGD m/V 0
LPRWGS m/V 0
LPSAT — 0
LPSATCV — 0
LPTWG — 0
LPTWGR — 0
LPTWGT — 0
LPVAG — 0
LQMFACTOR — 0
LQMTCENCV — 0
LQMTCENCVA — 0
LRDSW — 0
LRDW — 0
LRSD Length of the source/drain m 0
LRSW — 0

576
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
LSII0 m/V 0
LSII1 — 0
LSII2 — 0
LSIID — 0
LSP Thickness of the gate sidewall spacer m 0
LSTTHETASAT — 0
LTGIDL — 0
LTII — 0
LTSS — 0
LU0 — 0
LU0R — 0
LUA — 0
LUA1 — 0
LUA1R — 0
LUAR — 0
LUC — 0
LUC1 — 0
LUC1R — 0
LUCR — 0
LUCS — 0
LUCSTE — 0
LUD — 0
LUD1 — 0
LUD1R — 0
LUDR — 0
LUP — 0
LUPR — 0
LUTE — 0
LUTER — 0
LUTL — 0
LUTLR — 0
LVSAT — 0
LVSAT1 — 0
LVSAT1R — 0
LVSATCV — 0
LVSATR — 0

577
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
LWR — 0
LXRCRG1 — 0
LXRCRG2 — 0
MEXP Smoothing function factor for Vdsat — 4
MEXPR Reverse-mode smoothing function factor for Vdsat — 0
MJD Drain bottom junction capacitance grading coefficient — 0
Drain bottom two-step second junction capacitance
MJD2 — 0
grading coefficient
Source bottom junction capacitance grading
MJS — 0.5
coefficient
Source bottom two-step second junction capacitance
MJS2 — 0.125
grading coefficient
Drain sidewall junction capacitance grading
MJSWD — 0
coefficient
Drain sidewall two-step second junction capacitance
MJSWD2 — 0
grading coefficient
Drain-side gate sidewall junction capacitance grading
MJSWGD — 0
coefficient
Drain-side gate sidewall two-step second junction
MJSWGD2 — 0
capacitance grading coefficient
Source-side gate sidewall junction capacitance
MJSWGS — 0
grading coefficient
Source-side gate sidewall two-step second junction
MJSWGS2 — 0
capacitance grading coefficient
Source sidewall junction capacitance grading
MJSWS — 0.33
coefficient
Source sidewall two-step second junction capacitance
MJSWS2 — 0.083
grading coefficient
NA1 — 0
NA11 — 0
NA2 — 0
NA21 — 0
NAGIDL — 0
NAGISL — 0
NAIGBACC — 0
NAIGBACC1 — 0
NAIGBINV — 0
NAIGBINV1 — 0
NAIGC — 0

578
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
NAIGC1 — 0
NAIGD — 0
NAIGD1 — 0
NAIGEN — 0
NAIGS — 0
NAIGS1 — 0
NALPHA0 — 0
NALPHA1 m/V 0
NALPHAII0 — 0
NALPHAII1 m/V 0
NAT — 0
NATCV — 0
NATR — 0
NBETA0 m/V 0
NBETAII0 m/V 0
NBETAII1 — 0
NBETAII2 — 0
NBGIDL V 0
NBGISL V 0
NBIGBACC — 0
NBIGBINV — 0
NBIGC — 0
NBIGD — 0
NBIGEN — 0
NBIGS — 0
NBODY Channel (body) doping — 1e+22
NBODYN1 NFIN dependence of channel (body) doping — 0
NBODYN2 NFIN dependence of channel (body) doping — 100000
NC0SUB Conduction band density of states — 2.86e+25
NCDSC — 0
NCDSCD — 0
NCDSCDR — 0
NCFD F 0
NCFS F 0
NCGBL F 0
NCGDL F 0

579
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
NCGIDL — 0
NCGISL — 0
NCGSL F 0
NCIGBACC m/V 0
NCIGBINV m/V 0
NCIGC m/V 0
NCIGD m/V 0
NCIGS m/V 0
NCIT — 0
NCITR — 0
NCKAPPAB — 0
NCKAPPAD — 0
NCKAPPAS — 0
NCOVD F 0
NCOVS F 0
NDELTAVSAT — 0
NDELTAVSATCV — 0
NDROUT — 0
NDSUB — 0
NDVT0 — 0
NDVT1 — 0
NDVT1SS — 0
NDVTB — 0
NDVTSHIFT — 0
NDVTSHIFTR — 0
NEGIDL — 0
NEGISL — 0
NEIGBINV — 0
NEMOBT — 0
NESATII V 0
NETA0 — 0
NETA0R — 0
NETAMOB — 0
NEU — 0
NEUR — 0
NF Number of fingers — 1

580
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
Number of fins per finger (real number enables
NFIN — 1
optimization)
Parameter for poly gate doping. For metal gate please
NGATE — 0
set NGATE = 0
NGCON Number of gate contact (1 or 2 sided) — 1
NI0SUB Intrinsic carrier constant at 300.15K — 1.1e+16
NIGBACC Parameter for Igb in accumulation — 1
NIGBINV Parameter for Igb in inversion — 3
NIGT — 0
NIIT — 0
NJD Drain junction emission coefficient — 0
NJS Source junction emission coefficient — 1
NJTS Non-ideality factor for JTSS — 20
NJTSD Non-ideality factor for JTSD — 0
NJTSSW Non-ideality factor for JTSSWS — 20
NJTSSWD Non-ideality factor for JTSSWD — 0
NJTSSWG Non-ideality factor for JTSSWGS — 20
NJTSSWGD Non-ideality factor for JTSSWGD — 0
NK0 — 0
NK01 — 0
NK0SI — 0
NK0SI1 — 0
NK0SISAT — 0
NK0SISAT1 — 0
NK1 — 0
NK11 — 0
NK1RSCE — 0
NK2 — 0
NK21 — 0
NK2SAT — 0
NK2SAT1 — 0
NK2SI — 0
NK2SI1 — 0
NK2SISAT — 0
NK2SISAT1 — 0
NKSATIV — 0

581
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
NKSATIVR — 0
NKT1 — 0
NLII — 0
NLPE0 m2 0
NLPEB — 0
NMEXP — 0
NMEXPR — 0
NNBODY — 0
NNGATE — 0
NNIGBACC — 0
NNIGBINV — 0
NNTGEN — 0
NNTOX — 0
NOIA Flicker noise parameter — 6.25e+39
NOIB Flicker noise parameter — 3.125e+24
NOIC Flicker noise parameter — 8.75e+07
NPCLM — 0
NPCLMCV — 0
NPCLMG — 0
NPCLMR — 0
NPDIBL1 — 0
NPDIBL1R — 0
NPDIBL2 — 0
NPDIBL2R — 0
NPGIDL — 0
NPGISL — 0
NPHIBE — 0
NPHIG — 0
NPHIN — 0
NPIGCD — 0
NPOXEDGE — 0
NPRT — 0
NPRWGD m/V 0
NPRWGS m/V 0
NPSAT — 0
NPSATCV — 0

582
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
NPTWG — 0
NPTWGR — 0
NPTWGT — 0
NPVAG — 0
NQMFACTOR — 0
NQMTCENCV — 0
NQMTCENCVA — 0
0: Turn off NQS model; 1: NQS gate resistance (with
NQSMOD gi node); 2: NQS charge deficit model from BSIM4 — 0
(with q node)
NRD Number of source diffusion squares — 0
NRDSW — 0
NRDW — 0
NRS Number of source diffusion squares — 0
NRSW — 0
NSD Source/drain active doping concentration — 2e+26
NSDE Source/drain active doping concentration at Leff edge — 2e+25
Number of segments for NQSMOD=3 (3, 5 and 10
NSEG — 4
supported)
NSII0 m/V 0
NSII1 — 0
NSII2 — 0
NSIID — 0
NSTTHETASAT — 0
NTGEN Thermal generation current parameter — 1
NTGIDL — 0
NTII — 0
NTNOI Thermal noise parameter — 1
NTOX Exponent for Tox ratio — 1
NTSS — 0
NU0 — 0
NU0R — 0
NUA — 0
NUA1 — 0
NUA1R — 0
NUAR — 0
NUC — 0

583
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
NUC1 — 0
NUC1R — 0
NUCR — 0
NUCS — 0
NUCSTE — 0
NUD — 0
NUD1 — 0
NUD1R — 0
NUDR — 0
NUP — 0
NUPR — 0
NUTE — 0
NUTER — 0
NUTL — 0
NUTLR — 0
NVSAT — 0
NVSAT1 — 0
NVSAT1R — 0
NVSATCV — 0
NVSATR — 0
Subthreshold swing factor multiplied by Vtm. If
NVTM V 0
defined by user, it will overwrite nVtm in the code
NWR — 0
NXRCRG1 — 0
NXRCRG2 — 0
PA1 — 0
PA11 — 0
PA2 — 0
PA21 — 0
PAGIDL — 0
PAGISL — 0
PAIGBACC — 0
PAIGBACC1 — 0
PAIGBINV — 0
PAIGBINV1 — 0
PAIGC — 0

584
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
PAIGC1 — 0
PAIGD — 0
PAIGD1 — 0
PAIGEN — 0
PAIGS — 0
PAIGS1 — 0
PALPHA0 — 0
PALPHA1 — 0
PALPHAII0 — 0
PALPHAII1 — 0
PAT — 0
PATCV — 0
PATR — 0
PBD Drain-side bulk junction built-in potential V 0
PBETA0 — 0
PBETAII0 — 0
PBETAII1 — 0
PBETAII2 — 0
PBGIDL — 0
PBGISL — 0
PBIGBACC — 0
PBIGBINV — 0
PBIGC — 0
PBIGD — 0
PBIGEN — 0
PBIGS — 0
PBS Source-side bulk junction built-in potential V 1
Built-in potential for Drain-side sidewall junction
PBSWD V 0
capacitance
Built-in potential for Drain-side gate sidewall junction
PBSWGD V 0
capacitance
Built-in potential for Source-side gate sidewall
PBSWGS V 0
junction capacitance
Built-in potential for Source-side sidewall junction
PBSWS V 1
capacitance
PCDSC F 0
PCDSCD F 0

585
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
PCDSCDR F 0
PCFD — 0
PCFS — 0
PCGBL — 0
PCGDL — 0
PCGIDL — 0
PCGISL — 0
PCGSL — 0
PCIGBACC — 0
PCIGBINV — 0
PCIGC — 0
PCIGD — 0
PCIGS — 0
PCIT F 0
PCITR — 0
PCKAPPAB — 0
PCKAPPAD — 0
PCKAPPAS — 0
PCLM Channel length modulation (CLM) parameter — 0.013
PCLMCV CLM parameter for short-channel CV — 0
PCLMG Gate bias dependence parameter for CLM — 0
PCLMR Reverse model PCLM parameter — 0
PCOVD — 0
PCOVS — 0
Drain-to-substrate PN junction perimeter
PDEJ m 0
(BULKMOD=1 or 2)
PDELTAVSAT — 0
PDELTAVSATCV — 0
Perimeter of drain-to-substrate overlap region through
PDEO m 0
oxide
PDIBL1 DIBL output conductance parameter - forward mode — 1.3
PDIBL1R DIBL output conductance parameter - reverse mode — 0
PDIBL2 DIBL output conductance parameter — 0.0002
PDIBL2R DIBL output conductance parameter - reverse mode — 0
PDROUT — 0
PDSUB — 0

586
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
PDVT0 — 0
PDVT1 — 0
PDVT1SS — 0
PDVTB — 0
PDVTSHIFT — 0
PDVTSHIFTR — 0
PEGIDL — 0
PEGISL — 0
PEIGBINV — 0
PEMOBT — 0
PESATII — 0
PETA0 — 0
PETA0R — 0
PETAMOB — 0
PEU — 0
PEUR — 0
PGIDL Parameter for body-bias effect on GIDL — 0
PGISL Parameter for body-bias effect on GISL — 1
Body effect voltage parameter. After binning it should
PHIBE V 0.7
be within [0.2:1.2]
PHIG Gate workfunction — 4.61
PHIGL Length dependence of gate workfunction — 0
Coupled NFIN and length dependence of gate
PHIGLT — 0
workfunction
PHIGN1 NFIN dependence of gate workfunction — 0
PHIGN2 NFIN dependence of gate workfunction — 100000
PHIN Nonuniform vertical doping effect on surface potential V 0.05
PIGCD Parameter for Igc partition — 1
PIGT — 0
PIIT — 0
PK0 — 0
PK01 — 0
PK0SI — 0
PK0SI1 — 0
PK0SISAT — 0
PK0SISAT1 — 0

587
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
PK1 — 0
PK11 — 0
PK1RSCE — 0
PK2 — 0
PK21 — 0
PK2SAT — 0
PK2SAT1 — 0
PK2SI — 0
PK2SI1 — 0
PK2SISAT — 0
PK2SISAT1 — 0
PKSATIV — 0
PKSATIVR — 0
PKT1 — 0
PLII — 0
PLPE0 — 0
PLPEB — 0
PMEXP — 0
PMEXPR — 0
PNBODY — 0
PNGATE — 0
PNIGBACC — 0
PNIGBINV — 0
PNTGEN — 0
PNTOX — 0
POXEDGE Factor for the gate edge Tox — 1
PPCLM — 0
PPCLMCV — 0
PPCLMG — 0
PPCLMR — 0
PPDIBL1 — 0
PPDIBL1R — 0
PPDIBL2 — 0
PPDIBL2R — 0
PPGIDL — 0
PPGISL — 0

588
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
PPHIBE — 0
PPHIG — 0
PPHIN — 0
PPIGCD — 0
PPOXEDGE — 0
PPRT — 0
PPRWGD — 0
PPRWGS — 0
PPSAT — 0
PPSATCV — 0
PPTWG — 0
PPTWGR — 0
PPTWGT — 0
PPVAG — 0
PQM Slope of normalized Tcen in inversion — 0.66
PQMACC Slope of normalized Tcen in accumulation — 0.66
PQMFACTOR — 0
PQMTCENCV — 0
PQMTCENCVA — 0
PRDDR Drain-side quasi-saturation parameter — 0
PRDSW — 0
PRDW — 0
Extra silicon/silicide interface perimeter at the two
PRSDEND m 0
ends of the FinFET
PRSDR Source-side quasi-saturation parameter — 1
PRSW — 0
PRT Series resistance temperature coefficient — 0.001
PRWGD Gate bias dependence of drain extension resistance V−1 0
PRWGS Gate bias dependence of source extension resistance V−1 0
Velocity saturation exponent, after binnig should be
PSAT — 2
from [2.0:inf)
PSATCV Velocity saturation exponent for C-V — 0
Source-to-substrate PN junction perimeter
PSEJ m 0
(BULKMOD=1 or 2)
Perimeter of source-to-substrate overlap region
PSEO m 0
through oxide
PSII0 — 0

589
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
PSII1 — 0
PSII2 — 0
PSIID — 0
PSTTHETASAT — 0
PTGIDL — 0
PTII — 0
PTSS — 0
PTWG Gmsat degradation parameter - forward mode — 0
PTWGR Gmsat degradation parameter - reverse mode — 0
PTWGT PTWG temperature coefficient — 0.004
PU0 — 0
PU0R — 0
PUA — 0
PUA1 — 0
PUA1R — 0
PUAR — 0
PUC — 0
PUC1 — 0
PUC1R — 0
PUCR — 0
PUCS — 0
PUCSTE — 0
PUD — 0
PUD1 — 0
PUD1R — 0
PUDR — 0
PUP — 0
PUPR — 0
PUTE — 0
PUTER — 0
PUTL — 0
PUTLR — 0
PVAG Vgs dependence on early voltage — 1
PVSAT — 0
PVSAT1 — 0
PVSAT1R — 0

590
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
PVSATCV — 0
PVSATR — 0
PWR — 0
PXRCRG1 — 0
PXRCRG2 — 0
Knee-point for Tcen in inversion (Charge normalized
QM0 V 0.001
to Cox)
Knee-point for Tcen in accumulation (Charge
QM0ACC V 0.001
normalized to Cox)
QMFACTOR Prefactor + switch for QM Vth correction — 0
QMFACTORCV Charge dependence taking QM effects into account — 0
Prefactor + switch for QM Width and Toxeff
QMTCENCV — 0
correction for CV
Prefactor + switch for QM Width and Toxeff
QMTCENCVA — 0
correction for CV (accumulation region)
RDDR Drain-side drift resistance parameter - forward mode — 0
RDDRR Drain-side drift resistance parameter - reverse mode — 0
0: Internal S/D resistance model; 1: External S/D
RDSMOD resistance model; 2: Both bias dependent and — 0
independent part of S/D resistance internal
RDSMOD = 0 zero bias S/D extension resistance per
RDSW — 100
unit width
RDSMOD = 0 S/D extension resistance per unit width
RDSWMIN — 0
at high Vgs
RDSMOD = 1 zero bias drain extension resistance per
RDW — 50
unit width
RDSMOD = 1 drain extension resistance per unit
RDWMIN — 0
width at high Vgs
0: Turn off gate electrode resistance (without ge node);
RGATEMOD — 0
1: Turn on gate electrode resistance (with ge node)
RGEOA Fitting parameter for RGEOMOD=1 — 1
RGEOB Fitting parameter for RGEOMOD=1 — 0
RGEOC Fitting parameter for RGEOMOD=1 — 0
RGEOD Fitting parameter for RGEOMOD=1 — 0
RGEOE Fitting parameter for RGEOMOD=1 — 0
Geometry-dependent source/drain resistance; 0:
RGEOMOD — 0
RSH-based; 1: Holistic
RGEXT Effective gate electrode external resistance — 0
RGFIN Effective gate electrode per finger per fin resistance — 0.001

591
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
RHOC Contact resistivity at the silicon/silicide interface — 1e-12
Average resistivity of silicon in the raised source/drain
RHORSD — 1
region
RNOIA Thermal noise coefficient — 0.577
RNOIB Thermal noise coefficient — 0.37
RNOIC Thermal noise coefficient for TNOIMOD=1 — 0.395
RSDR Source-side drift resistance parameter - forward mode — 0
RSDRR Source-side drift resistance parameter - reverse mode — 0
RSHD Drain-side sheet resistance — 0
RSHS Source-side sheet resistance — 0
RDSMOD = 1 zero bias source extension resistance
RSW — 50
per unit width
RDSMOD = 1 source extension resistance per unit
RSWMIN — 0
width at high Vgs
RTH0 Thermal resistance — 0.01
SCALEN Noise scaling parameter for TNOIMOD=1 — 100000
Indicator of whether the source/drain are terminated
SDTERM — 0
with silicide
0: Disable self-heating warnings; 1: Enable
SH_WARN — 0
self-heating warnings
SHMOD 0: Turn off self-heating; 1: Turn on self-heating — 0
SII0 Vgs dependence parameter of Iii V−1 0.5
SII1 1st Vgs dependence parameter of Iii — 0.1
SII2 2nd Vgs dependence parameter of Iii V 0
SIID 3rd Vds dependence parameter of Iii V 0
SJD Constant for drain-side two-step second junction — 0
SJS Constant for source-side two-step second junction — 0
Constant for drain-side sidewall two-step second
SJSWD — 0
junction
Constant for source-side gate sidewall two-step
SJSWGD — 0
second junction
Constant for source-side gate sidewall two-step
SJSWGS — 0
second junction
Constant for source-side sidewall two-step second
SJSWS — 0
junction
TBGASUB Bandgap temperature coefficient — 0.000702
TBGBSUB Bandgap temperature coefficient K 1108
TCJ Temperature coefficient for CJS/CJD — 0

592
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
TCJSW Temperature coefficient for CJSWS/CJSWD — 0
TCJSWG Temperature coefficient for CJSWGS/CJSWGD — 0
1: Change temperature dependence of specific
TEMPMOD — 0
parameters
TETA0 Temperature dependence of DIBL coefficient — 0
Temperature dependence of reverse-mode DIBL
TETA0R — 0
coefficient
TFIN Body (fin) thickness m 1.5e-08
TFIN_BASE Base body (fin) thickness for trapezoidal triple gate m 1.5e-08
TFIN_TOP Top body (fin) thickness for trapezoidal triple gate m 1.5e-08
TGATE Gate height on top of the hard mask m 3e-08
TGIDL GIDL/GISL temperature dependence — -0.003
DIBL length dependence. If defined by user, will
THETADIBL — 0
overwrite Theta_DIBL in the code
Vth roll-off length dependence. If defined by user, it
THETASCE — 0
will overwrite Theta_SCE in the code
Subthreshold swing length dependence. If defined by
THETASW — 0
user, it will overwrite Theta_SW in the code
Impact ionization temperature dependence for IIMOD
TII — 0
=2
TMASK Height of hard mask on top of the fin m 3e-08
TMEXP Temperature coefficient for Vdseff smoothing — 0
Reverse-mode temperature coefficient for Vdseff
TMEXPR — 0
smoothing
TNJTS Temperature coefficient for NJTS — 0
TNJTSD Temperature coefficient for NJTSD — 0
TNJTSSW Temperature coefficient for NJTSSW — 0
TNJTSSWD Temperature coefficient for NJTSSWD — 0
TNJTSSWG Temperature coefficient for NJTSSWG — 0
TNJTSSWGD Temperature coefficient for NJTSSWGD — 0
TNOIA Thermal noise parameter — 1.5
TNOIB Thermal noise parameter — 3.5
TNOIC Thermal noise parameter for TNOIMOD=1 — 3.5
TNOIMOD 0: Charge-based, 1: Correlated thermal noise model — 0
TNOM Temperature at which the model is extracted — 27
TOXG Oxide thickness for gate current model m 0
TOXP Physical oxide thickness m 1.2e-09
TOXREF Target tox value m 1.2e-09

593
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
TPB Temperature coefficient for PBS/PBD — 0
TPBSW Temperature coefficient for PBSWS/PBSWD — 0
TPBSWG Temperature coefficient for PBSWGS/PBSWGD — 0
TRDDR Drain-side drift resistance temperature coefficient — 0
TRSDR Source-side drift resistance temperature coefficient — 0
Thickness of the silicide on top of the raised
TSILI m 1e-08
source/drain
TSS Swing temperature coefficient — 0
TYPE 0: PMOS; 1: NMOS — 0
U0 Low-field mobility — 0.03
U0LT Coupled NFIN and length dependence of U0 — 0
U0MULT Variability in carrier mobility — 1
U0N1 NFIN dependence of U0 — 0
U0N1R Reverse-mode NFIN dependence of U0 — 0
U0N2 NFIN dependence of U0 — 100000
U0N2R Reverse-mode NFIN dependence of U0 — 0
U0R Reverse-mode low-field mobility — 0
UA Phonon/surface roughness scattering parameter — 0.3
UA1 Mobility temperature coefficient for UA — 0.001032
UA1R Reverse-mode mobility temperature coefficient for UA — 0
Reverse-mode phonon/surface roughness scattering
UAR — 0
parameter
Body effect for mobility degradation parameter -
UC — 0
BULKMOD=1 or 2
UC1 Mobility temperature coefficient for UC — 5.6e-11
UC1R Reverse-mode mobility temperature coefficient for UC — 0
Reverse-mode body effect for mobility degradation
UCR — 0
parameter - BULKMOD=1 or 2
UCS Columbic scattering parameter — 1
UCSTE Mobility temperature coefficient — -0.004775
UD Columbic scattering parameter — 0
UD1 Mobility temperature coefficient for UC — 0
UD1R Reverse-mode mobility temperature coefficient for UD — 0
UDR Reverse-mode columbic scattering parameter — 0
UP Mobility L coefficient — 0
UPR Reverse-mode mobility L coefficient — 0
UTE Mobility temperature coefficient — 0

594
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
UTER Reverse-mode for mobility temperature coefficient — 0
UTL Mobility temperature coefficient — -0.0015
UTLR Reverse-mode for mobility temperature coefficient — 0
VFBSD Flatband voltage for S/D region V 0
VFBSDCV Flatband voltage for S/D region for C-V calculations V 0
VSAT Saturation velocity for the saturation region — 85000
Velocity saturation parameter for Ion degradation -
VSAT1 — 0
forward mode
VSAT1N1 NFIN dependence of VSAT1 — 0
VSAT1N2 NFIN dependence of VSAT1 — 0
Velocity saturation parameter for Ion degradation -
VSAT1R — 0
reverse mode
VSAT1RN1 NFIN dependence of VSAT1R — 0
VSAT1RN2 NFIN dependence of VSAT1R — 0
VSATCV Velocity saturation parameter for CV — 0
VSATN1 NFIN dependence of VSAT — 0
VSATN2 NFIN dependence of VSAT — 100000
Saturation velocity for the saturation region in the
VSATR — 0
reverse mode
VSATRN1 NFIN dependence of VSATR — 0
VSATRN2 NFIN dependence of VSATR — 0
Bottom drain junction trap-assisted current voltage
VTSD V 0
dependent parameter
Bottom source junction trap-assisted current voltage
VTSS V 10
dependent parameter
Unit length trap-assisted current voltage dependent
VTSSWD V 0
parameter for sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGD V 0
parameter for gate-edge sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGS V 10
parameter for gate-edge sidewall source junction
Unit length trap-assisted current voltage dependent
VTSSWS V 10
parameter for sidewall source junction
W_UFCM Effective channel width for the unified Model m 1
WR W dependence parameter of S/D extension resistance — 1
WTH0 Width dependence coefficient for Rth and Cth m 0
XJBVD Fitting parameter for drain diode breakdown current — 0
XJBVS Fitting parameter for source diode breakdown current — 1

595
 Table 2-126. BSIM-CMG FINFET v110.0.0 Device Model Parameters
Parameter Description Units Default
XL L offset for channel length due to mask/etch effect m 0
Parameter for non-quasistatic gate resistance
XRCRG1 — 12
(NQSMOD = 1) and NQSMOD = 2
Parameter for non-quasistatic gate resistance
XRCRG2 — 1
(NQSMOD = 1) and NQSMOD = 2
XTID Drain junction current temperature exponent — 0
XTIS Source junction current temperature exponent — 3
XTSD Power dependence of JTSD on temperature — 0
XTSS Power dependence of JTSS on temperature — 0.02
XTSSWD Power dependence of JTSSWD on temperature — 0
XTSSWGD Power dependence of JTSSWGD on temperature — 0
XTSSWGS Power dependence of JTSSWGS on temperature — 0.02
XTSSWS Power dependence of JTSSWS on temperature — 0.02

Table 2-127. MOSFET level 110 Output Variables

Parameter Description Units Default
WEFF WEFF – none
LEFF LEFF – none
WEFFCV WEFFCV – none
LEFFCV LEFFCV – none
IDS IDS – none
IDEFF IDEFF – none
ISEFF ISEFF – none
IGTOT IGTOT – none
IDSGEN IDSGEN – none
III III – none
IGS IGS – none
IGD IGD – none
IGCS IGCS – none
IGCD IGCD – none
IGBS IGBS – none
IGBD IGBD – none
IGIDL IGIDL – none
IGISL IGISL – none
IJSB IJSB – none

596
 Table 2-127. MOSFET level 110 Output Variables
Parameter Description Units Default
IJDB IJDB – none
ISUB ISUB – none
BETA BETA – none
VTH VTH – none
VDSSAT VDSSAT – none
VFB VFB – none
GM GM – none
GDS GDS – none
GMBS GMBS – none
QGI QGI – none
QDI QDI – none
QSI QSI – none
QBI QBI – none
QG QG – none
QD QD – none
QS QS – none
QB QB – none
CGGI CGGI – none
CGSI CGSI – none
CGDI CGDI – none
CGEI CGEI – none
CDGI CDGI – none
CDDI CDDI – none
CDSI CDSI – none
CDEI CDEI – none
CSGI CSGI – none
CSDI CSDI – none
CSSI CSSI – none
CSEI CSEI – none
CEGI CEGI – none
CEDI CEDI – none
CESI CESI – none
CEEI CEEI – none
CGG CGG – none
CGS CGS – none
CGD CGD – none

597
 Table 2-127. MOSFET level 110 Output Variables
Parameter Description Units Default
CGE CGE – none
CDG CDG – none
CDD CDD – none
CDS CDS – none
CDE CDE – none
CSG CSG – none
CSD CSD – none
CSS CSS – none
CSE CSE – none
CEG CEG – none
CED CED – none
CES CES – none
CEE CEE – none
CGSEXT CGSEXT – none
CGDEXT CGDEXT – none
CGBOV CGBOV – none
CJST CJST – none
CJDT CJDT – none
RSGEO RSGEO – none
RDGEO RDGEO – none
CFGEO CFGEO – none
T_TOTAL_K T_TOTAL_K – none
T_TOTAL_C T_TOTAL_C – none
T_DELTA_SH T_DELTA_SH – none
IGBACC IGBACC – none
IGBINV IGBINV – none
DIDSDVG DIDSDVG – none
DIDSDVS DIDSDVS – none
DIDSDVD DIDSDVD – none
DIGSDVG DIGSDVG – none
DIGSDVS DIGSDVS – none
DIGSDVD DIGSDVD – none
DIGDDVG DIGDDVG – none
DIGDDVS DIGDDVS – none
DIGDDVD DIGDDVD – none
DIIIDVG DIIIDVG – none

598
 Table 2-127. MOSFET level 110 Output Variables
Parameter Description Units Default
DIIIDVS DIIIDVS – none
DIIIDVD DIIIDVD – none
DIGIDLDVG DIGIDLDVG – none
DIGIDLDVS DIGIDLDVS – none
DIGIDLDVD DIGIDLDVD – none
DIGISLDVG DIGISLDVG – none
DIGISLDVS DIGISLDVS – none
DIGISLDVD DIGISLDVD – none
CGT CGT – none
CST CST – none
CDT CDT – none
DIDSDVTH DIDSDVTH – none
DIGSDVTH DIGSDVTH – none
DIGDDVTH DIGDDVTH – none
DIIIDVTH DIIIDVTH – none
DIGIDLDVTH DIGIDLDVTH – none
DIGISLDVTH DIGISLDVTH – none
DITHDVTH DITHDVTH – none
ITH ITH – none
DITHDVG DITHDVG – none
DITHDVS DITHDVS – none
DITHDVD DITHDVD – none
VDSAT Synonym for VDSSAT – none
VDS Drain-source voltage – none
VGS Gate-source voltage – none
VBS Bulk-source voltage – none

2.3.20.14. Level 107 and 108 MOSFET Tables (BSIM CMG versions 107.0.0 and 108.0.0)

Xyce includes the legacy BSIM CMG Common Multi-gate model versions 107 and 108. These models have
been superceded by the level 110 version, but has been retained for backward compatibility with previous
versions of Xyce and older model cards and PDKs. The code in Xyce was generated from the BSIM group’s
Verilog-A input using the default “ifdef” lines provided, and therefore supports only the subset of BSIM
CMG features those defaults enable. Instance and model parameters for the BSIM CMG model are given in
tables 2-128, 2-129, 2-131, and 2-132. Details of the model are documented in the BSIM-CMG technical
report[30], available from the BSIM web site at http://bsim.berkeley.edu/models/bsimcmg/.
Note that the TNOIMOD=1 option of BSIM-CMG 108 is not supported in Xyce, as it uses features of
Verilog-A that are not supported in our Verilog-A compiler. This noise model was added in version 108 and

599
removed in version 109. The TNOIMOD=2 option of BSIM-CMG 108 is the same as the TNOIMOD=1
option of BSIM-CMG 110.

Table 2-128. BSIM-CMG FINFET v107.0.0 Device Instance Parameters

Parameter Description Units Default
ADEJ Drain junction area (BULKMOD=1) – 0
ADEO Drain to substrate overlap area through oxide – 0
ASEJ Source junction area (BULKMOD=1) – 0
ASEO Source to substrate overlap area through oxide – 0
Constant drain-to-source fringe capacitance (All
CDSP – 0
CGEOMOD)
Constant gate-to-drain fringe capacitance
CGDP – 0
(CGEOMOD=1)
Constant gate-to-source fringe capacitance
CGSP – 0
(CGEOMOD=1)
COVD Constant g/d overlap capacitance (CGEOMOD=1) – 0
COVS Constant g/s overlap capacitance (CGEOMOD=1) – 0
D Diameter of the cylinder (GEOMOD=3) – 4e-08
FPITCH Fin pitch – 8e-08
L Designed Gate Length – 3e-08
LRSD Length of the source/drain – 0
Number of fins per finger (real number enables
NFIN – 1
optimization)
NGCON number of gate contact (1 or 2 sided) – 1
NRD Number of source diffusion squares – 0
NRS Number of source diffusion squares – 0
Drain to substrate PN junction perimeter
PDEJ – 0
(BULKMOD=1)
Perimeter of drain to substrate overlap region through
PDEO – 0
oxide
Source to substrate PN junction perimeter
PSEJ – 0
(BULKMOD=1)
Perimeter of source to substrate overlap region
PSEO – 0
through oxide
TFIN Body (Fin) thickness – 1.5e-08

Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters

Parameter Description Units Default
Non-saturation effect parameter for strong inversion
A1 – 0
region

600
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
A11 Temperature dependence of A1 – 0
Non-saturation effect parameter for moderate
A2 – 0
inversion region
A21 Temperature dependence of A2 – 0
ADEJ Drain junction area (BULKMOD=1) – 0
ADEO Drain to substrate overlap area through oxide – 0
AEU – 0
AGIDL pre-exponential coeff. for GIDL in mho – 0
AGISL pre-exponential coeff. for GISL in mho – 6.055e-12
AIGBACC parameter for Igb in accumulation – 0.0136
AIGBACC1 parameter for Igb in accumulation – 0
AIGBINV parameter for Igb in inversion – 0.0111
AIGBINV1 parameter for Igb in inversion – 0
AIGC parameter for Igc in inversion – 0.0136
AIGC1 parameter for Igc in inversion – 0
AIGD parameter for Igd in inversion – 0
AIGD1 parameter for Igd in inversion – 0
AIGEN Thermal Generation Current Parameter – 0
AIGS parameter for Igs in inversion – 0.0136
AIGS1 parameter for Igs in inversion – 0
ALPHA0 first parameter of Iii m/V 0
ALPHA01 Temperature dependence of ALPHA0, m/V/degrees – 0
ALPHA1 L scaling parameter of Iii V−1 0
ALPHA11 Temperature dependence ALPHA1, 1/V/degree – 0
ALPHAII0 first parameter of Iii for IIMOD=2, m/V – 0
ALPHAII01 Temperature dependence of ALPHAII0, m/V/degrees – 0
ALPHAII1 L scaling parameter of Iii for IIMOD=2 V−1 0
ALPHAII11 Temperature dependence of ALPHAII1, 1/V/degrees – 0
AMEXP – 0
AMEXPR – 0
APCLM – 0
APSAT – 0
APSATCV – 0
APTWG – 0
Parameter for Geometric dependence of Tcen on
AQMTCEN – 0
R/TFIN/HFIN

601
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
ARDSW – 0
ARDW – 0
ARSDEND – 0
ARSW – 0
ASEJ Source junction area (BULKMOD=1) – 0
ASEO Source to substrate overlap area through oxide – 0
ASILIEND – 0
ASYMMOD Asymmetric model selector – 0
AT – -0.00156
AUA – 0
AUD – 0
AVSAT – 0
AVSAT1 – 0
AVSATCV – 0
BETA0 Vds dependent parameter of Iii V−1 0
BETAII0 Vds dependent parameter of Iii V−1 0
BETAII1 Vds dependent parameter of Iii – 0
BETAII2 Vds dependent parameter of Iii, V – 0.1
BEU – 1e-07
BG0SUB Band gap of substrate at 300.15K, eV – 1.12
BGIDL exponential coeff. for GIDL V/m 0
BGISL exponential coeff. for GISL V/m 3e+08
BIGBACC parameter for Igb in accumulation – 0.00171
BIGBINV parameter for Igb in inversion – 0.000949
BIGC parameter for Igc in inversion – 0.00171
BIGD parameter for Igd in inversion – 0
BIGEN Thermal Generation Current Parameter – 0
BIGS parameter for Igs in inversion – 0.00171
BMEXP – 1
BMEXPR – 0
BPCLM – 1e-07
BPSAT – 1
BPSATCV – 0
BPTWG – 1e-07
Parameter for Geometric dependence of Tcen on
BQMTCEN – 1.2e-08
R/TFIN/HFIN

602
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
BRDSW – 1e-07
BRDW – 1e-07
BRSW – 1e-07
BUA – 1e-07
BUD – 5e-08
BULKMOD Bulk model – 0
BVD Drain diode breakdown voltage – 0
BVS Source diode breakdown voltage – 10
BVSAT – 1e-07
BVSAT1 – 0
BVSATCV – 0
CAPMOD Accumulation region capacitance model selector – 0
CDSC coupling capacitance between S/D and channel – 0.007
CDSCD drain-bias sensitivity of CDSC – 0.007
CDSCDN1 NFIN dependence of CDSCD – 0
CDSCDN2 NFIN dependence of CDSCD – 100000
Reverse-mode drain-bais sensitivity of CDSC
CDSCDR – 0
(Experimental)
CDSCDRN1 NFIN dependence of CDSCD – 0
CDSCDRN2 NFIN dependence of CDSCD – 0
CDSCN1 NFIN dependence of CDSC – 0
CDSCN2 NFIN dependence of CDSC – 100000
Constant drain-to-source fringe capacitance (All
CDSP – 0
CGEOMOD)
CFD Outer Fringe Cap (drain side) – 0
CFS Outer Fringe Cap (source side) – 2.5e-11
Bias dependent component of Gate to substrate
CGBL – 0
overlap cap
Gate to substrate overlap cap per unit channel length
CGBN – 0
per fin per finger
Gate to substrate overlap cap per unit channel length
CGBO – 0
per finger per NGCON
CGDL – 0
Non LDD region drain-gate overlap capacitance per
CGDO – 0
unit channel width
Constant gate-to-drain fringe capacitance
CGDP – 0
(CGEOMOD=1)
CGEO1SW – 0

603
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
CGEOA Fitting parameter for CGEOMOD=2 – 1
CGEOB Fitting parameter for CGEOMOD=2 – 0
CGEOC Fitting parameter for CGEOMOD=2 – 0
CGEOD Fitting parameter for CGEOMOD=2 – 0
CGEOE Fitting parameter for CGEOMOD=2 – 1
CGEOMOD parasitic capacitance model selector – 0
CGIDL parameter for body-effect of GIDL V3 0
CGISL parameter for body-effect of GISL V3 0.5
CGSL – 0
Non LDD region source-gate overlap capacitance per
CGSO – 0
unit channel width
Constant gate-to-source fringe capacitance
CGSP – 0
(CGEOMOD=1)
Average Channel Charge Weighting Factor,
CHARGEWF – 0
+1:source-side, 0:middle, -1:drain-side
CIGBACC parameter for Igb in accumulation – 0.075
CIGBINV parameter for Igb in inversion – 0.006
CIGC parameter for Igc in inversion – 0.075
CIGD parameter for Igd in inversion – 0
CIGS parameter for Igs in inversion – 0.075
CIT parameter for interface trap – 0
CJD Unit area drain-side junction capacitance at zero bias – 0
CJS Unit area source-side junction capacitance at zero bias – 0.0005
Unit length drain-side sidewall junction capacitance at
CJSWD – 0
zero bias
Unit length drain-side gate sidewall junction
CJSWGD – 0
capacitance at zero bias
Unit length source-side gate sidewall junction
CJSWGS – 0
capacitance at zero bias
Unit length source-side sidewall junction capacitance
CJSWS – 5e-10
at zero bias
CKAPPAB – 0.6
CKAPPAD – 0
CKAPPAS – 0.6
COREMOD Surface potential algorithm – 0
COVD Constant g/d overlap capacitance (CGEOMOD=1) – 0
COVS Constant g/s overlap capacitance (CGEOMOD=1) – 0
CRATIO – 0.5

604
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
CSDESW Coefficient for source/drain to substrate sidewall cap – 0
CTH0 Thermal capacitance – 1e-05
D Diameter of the cylinder (GEOMOD=3) – 4e-08
DELTAPRSD – 0
DELTAVSAT – 1
DELTAVSATCV – 0
DELTAW change of effective width due to shape of fin/cylinder – 0
CV change of effective width due to shape of
DELTAWCV – 0
fin/cylinder
Change in Flatband Voltage;
DELVFBACC – 0
Vfb_accumulation-Vfb_inversion
DELVTRAND Variability in Vth – 0
DEVTYPE – 1
DLBIN Delta L for Binning – 0
DLC Delta L for C-V model – 0
Delta L for C-V model in accumulation region
DLCACC – 0
(CAPMOD=1, BULKMOD=1)
DLCIGD Delta L for Igd model – 0
DLCIGS Delta L for Igs model – 0
DROUT – 1.06
DSUB DIBL exponent coefficient – 1.06
DTEMP Variability in Device Temperature – 0
DVT0 SCE coefficient – 0
SCE exponent coefficient, after binning should be in
DVT1 – 0.6
(0:inf)
Subthreshold Swing exponent coefficient, after
DVT1SS – 0
binning should be in (0:inf)
DVTP0 Coefficient for Drain-Induced Vth Shift (DITS) – 0
DVTP1 DITS exponent coefficient – 0
DVTSHIFT Vth shift handle – 0
EASUB Electron affinity of substrate, eV – 4.05
EF – 1
EGIDL band bending parameter for GIDL V 0
EGISL band bending parameter for GISL V 0.2
EIGBINV parameter for Igb in inversion – 1.1
EM – 4.1e+07
EMOBT – 0

605
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
EOT equivalent oxide thickness in meters – 1e-09
equivalent oxide thickness for accumulation region in
EOTACC m 0
meters
equivalent oxide thickness of the buried oxide (SOI
EOTBOX – 1.4e-07
FinFET) or STI (bulk FinFET) in meters
EPSROX Relative dielectric constant of the gate dielectric – 3.9
EPSRSP Relative dielectric constant of the spacer – 3.9
EPSRSUB Relative dielectric constant of the channel material – 11.9
ESATII Saturation channel E-Field for Iii V/m 1e+07
ETA0 DIBL coefficient – 0.6
ETA0N1 NFIN dependence of ETA0 – 0
ETA0N2 NFIN dependence of ETA0 – 100000
ETA0R Reverse-mode DIBL coefficient (Experimental) – 0
ETAMOB – 2
ETAQM Bulk charge coefficient for Tcen – 0.54
EU – 2.5
FECH End-channel factor, for different orientation/shape – 1
FECHCV CV end-channel factor, for different orientaion/shape – 1
FPITCH Fin pitch – 8e-08
GEOMOD Geometry mode selector – 1
GIDLMOD GIDL/GISL current switcher – 0
HEPI Height of the raised source/drain on top of the fin – 1e-08
HFIN Fin height in meters – 3e-08
IDS0MULT Variability in Drain current for misc. reasons – 1
IGBMOD model selector for Igb – 0
IGCMOD model selector for Igc, Igs, and Igd – 0
IGT Gate Current Temperature Dependence – 2.5
IIMOD Impact ionization model switch – 0
Impact Ionization Temperature Dependence,
IIT – -0.5
IIMOD=1
IJTHDFWD Forward drain diode breakdown limiting current – 0
IJTHDREV Reverse drain diode breakdown limiting current – 0
IJTHSFWD Forward source diode breakdown limiting current – 0.1
IJTHSREV Reverse source diode breakdown limiting current – 0.1
Parameter for Vgs Clamping for inversion region calc.
IMIN – 1e-15
in accumulation

606
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
Bottom drain junction reverse saturation current
JSD – 0
density
Bottom source junction reverse saturation current
JSS – 0.0001
density
Unit length reverse saturation current for sidewall
JSWD – 0
drain junction
Unit length reverse saturation current for gate-edge
JSWGD – 0
sidewall drain junction
Unit length reverse saturation current for gate-edge
JSWGS – 0
sidewall source junction
Unit length reverse saturation current for sidewall
JSWS – 0
source junction
Bottom drain junction trap-assisted saturation current
JTSD – 0
density
Bottom source junction trap-assisted saturation
JTSS – 0
current density
Unit length trap-assisted saturation current for
JTSSWD – 0
sidewall drain junction
Unit length trap-assisted saturation current for
JTSSWGD – 0
gate-edge sidewall drain junction
Unit length trap-assisted saturation current for
JTSSWGS – 0
gate-edge sidewall source junction
Unit length trap-assisted saturation current for
JTSSWS – 0
sidewall source junction
JTWEFF Trap assisted tunneling current width dependence – 0
K0 Lateral NUD voltage parameter, V – 0
Temperature dependence of lateral NUD voltage
K01 – 0
parameter, V/K
Correction factor for strong inversion, used in Mnud,
K0SI – 1
after binnig should be from (0:inf)
K0SI1 Temperature dependence of K0SI, 1/K – 0
K1 Body effect coefficient for sub-threshold region – 0
K11 Temperature dependence of K1 – 0
K1RSCE K1 for reverse short channel effect calculation – 0
K1SAT Correction factor for K1 in saturation (high Vds) – 0
K1SAT1 Temperature dependence of K1SAT1 – 0
K1SI Correction factor for strong inversion, used in Mob – 0
K1SI1 Temperature dependence of K1SI, 1/K – 0
KSATIV – 1
KT1 Vth Temperature Coefficient (V) – 0

607
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
KT1L Vth Temperature L Coefficient (m-V) – 0
L Designed Gate Length – 3e-08
LA1 – 0
LA11 – 0
LA2 – 0
LA21 – 0
LAGIDL – 0
LAGISL – 0
LAIGBACC – 0
LAIGBACC1 – 0
LAIGBINV – 0
LAIGBINV1 – 0
LAIGC – 0
LAIGC1 – 0
LAIGD – 0
LAIGD1 – 0
LAIGEN – 0
LAIGS – 0
LAIGS1 – 0
LALPHA0 – 0
LALPHA1 – 0
LALPHAII0 – 0
LALPHAII1 – 0
LAT – 0
LBETA0 – 0
LBETAII0 – 0
LBETAII1 – 0
LBETAII2 – 0
LBGIDL – 0
LBGISL – 0
LBIGBACC – 0
LBIGBINV – 0
LBIGC – 0
LBIGD – 0
LBIGEN – 0
LBIGS – 0

608
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
LCDSC – 0
LCDSCD – 0
LCDSCDR – 0
LCFD – 0
LCFS – 0
LCGBL – 0
LCGDL – 0
LCGIDL – 0
LCGISL – 0
LCGSL – 0
LCIGBACC – 0
LCIGBINV – 0
LCIGC – 0
LCIGD – 0
LCIGS – 0
LCIT – 0
LCKAPPAB – 0
LCKAPPAD – 0
LCKAPPAS – 0
LCOVD – 0
LCOVS – 0
LDELTAVSAT – 0
LDELTAVSATCV – 0
LDROUT – 0
LDSUB – 0
LDVT0 – 0
LDVT1 – 0
LDVT1SS – 0
LDVTB – 0
LDVTSHIFT – 0
LEGIDL – 0
LEGISL – 0
LEIGBINV – 0
LEMOBT – 0
LESATII – 0
LETA0 – 0

609
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
LETA0R – 0
LETAMOB – 0
LEU – 0
LIGT – 0
LII Channel length dependent parameter of Iii Vm 5e-10
LIIT – 0
LINT Length reduction parameter (dopant diffusion effect) – 0
LINTIGEN Lint for Thermal Generation Current – 0
LINTNOI – 0
LK0 – 0
LK01 – 0
LK0SI – 0
LK0SI1 – 0
LK1 – 0
LK11 – 0
LK1RSCE – 0
LK1SAT – 0
LK1SAT1 – 0
LK1SI – 0
LK1SI1 – 0
LKSATIV – 0
LKT1 – 0
LL Length reduction parameter (dopant diffusion effect) – 0
LLC Length reduction parameter (dopant diffusion effect) – 0
LLII – 0
LLN Length reduction parameter (dopant diffusion effect) – 1
LLPE0 – 0
LLPEB – 0
LMEXP – 0
LMEXPR – 0
LNBODY – 0
LNGATE – 0
LNIGBACC – 0
LNIGBINV – 0
LNTGEN – 0
LNTOX – 0

610
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
LPA – 1
LPCLM – 0
LPCLMCV – 0
LPCLMG – 0
LPDIBL1 – 0
LPDIBL1R – 0
LPDIBL2 – 0
LPE0 Equivalent length of pocket region at zero bias – 5e-09
LPGIDL – 0
LPGISL – 0
LPHIBE – 0
LPHIG – 0
LPHIN – 0
LPIGCD – 0
LPOXEDGE – 0
LPRT – 0
LPRWGD – 0
LPRWGS – 0
LPSAT – 0
LPSATCV – 0
LPTWG – 0
LPTWGR – 0
LPTWGT – 0
LPVAG – 0
LQMFACTOR – 0
LQMTCENCV – 0
LQMTCENCVA – 0
LQMTCENIV – 0
LRDSW – 0
LRDW – 0
LRSD Length of the source/drain – 0
LRSW – 0
LSII0 – 0
LSII1 – 0
LSII2 – 0
LSIID – 0

611
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
LSP – 0
LSTTHETASAT – 0
LTGIDL – 0
LTII – 0
LTSS – 0
LU0 – 0
LUA – 0
LUA1 – 0
LUC – 0
LUC1 – 0
LUCS – 0
LUCSTE – 0
LUD – 0
LUD1 – 0
LUP – 0
LUTE – 0
LUTL – 0
LVSAT – 0
LVSAT1 – 0
LVSAT1R – 0
LVSATCV – 0
LWR – 0
LXRCRG1 – 0
LXRCRG2 – 0
MEXP – 4
MEXPR – 0
MJD Drain bottom junction capacitance grading coefficient – 0
Drain bottom two-step second junction capacitance
MJD2 – 0
grading coefficient
Source bottom junction capacitance grading
MJS – 0.5
coefficient
Source bottom two-step second junction capacitance
MJS2 – 0.125
grading coefficient
Drain sidewall junction capacitance grading
MJSWD – 0
coefficient
Drain sidewall two-step second junction capacitance
MJSWD2 – 0
grading coefficient

612
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
Drain-side gate sidewall junction capacitance grading
MJSWGD – 0
coefficient
MJSWGD2 Drain-side gate sidewall two-step – 0
Source-side gate sidewall junction capacitance
MJSWGS – 0
grading coefficient
MJSWGS2 Source-side gate sidewall two-step – 0
Source sidewall junction capacitance grading
MJSWS – 0.33
coefficient
Source sidewall two-step second junction capacitance
MJSWS2 – 0.083
grading coefficient
NA1 – 0
NA11 – 0
NA2 – 0
NA21 – 0
NAGIDL – 0
NAGISL – 0
NAIGBACC – 0
NAIGBACC1 – 0
NAIGBINV – 0
NAIGBINV1 – 0
NAIGC – 0
NAIGC1 – 0
NAIGD – 0
NAIGD1 – 0
NAIGEN – 0
NAIGS – 0
NAIGS1 – 0
NALPHA0 – 0
NALPHA1 – 0
NALPHAII0 – 0
NALPHAII1 – 0
NAT – 0
NBETA0 – 0
NBETAII0 – 0
NBETAII1 – 0
NBETAII2 – 0
NBGIDL – 0

613
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
NBGISL – 0
NBIGBACC – 0
NBIGBINV – 0
NBIGC – 0
NBIGD – 0
NBIGEN – 0
NBIGS – 0
NBODY channel (body) doping – 1e+22
NBODYN1 NFIN dependence of channel (body) doping – 0
NBODYN2 NFIN dependence of channel (body) doping – 100000
NC0SUB Conduction band density of states, m-3 – 2.86e+25
NCDSC – 0
NCDSCD – 0
NCDSCDR – 0
NCFD – 0
NCFS – 0
NCGBL – 0
NCGDL – 0
NCGIDL – 0
NCGISL – 0
NCGSL – 0
NCIGBACC – 0
NCIGBINV – 0
NCIGC – 0
NCIGD – 0
NCIGS – 0
NCIT – 0
NCKAPPAB – 0
NCKAPPAD – 0
NCKAPPAS – 0
NCOVD – 0
NCOVS – 0
NDELTAVSAT – 0
NDELTAVSATCV – 0
NDROUT – 0
NDSUB – 0

614
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
NDVT0 – 0
NDVT1 – 0
NDVT1SS – 0
NDVTB – 0
NDVTSHIFT – 0
NEGIDL – 0
NEGISL – 0
NEIGBINV – 0
NEMOBT – 0
NESATII – 0
NETA0 – 0
NETA0R – 0
NETAMOB – 0
NEU – 0
NF Number of fingers – 1
Number of fins per finger (real number enables
NFIN – 1
optimization)
Parameter for Poly Gate Doping, for metal gate please
NGATE – 0
set NGATE = 0
NGCON number of gate contact (1 or 2 sided) – 1
NI0SUB Intrinsic carrier constant at 300.15K, m-3 – 1.1e+16
NIGBACC parameter for Igb in accumulation – 1
NIGBINV parameter for Igb in inversion – 3
NIGT – 0
NIIT – 0
NJD Drain junction emission coefficient – 0
NJS Source junction emission coefficient – 1
NJTS Non-ideality factor for JTSS – 20
NJTSD Non-ideality factor for JTSD – 0
NJTSSW Non-ideality factor for JTSSWS – 20
NJTSSWD Non-ideality factor for JTSSWD – 0
NJTSSWG Non-ideality factor for JTSSWGS – 20
NJTSSWGD Non-ideality factor for JTSSWGD – 0
NK0 – 0
NK01 – 0
NK0SI – 0

615
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
NK0SI1 – 0
NK1 – 0
NK11 – 0
NK1RSCE – 0
NK1SAT – 0
NK1SAT1 – 0
NK1SI – 0
NK1SI1 – 0
NKSATIV – 0
NKT1 – 0
NLII – 0
NLPE0 – 0
NLPEB – 0
NMEXP – 0
NMEXPR – 0
NNBODY – 0
NNGATE – 0
NNIGBACC – 0
NNIGBINV – 0
NNTGEN – 0
NNTOX – 0
NOIA – 6.25e+39
NOIB – 3.125e+24
NOIC – 8.75e+07
NPCLM – 0
NPCLMCV – 0
NPCLMG – 0
NPDIBL1 – 0
NPDIBL1R – 0
NPDIBL2 – 0
NPGIDL – 0
NPGISL – 0
NPHIBE – 0
NPHIG – 0
NPHIN – 0
NPIGCD – 0

616
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
NPOXEDGE – 0
NPRT – 0
NPRWGD – 0
NPRWGS – 0
NPSAT – 0
NPSATCV – 0
NPTWG – 0
NPTWGR – 0
NPTWGT – 0
NPVAG – 0
NQMFACTOR – 0
NQMTCENCV – 0
NQMTCENCVA – 0
NQMTCENIV – 0
NQSMOD – 0
NRD Number of source diffusion squares – 0
NRDSW – 0
NRDW – 0
NRS Number of source diffusion squares – 0
NRSW – 0
NSD Source/drain active doping concentration in m-3 – 2e+26
NSDE Source/drain active doping concentration at Leff edge – 2e+25
Number of segments for NQSMOD=3 (3,5 and 10
NSEG – 4
supported)
NSII0 – 0
NSII1 – 0
NSII2 – 0
NSIID – 0
NSTTHETASAT – 0
NTGEN Thermal Generation Current Parameter – 1
NTGIDL – 0
NTII – 0
NTNOI – 1
NTOX Exponent for Tox ratio – 1
NTSS – 0
NU0 – 0

617
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
NUA – 0
NUA1 – 0
NUC – 0
NUC1 – 0
NUCS – 0
NUCSTE – 0
NUD – 0
NUD1 – 0
NUP – 0
NUTE – 0
NUTL – 0
NVSAT – 0
NVSAT1 – 0
NVSAT1R – 0
NVSATCV – 0
NWR – 0
NXRCRG1 – 0
NXRCRG2 – 0
PA1 – 0
PA11 – 0
PA2 – 0
PA21 – 0
PAGIDL – 0
PAGISL – 0
PAIGBACC – 0
PAIGBACC1 – 0
PAIGBINV – 0
PAIGBINV1 – 0
PAIGC – 0
PAIGC1 – 0
PAIGD – 0
PAIGD1 – 0
PAIGEN – 0
PAIGS – 0
PAIGS1 – 0
PALPHA0 – 0

618
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
PALPHA1 – 0
PALPHAII0 – 0
PALPHAII1 – 0
PAT – 0
PBD Drain-side bulk junction built-in potential – 0
PBETA0 – 0
PBETAII0 – 0
PBETAII1 – 0
PBETAII2 – 0
PBGIDL – 0
PBGISL – 0
PBIGBACC – 0
PBIGBINV – 0
PBIGC – 0
PBIGD – 0
PBIGEN – 0
PBIGS – 0
PBS Source-side bulk junction built-in potential – 1
Built-in potential for Drain-side sidewall junction
PBSWD – 0
capacitance
Built-in potential for Drain-side gate sidewall junction
PBSWGD – 0
capacitance
Built-in potential for Source-side gate sidewall
PBSWGS – 0
junction capacitance
Built-in potential for Source-side sidewall junction
PBSWS – 1
capacitance
PCDSC – 0
PCDSCD – 0
PCDSCDR – 0
PCFD – 0
PCFS – 0
PCGBL – 0
PCGDL – 0
PCGIDL – 0
PCGISL – 0
PCGSL – 0
PCIGBACC – 0

619
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
PCIGBINV – 0
PCIGC – 0
PCIGD – 0
PCIGS – 0
PCIT – 0
PCKAPPAB – 0
PCKAPPAD – 0
PCKAPPAS – 0
PCLM – 0.013
PCLMCV CLM parameter for Short Channel CV – 0
PCLMG – 0
PCOVD – 0
PCOVS – 0
Drain to substrate PN junction perimeter
PDEJ – 0
(BULKMOD=1)
PDELTAVSAT – 0
PDELTAVSATCV – 0
Perimeter of drain to substrate overlap region through
PDEO – 0
oxide
PDIBL1 DIBL Output Conductance parameter - forward mode – 1.3
PDIBL1R DIBL Output Conductance parameter - reverse mode – 0
PDIBL2 DIBL Output Conductance parameter – 0.0002
PDROUT – 0
PDSUB – 0
PDVT0 – 0
PDVT1 – 0
PDVT1SS – 0
PDVTB – 0
PDVTSHIFT – 0
PEGIDL – 0
PEGISL – 0
PEIGBINV – 0
PEMOBT – 0
PESATII – 0
PETA0 – 0
PETA0R – 0

620
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
PETAMOB – 0
PEU – 0
PGIDL parameter for body-bias effect on GIDL – 0
PGISL parameter for body-bias effect on GISL – 1
Body effect voltage parameter, V, after binnig should
PHIBE – 0.7
be from [0.2:1.2]
PHIG Gate workfunction, eV – 4.61
PHIGL Length dependence of Gate workfunction, eV/m – 0
PHIGN1 NFIN dependence of Gate workfunction – 0
PHIGN2 NFIN dependence of Gate workfunction – 100000
Nonuniform vertical doping effect on surface
PHIN – 0.05
potential, V
PIGCD parameter for Igc partition – 1
PIGT – 0
PIIT – 0
PK0 – 0
PK01 – 0
PK0SI – 0
PK0SI1 – 0
PK1 – 0
PK11 – 0
PK1RSCE – 0
PK1SAT – 0
PK1SAT1 – 0
PK1SI – 0
PK1SI1 – 0
PKSATIV – 0
PKT1 – 0
PLII – 0
PLPE0 – 0
PLPEB – 0
PMEXP – 0
PMEXPR – 0
PNBODY – 0
PNGATE – 0
PNIGBACC – 0

621
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
PNIGBINV – 0
PNTGEN – 0
PNTOX – 0
POXEDGE Factor for the gate edge Tox – 1
PPCLM – 0
PPCLMCV – 0
PPCLMG – 0
PPDIBL1 – 0
PPDIBL1R – 0
PPDIBL2 – 0
PPGIDL – 0
PPGISL – 0
PPHIBE – 0
PPHIG – 0
PPHIN – 0
PPIGCD – 0
PPOXEDGE – 0
PPRT – 0
PPRWGD – 0
PPRWGS – 0
PPSAT – 0
PPSATCV – 0
PPTWG – 0
PPTWGR – 0
PPTWGT – 0
PPVAG – 0
PQM Slope of normalized Tcen in inversion – 0.66
PQMACC Slope of normalized Tcen in accumulation – 0.66
PQMFACTOR – 0
PQMTCENCV – 0
PQMTCENCVA – 0
PQMTCENIV – 0
PRDDR Drain side quasi-saturation parameter – 0
PRDSW – 0
PRDW – 0
PRSDEND – 0

622
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
PRSDR Source side quasi-saturation parameter – 1
PRSW – 0
PRT – 0.001
PRWGD Gate bias dependence of drain extension resistance V−1 0
PRWGS Gate bias dependence of source extension resistance V−1 0
Velocity saturation exponent, after binnig should be
PSAT – 2
from [2.0:inf)
PSATCV Velocity saturation exponent for C-V – 0
Source to substrate PN junction perimeter
PSEJ – 0
(BULKMOD=1)
Perimeter of source to substrate overlap region
PSEO – 0
through oxide
PSII0 – 0
PSII1 – 0
PSII2 – 0
PSIID – 0
PSTTHETASAT – 0
PTGIDL – 0
PTII – 0
PTSS – 0
PTWG Gmsat degradation parameter - forward mode – 0
PTWGR Gmsat degradation parameter - reverse mode – 0
PTWGT – 0.004
PU0 – 0
PUA – 0
PUA1 – 0
PUC – 0
PUC1 – 0
PUCS – 0
PUCSTE – 0
PUD – 0
PUD1 – 0
PUP – 0
PUTE – 0
PUTL – 0
PVAG – 1
PVSAT – 0

623
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
PVSAT1 – 0
PVSAT1R – 0
PVSATCV – 0
PWR – 0
PXRCRG1 – 0
PXRCRG2 – 0
Knee-Point for Tcen in inversion (Charge normalized
QM0 – 0.001
to Cox)
Knee-Point for Tcen in accumulation (Charge
QM0ACC – 0.001
normalized to Cox)
QMFACTOR Prefactor + switch for QM Vth correction – 0
Prefactor + switch for QM Width and Toxeff
QMTCENCV – 0
correction for CV
Prefactor + switch for QM Width and Toxeff
QMTCENCVA – 0
correction for CV (accumulation region)
QMTCENIV Prefactor + switch for QM Width correction for IV – 0
RDDR Drain side drift resistance parameter - forward mode – 0
RDDRR Drain side drift resistance parameter - reverse mode – 0
RDSMOD Resistance model selector – 0
RDSW – 100
RDSWMIN – 0
RDW – 50
RDWMIN – 0
Gate electrode resistor and ge node switcher — NOT
RGATEMOD – 0
USED IN XYCE
RGEOA Fitting parameter for RGEOMOD=1 – 1
RGEOB Fitting parameter for RGEOMOD=1 – 0
RGEOC Fitting parameter for RGEOMOD=1 – 0
RGEOD Fitting parameter for RGEOMOD=1 – 0
RGEOE Fitting parameter for RGEOMOD=1 – 0
RGEOMOD Bias independent parasitic resistance model selector – 0
RGEXT Effective gate electrode external resistance – 0
RGFIN Effective gate electrode per finger per fin resistance – 0.001
RHOC – 1e-12
RHORSD – 1
RSDR Source side drift resistance parameter - forward mode – 0
RSDRR Source side drift resistance parameter - reverse mode – 0

624
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
RSHD Drain-side sheet resistance – 0
RSHS Source-side sheet resistance – 0
RSW – 50
RSWMIN – 0
RTH0 Thermal resistance – 0.01
SDTERM – 0
Self heating and T node switcher — NOT USED IN
SHMOD – 0
XYCE
SII0 Vgs dependent parameter of Iii V−1 0.5
SII1 1st Vgs dependent parameter of Iii V−1 0.1
SII2 2nd Vgs dependent parameter of Iii – 0
SIID 3rd Vds dependent parameter of Iii V−1 0
SJD Constant for drain-side two-step second junction – 0
SJS Constant for source-side two-step second junction – 0
Constant for drain-side sidewall two-step second
SJSWD – 0
junction
Constant for source-side gate sidewall two-step
SJSWGD – 0
second junction
Constant for source-side gate sidewall two-step
SJSWGS – 0
second junction
Constant for source-side sidewall two-step second
SJSWS – 0
junction
TBGASUB Bandgap Temperature Coefficient (eV / degrees) – 0.000702
TBGBSUB Bandgap Temperature Coefficient (degrees) – 1108
TCJ Temperature coefficient for CJS/CJD – 0
TCJSW Temperature coefficient for CJSWS/CJSWD – 0
TCJSWG Temperature coefficient for CJSWGS/CJSWGD – 0
TETA0 Temperature dependence of DIBL coefficient, 1/K – 0
Temperature dependence of Reverse-mode DIBL
TETA0R – 0
coefficient, 1/K
TFIN Body (Fin) thickness – 1.5e-08
TGATE Gate height on top of the hard mask – 3e-08
TGIDL GIDL/GISL Temperature Dependence – -0.003
Impact Ionization Temperature Dependence,
TII – 0
IIMOD=2
TMASK Height of hard mask on top of the fin – 3e-08
TMEXP – 0
TMEXPR – 0

625
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
TNJTS Temperature coefficient for NJTS – 0
TNJTSD Temperature coefficient for NJTSD – 0
TNJTSSW Temperature coefficient for NJTSSW – 0
TNJTSSWD NTemperature coefficient for NJTSSWD – 0
TNJTSSWG Temperature coefficient for NJTSSWG – 0
TNJTSSWGD Temperature coefficient for NJTSSWGD – 0
TNOM Temperature at which the model is extracted (degrees) – 27
oxide thickness for gate current model in meters,
TOXG m 0
Introduced in BSIM-CMG106.1.0
TOXP physical oxide thickness in meters – 1.2e-09
TOXREF Target tox value [m] – 1.2e-09
TPB Temperature coefficient for PBS/PBD – 0
TPBSW Temperature coefficient for PBSWS/PBSWD – 0
TPBSWG Temperature coefficient for PBSWGS/PBSWGD – 0
TRDDR – 0
TRSDR – 0
Thickness of the silicide on top of the raised
TSILI – 1e-08
source/drain
TSS SSwing Temperature Coefficient (/ degrees) – 0
U0 – 0.03
U0MULT Variability in carrier mobility – 1
U0N1 NFIN dependence of U0 – 0
U0N2 NFIN dependence of U0 – 100000
UA – 0.3
UA1 – 0.001032
Body effect for mobility degradation parameter -
UC – 0
BULKMOD=1
UC1 – 5.6e-11
UCS – 1
UCSTE – -0.004775
UD – 0
UD1 – 0
UP – 0
UTE – 0
UTL – -0.0015
VSAT – 85000

626
 Table 2-129. BSIM-CMG FINFET v107.0.0 Device Model Parameters
Parameter Description Units Default
Velocity Saturation parameter for I_on degradation -
VSAT1 – 0
forward mode
VSAT1N1 NFIN dependence of VSAT1 – 0
VSAT1N2 NFIN dependence of VSAT1 – 0
Velocity Saturation parameter for I_on degradation -
VSAT1R – 0
reverse mode
VSAT1RN1 NFIN dependence of VSAT1R – 0
VSAT1RN2 NFIN dependence of VSAT1R – 0
VSATCV Velocity Saturation parameter for CV – 0
VSATN1 NFIN dependence of VSAT – 0
VSATN2 NFIN dependence of VSAT – 100000
Bottom drain junction trap-assisted current voltage
VTSD – 0
dependent parameter
Bottom source junction trap-assisted current voltage
VTSS – 10
dependent parameter
Unit length trap-assisted current voltage dependent
VTSSWD – 0
parameter for sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGD – 0
parameter for gate-edge sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGS – 10
parameter for gate-edge sidewall source junction
Unit length trap-assisted current voltage dependent
VTSSWS – 10
parameter for sidewall source junction
WR – 1
WTH0 Width dependence coefficient for Rth and Cth – 0
XJBVD Fitting parameter for drain diode breakdown current – 0
XJBVS Fitting parameter for source diode breakdown current – 1
XL L offset for channel length due to mask/etch effect – 0
XRCRG1 – 12
XRCRG2 – 1
XTID Drain junction current temperature exponent – 0
XTIS Source junction current temperature exponent – 3
XTSD Power dependence of JTSD on temperature – 0
XTSS Power dependence of JTSS on temperature – 0.02
XTSSWD Power dependence of JTSSWD on temperature – 0
XTSSWGD Power dependence of JTSSWGD on temperature – 0
XTSSWGS Power dependence of JTSSWGS on temperature – 0.02
XTSSWS Power dependence of JTSSWS on temperature – 0.02

627
 Table 2-130. MOSFET level 107 Output Variables
Parameter Description Units Default
WEFF WEFF – none
LEFF LEFF – none
WEFFCV WEFFCV – none
LEFFCV LEFFCV – none
IDS IDS – none
IDEFF IDEFF – none
ISEFF ISEFF – none
IGTOT IGTOT – none
IDSGEN IDSGEN – none
III III – none
IGIDL IGIDL – none
IGISL IGISL – none
IJSB IJSB – none
IJDB IJDB – none
ISUB ISUB – none
BETA BETA – none
VTH VTH – none
VDSSAT VDSSAT – none
VFB VFB – none
GM GM – none
GDS GDS – none
GMBS GMBS – none
QGI QGI – none
QDI QDI – none
QSI QSI – none
QBI QBI – none
QG QG – none
QD QD – none
QS QS – none
QB QB – none
CGGI CGGI – none
CGSI CGSI – none
CGDI CGDI – none
CGEI CGEI – none
CDGI CDGI – none
CDDI CDDI – none

628
 Table 2-130. MOSFET level 107 Output Variables
Parameter Description Units Default
CDSI CDSI – none
CDEI CDEI – none
CSGI CSGI – none
CSDI CSDI – none
CSSI CSSI – none
CSEI CSEI – none
CEGI CEGI – none
CEDI CEDI – none
CESI CESI – none
CEEI CEEI – none
CGG CGG – none
CGS CGS – none
CGD CGD – none
CGE CGE – none
CDG CDG – none
CDD CDD – none
CDS CDS – none
CDE CDE – none
CSG CSG – none
CSD CSD – none
CSS CSS – none
CSE CSE – none
CEG CEG – none
CED CED – none
CES CES – none
CEE CEE – none
CGSEXT CGSEXT – none
CGDEXT CGDEXT – none
CGBOV CGBOV – none
CJST CJST – none
CJDT CJDT – none
RSGEO RSGEO – none
RDGEO RDGEO – none
CFGEO CFGEO – none

629
 Table 2-131. BSIM-CMG FINFET v108.0.0 Device Instance Parameters
Parameter Description Units Default
ADEJ Drain junction area (BULKMOD=1) – 0
ADEO Drain to substrate overlap area through oxide – 0
ASEJ Source junction area (BULKMOD=1) – 0
ASEO Source to substrate overlap area through oxide – 0
Constant drain-to-source fringe capacitance (All
CDSP – 0
CGEOMOD)
Constant gate-to-drain fringe capacitance
CGDP – 0
(CGEOMOD=1)
Constant gate-to-source fringe capacitance
CGSP – 0
(CGEOMOD=1)
COVD Constant g/d overlap capacitance (CGEOMOD=1) – 0
COVS Constant g/s overlap capacitance (CGEOMOD=1) – 0
D Diameter of the cylinder (GEOMOD=3) – 4e-08
FPITCH Fin pitch – 8e-08
L Designed Gate Length – 3e-08
LRSD Length of the source/drain – 0
M multiplicity factor — 1
NF Number of fingers – 1
Number of fins per finger (real number enables
NFIN – 1
optimization)
NGCON number of gate contact (1 or 2 sided) – 1
NRD Number of source diffusion squares – 0
NRS Number of source diffusion squares – 0
Drain to substrate PN junction perimeter
PDEJ – 0
(BULKMOD=1)
Perimeter of drain to substrate overlap region through
PDEO – 0
oxide
Source to substrate PN junction perimeter
PSEJ – 0
(BULKMOD=1)
Perimeter of source to substrate overlap region
PSEO – 0
through oxide
TFIN Body (Fin) thickness – 1.5e-08

Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters

Parameter Description Units Default
Non-saturation effect parameter for strong inversion
A1 – 0
region

630
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
A11 Temperature dependence of A1 – 0
Non-saturation effect parameter for moderate
A2 – 0
inversion region
A21 Temperature dependence of A2 – 0
ADEJ Drain junction area (BULKMOD=1) – 0
ADEO Drain to substrate overlap area through oxide – 0
AEU – 0
AEUR – 0
AGIDL pre-exponential coeff. for GIDL in mho – 0
AGISL pre-exponential coeff. for GISL in mho – 6.055e-12
AIGBACC parameter for Igb in accumulation – 0.0136
AIGBACC1 parameter for Igb in accumulation – 0
AIGBINV parameter for Igb in inversion – 0.0111
AIGBINV1 parameter for Igb in inversion – 0
AIGC parameter for Igc in inversion – 0.0136
AIGC1 parameter for Igc in inversion – 0
AIGD parameter for Igd in inversion – 0
AIGD1 parameter for Igd in inversion – 0
AIGEN Thermal Generation Current Parameter – 0
AIGS parameter for Igs in inversion – 0.0136
AIGS1 parameter for Igs in inversion – 0
ALPHA0 first parameter of Iii, m/V – 0
ALPHA01 Temperature dependence of ALPHA0, m/V/degrees – 0
ALPHA1 L scaling parameter of Iii, 1/V – 0
ALPHA11 Temperature dependence ALPHA1, 1/V/degree – 0
ALPHAII0 first parameter of Iii for IIMOD=2, m/V – 0
ALPHAII01 Temperature dependence of ALPHAII0, m/V/degrees – 0
ALPHAII1 L scaling parameter of Iii for IIMOD=2, 1/V – 0
ALPHAII11 Temperature dependence of ALPHAII1, 1/V/degrees – 0
AMEXP – 0
AMEXPR – 0
APCLM – 0
APCLMR – 0
APSAT – 0
APSATCV – 0
APTWG – 0

631
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
Parameter for Geometric dependence of Tcen on
AQMTCEN – 0
R/TFIN/HFIN
ARDSW – 0
ARDW – 0
ARSDEND – 0
ARSW – 0
ASEJ Source junction area (BULKMOD=1) – 0
ASEO Source to substrate overlap area through oxide – 0
ASILIEND – 0
ASYMMOD Asymmetric model selector – 0
AT – -0.00156
ATCV – 0
ATR – 0
AUA – 0
AUAR – 0
AUD – 0
AUDR – 0
AVSAT – 0
AVSAT1 – 0
AVSATCV – 0
BETA0 Vds dependent parameter of Iii, 1/V – 0
BETAII0 Vds dependent parameter of Iii, 1/V – 0
BETAII1 Vds dependent parameter of Iii – 0
BETAII2 Vds dependent parameter of Iii, V – 0.1
BEU – 1e-07
BEUR – 0
BG0SUB Band gap of substrate at 300.15K, eV – 1.12
BGIDL exponential coeff. for GIDL in V/m – 0
BGISL exponential coeff. for GISL in V/m – 3e+08
BIGBACC parameter for Igb in accumulation – 0.00171
BIGBINV parameter for Igb in inversion – 0.000949
BIGC parameter for Igc in inversion – 0.00171
BIGD parameter for Igd in inversion – 0
BIGEN Thermal Generation Current Parameter – 0
BIGS parameter for Igs in inversion – 0.00171
BMEXP – 1

632
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
BMEXPR – 0
BPCLM – 1e-07
BPCLMR – 0
BPSAT – 1
BPSATCV – 0
BPTWG – 1e-07
Parameter for Geometric dependence of Tcen on
BQMTCEN – 1.2e-08
R/TFIN/HFIN
BRDSW – 1e-07
BRDW – 1e-07
BRSW – 1e-07
BUA – 1e-07
BUAR – 0
BUD – 5e-08
BUDR – 0
BULKMOD Bulk model – 0
BVD Drain diode breakdown voltage – 0
BVS Source diode breakdown voltage – 10
BVSAT – 1e-07
BVSAT1 – 0
BVSATCV – 0
CAPMOD Accumulation capacitance selector – 0
CDSC coupling capacitance between S/D and channel – 0.007
CDSCD drain-bias sensitivity of CDSC – 0.007
CDSCDN1 NFIN dependence of CDSCD – 0
CDSCDN2 NFIN dependence of CDSCD – 100000
Reverse-mode drain-bias sensitivity of CDSC
CDSCDR – 0
(Experimental)
CDSCDRN1 NFIN dependence of CDSCD – 0
CDSCDRN2 NFIN dependence of CDSCD – 0
CDSCN1 NFIN dependence of CDSC – 0
CDSCN2 NFIN dependence of CDSC – 100000
Constant drain-to-source fringe capacitance (All
CDSP – 0
CGEOMOD)
CFD Outer Fringe Cap (drain side) – 0
CFS Outer Fringe Cap (source side) – 2.5e-11

633
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
Bias dependent component of Gate to substrate
CGBL – 0
overlap cap per unit channel length per fin per finger
Gate to substrate overlap cap per unit channel length
CGBN – 0
per fin per finger
Gate to substrate overlap cap per unit channel length
CGBO – 0
per finger per NGCON
CGDL – 0
Non LDD region drain-gate overlap capacitance per
CGDO – 0
unit channel width
Constant gate-to-drain fringe capacitance
CGDP – 0
(CGEOMOD=1)
CGEO1SW – 0
CGEOA Fitting parameter for CGEOMOD=2 – 1
CGEOB Fitting parameter for CGEOMOD=2 – 0
CGEOC Fitting parameter for CGEOMOD=2 – 0
CGEOD Fitting parameter for CGEOMOD=2 – 0
CGEOE Fitting parameter for CGEOMOD=2 – 1
Geometry dependent parasitic capacitance model
CGEOMOD – 0
selector
CGIDL parameter for body-effect of GIDL in V**3 – 0
CGISL parameter for body-effect of GISL in V**3 – 0.5
CGSL – 0
Non LDD region source-gate overlap capacitance per
CGSO – 0
unit channel width
Constant gate-to-source fringe capacitance
CGSP – 0
(CGEOMOD=1)
Average Channel Charge Weighting Factor,
CHARGEWF – 0
+1:source-side, 0:middle, -1:drain-side
CIGBACC parameter for Igb in accumulation – 0.075
CIGBINV parameter for Igb in inversion – 0.006
CIGC parameter for Igc in inversion – 0.075
CIGD parameter for Igd in inversion – 0
CIGS parameter for Igs in inversion – 0.075
CIT parameter for interface trap – 0
parameter for interface trap in reverse mode for
CITR – 0
asymmetric model
CJD Unit area drain-side junction capacitance at zero bias – 0
CJS Unit area source-side junction capacitance at zero bias – 0.0005

634
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
Unit length drain-side sidewall junction capacitance at
CJSWD – 0
zero bias
Unit length drain-side gate sidewall junction
CJSWGD – 0
capacitance at zero bias
Unit length source-side gate sidewall junction
CJSWGS – 0
capacitance at zero bias
Unit length source-side sidewall junction capacitance
CJSWS – 5e-10
at zero bias
CKAPPAB – 0.6
CKAPPAD – 0
CKAPPAS – 0.6
COREMOD Surface potential algorithm – 0
COVD Constant g/d overlap capacitance (CGEOMOD=1) – 0
COVS Constant g/s overlap capacitance (CGEOMOD=1) – 0
CRATIO – 0.5
CSDESW Coefficient for source/drain to substrate sidewall cap – 0
CTH0 Thermal capacitance – 1e-05
D Diameter of the cylinder (GEOMOD=3) – 4e-08
DELTAPRSD – 0
DELTAVSAT – 1
DELTAVSATCV – 0
DELTAW change of effective width due to shape of fin/cylinder – 0
CV change of effective width due to shape of
DELTAWCV – 0
fin/cylinder
Change in Flatband Voltage;
DELVFBACC – 0
Vfb_accumulation-Vfb_inversion
DELVTRAND Variability in Vth – 0
DEVTYPE – 1
DLBIN Delta L for Binning – 0
DLC Delta L for C-V model – 0
Delta L for C-V model in accumulation region
DLCACC – 0
(CAPMOD=1, BULKMOD=1)
DLCIGD Delta L for Igd model – 0
DLCIGS Delta L for Igs model – 0
DROUT – 1.06
DSUB DIBL exponent coefficient – 1.06
DTEMP Variability in Device Temperature – 0
DVT0 SCE coefficient – 0

635
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
SCE exponent coefficient, after binning should be in
DVT1 – 0.6
(0:inf)
Subthreshold Swing exponent coefficient, after
DVT1SS – 0
binning should be in (0:inf)
DVTP0 Coefficient for Drain-Induced Vth Shift (DITS) – 0
DVTP1 DITS exponent coefficient – 0
DVTSHIFT Vth shift handle – 0
DVTSHIFTR Vth shift handle for asymmetric mode – 0
EASUB Electron affinity of substrate, eV – 4.05
EF Flicker Noise frequency exponent – 1
EGIDL band bending parameter for GIDL in V – 0
EGISL band bending parameter for GISL in V – 0.2
EIGBINV parameter for Igb in inversion – 1.1
EM – 4.1e+07
EMOBT – 0
EOT equivalent oxide thickness in meters – 1e-09
equivalent oxide thickness for accumulation region in
EOTACC – 0
meters
equivalent oxide thickness of the buried oxide (SOI
EOTBOX – 1.4e-07
FinFET) or STI (bulk FinFET) in meters
EPSROX Relative dielectric constant of the gate dielectric – 3.9
EPSRSP Relative dielectric constant of the spacer – 3.9
EPSRSUB Relative dielectric constant of the channel material – 11.9
ESATII Saturation channel E-Field for Iii, V/m – 1e+07
ETA0 DIBL coefficient – 0.6
ETA0N1 NFIN dependence of ETA0 – 0
ETA0N2 NFIN dependence of ETA0 – 100000
ETA0R Reverse-mode DIBL coefficient (Experimental) – 0
ETAMOB – 2
ETAQM Bulk charge coefficient for Tcen – 0.54
EU – 2.5
EUR – 0
FECH End-channel factor, for different orientation/shape – 1
FECHCV CV end-channel factor, for different orientation/shape – 1
FPITCH Fin pitch – 8e-08
GEOMOD Geometry mode selector – 1
GIDLMOD GIDL/GISL current switcher – 0

636
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
HEPI Height of the raised source/drain on top of the fin – 1e-08
HFIN Fin height in meters – 3e-08
IDS0MULT Variability in Drain current for misc. reasons – 1
IGBMOD Model selector for Igb – 0
IGCMOD Model selector for Igc, Igs, and Igd – 0
IGT Gate Current Temperature Dependence – 2.5
IIMOD Impact ionization model switch – 0
Impact Ionization Temperature Dependence,
IIT – -0.5
IIMOD=1
IJTHDFWD Forward drain diode breakdown limiting current – 0
IJTHDREV Reverse drain diode breakdown limiting current – 0
IJTHSFWD Forward source diode breakdown limiting current – 0.1
IJTHSREV Reverse source diode breakdown limiting current – 0.1
Parameter for Vgs Clamping for inversion region calc.
IMIN – 1e-15
in accumulation
Bottom drain junction reverse saturation current
JSD – 0
density
Bottom source junction reverse saturation current
JSS – 0.0001
density
Unit length reverse saturation current for sidewall
JSWD – 0
drain junction
Unit length reverse saturation current for gate-edge
JSWGD – 0
sidewall drain junction
Unit length reverse saturation current for gate-edge
JSWGS – 0
sidewall source junction
Unit length reverse saturation current for sidewall
JSWS – 0
source junction
Bottom drain junction trap-assisted saturation current
JTSD – 0
density
Bottom source junction trap-assisted saturation
JTSS – 0
current density
Unit length trap-assisted saturation current for
JTSSWD – 0
sidewall drain junction
Unit length trap-assisted saturation current for
JTSSWGD – 0
gate-edge sidewall drain junction
Unit length trap-assisted saturation current for
JTSSWGS – 0
gate-edge sidewall source junction
Unit length trap-assisted saturation current for
JTSSWS – 0
sidewall source junction
JTWEFF Trap assisted tunneling current width dependence – 0

637
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
K0 Lateral NUD voltage parameter, V – 0
Temperature dependence of lateral NUD voltage
K01 – 0
parameter, V/K
Correction factor for strong inversion, used in Mnud,
K0SI – 1
after binnig should be from (0:inf)
K0SI1 Temperature dependence of K0SI, 1/K – 0
K1 Body effect coefficient for sub-threshold region – 0
K11 Temperature dependence of K1 – 0
K1RSCE K1 for reverse short channel effect calculation – 0
K1SAT Correction factor for K1 in saturation (high Vds) – 0
K1SAT1 Temperature dependence of K1SAT1 – 0
K1SI Correction factor for strong inversion, used in Mob – 0
K1SI1 Temperature dependence of K1SI, 1/K – 0
KSATIV – 1
KSATIVR KSATIV in asymmetric mode – 0
KT1 Vth Temperature Coefficient (V) – 0
KT1L Vth Temperature L Coefficient (m-V) – 0
L Designed Gate Length – 3e-08
LA1 – 0
LA11 – 0
LA2 – 0
LA21 – 0
LAGIDL – 0
LAGISL – 0
LAIGBACC – 0
LAIGBACC1 – 0
LAIGBINV – 0
LAIGBINV1 – 0
LAIGC – 0
LAIGC1 – 0
LAIGD – 0
LAIGD1 – 0
LAIGEN – 0
LAIGS – 0
LAIGS1 – 0
LALPHA0 – 0

638
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
LALPHA1 – 0
LALPHAII0 – 0
LALPHAII1 – 0
LAT – 0
LATCV – 0
LATR – 0
LBETA0 – 0
LBETAII0 – 0
LBETAII1 – 0
LBETAII2 – 0
LBGIDL – 0
LBGISL – 0
LBIGBACC – 0
LBIGBINV – 0
LBIGC – 0
LBIGD – 0
LBIGEN – 0
LBIGS – 0
LCDSC – 0
LCDSCD – 0
LCDSCDR – 0
LCFD – 0
LCFS – 0
LCGBL – 0
LCGDL – 0
LCGIDL – 0
LCGISL – 0
LCGSL – 0
LCIGBACC – 0
LCIGBINV – 0
LCIGC – 0
LCIGD – 0
LCIGS – 0
LCIT – 0
LCITR – 0
LCKAPPAB – 0

639
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
LCKAPPAD – 0
LCKAPPAS – 0
LCOVD – 0
LCOVS – 0
LDELTAVSAT – 0
LDELTAVSATCV – 0
LDROUT – 0
LDSUB – 0
LDVT0 – 0
LDVT1 – 0
LDVT1SS – 0
LDVTB – 0
LDVTSHIFT – 0
LDVTSHIFTR – 0
LEGIDL – 0
LEGISL – 0
LEIGBINV – 0
LEMOBT – 0
LESATII – 0
LETA0 – 0
LETA0R – 0
LETAMOB – 0
LEU – 0
LEUR – 0
LIGT – 0
LII Channel length dependent parameter of Iii, V-m – 5e-10
LIIT – 0
LINT Length reduction parameter (dopant diffusion effect) – 0
LINTIGEN Lint for Thermal Generation Current – 0
LINTNOI – 0
LK0 – 0
LK01 – 0
LK0SI – 0
LK0SI1 – 0
LK1 – 0
LK11 – 0

640
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
LK1RSCE – 0
LK1SAT – 0
LK1SAT1 – 0
LK1SI – 0
LK1SI1 – 0
LKSATIV – 0
LKSATIVR – 0
LKT1 – 0
LL Length reduction parameter (dopant diffusion effect) – 0
LLC Length reduction parameter (dopant diffusion effect) – 0
LLII – 0
LLN Length reduction parameter (dopant diffusion effect) – 1
LLPE0 – 0
LLPEB – 0
LMAX Maximum length for which this model should be used. – 100
LMEXP – 0
LMEXPR – 0
LMIN Minimum length for which this model should be used. – 0
LNBODY – 0
LNGATE – 0
LNIGBACC – 0
LNIGBINV – 0
LNTGEN – 0
LNTOX – 0
LPA – 1
LPAR – 0
LPCLM – 0
LPCLMCV – 0
LPCLMG – 0
LPCLMR – 0
LPDIBL1 – 0
LPDIBL1R – 0
LPDIBL2 – 0
LPDIBL2R – 0
LPE0 Equivalent length of pocket region at zero bias – 5e-09
LPGIDL – 0

641
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
LPGISL – 0
LPHIBE – 0
LPHIG – 0
LPHIN – 0
LPIGCD – 0
LPOXEDGE – 0
LPRT – 0
LPRWGD – 0
LPRWGS – 0
LPSAT – 0
LPSATCV – 0
LPTWG – 0
LPTWGR – 0
LPTWGT – 0
LPVAG – 0
LQMFACTOR – 0
LQMTCENCV – 0
LQMTCENCVA – 0
LQMTCENIV – 0
LRDSW – 0
LRDW – 0
LRSD Length of the source/drain – 0
LRSW – 0
LSII0 – 0
LSII1 – 0
LSII2 – 0
LSIID – 0
LSP – 0
LSTTHETASAT – 0
LTGIDL – 0
LTII – 0
LTSS – 0
LU0 – 0
LU0R – 0
LUA – 0
LUA1 – 0

642
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
LUA1R – 0
LUAR – 0
LUC – 0
LUC1 – 0
LUC1R – 0
LUCR – 0
LUCS – 0
LUCSTE – 0
LUD – 0
LUD1 – 0
LUD1R – 0
LUDR – 0
LUP – 0
LUTE – 0
LUTER – 0
LUTL – 0
LUTLR – 0
LVSAT – 0
LVSAT1 – 0
LVSAT1R – 0
LVSATCV – 0
LVSATR – 0
LWR – 0
LXRCRG1 – 0
LXRCRG2 – 0
MEXP – 4
MEXPR – 0
MJD Drain bottom junction capacitance grading coefficient – 0
Drain bottom two-step second junction capacitance
MJD2 – 0
grading coefficient
Source bottom junction capacitance grading
MJS – 0.5
coefficient
Source bottom two-step second junction capacitance
MJS2 – 0.125
grading coefficient
Drain sidewall junction capacitance grading
MJSWD – 0
coefficient

643
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
Drain sidewall two-step second junction capacitance
MJSWD2 – 0
grading coefficient
Drain-side gate sidewall junction capacitance grading
MJSWGD – 0
coefficient
Drain-side gate sidewall two-step second junction
MJSWGD2 – 0
capacitance grading coefficient
Source-side gate sidewall junction capacitance
MJSWGS – 0
grading coefficient
Source-side gate sidewall two-step second junction
MJSWGS2 – 0
capacitance grading coefficient
Source sidewall junction capacitance grading
MJSWS – 0.33
coefficient
Source sidewall two-step second junction capacitance
MJSWS2 – 0.083
grading coefficient
NA1 – 0
NA11 – 0
NA2 – 0
NA21 – 0
NAGIDL – 0
NAGISL – 0
NAIGBACC – 0
NAIGBACC1 – 0
NAIGBINV – 0
NAIGBINV1 – 0
NAIGC – 0
NAIGC1 – 0
NAIGD – 0
NAIGD1 – 0
NAIGEN – 0
NAIGS – 0
NAIGS1 – 0
NALPHA0 – 0
NALPHA1 – 0
NALPHAII0 – 0
NALPHAII1 – 0
NAT – 0
NATCV – 0
NATR – 0

644
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
NBETA0 – 0
NBETAII0 – 0
NBETAII1 – 0
NBETAII2 – 0
NBGIDL – 0
NBGISL – 0
NBIGBACC – 0
NBIGBINV – 0
NBIGC – 0
NBIGD – 0
NBIGEN – 0
NBIGS – 0
NBODY channel (body) doping – 1e+22
NBODYN1 NFIN dependence of channel (body) doping – 0
NBODYN2 NFIN dependence of channel (body) doping – 100000
NC0SUB Conduction band density of states, m-3 – 2.86e+25
NCDSC – 0
NCDSCD – 0
NCDSCDR – 0
NCFD – 0
NCFS – 0
NCGBL – 0
NCGDL – 0
NCGIDL – 0
NCGISL – 0
NCGSL – 0
NCIGBACC – 0
NCIGBINV – 0
NCIGC – 0
NCIGD – 0
NCIGS – 0
NCIT – 0
NCITR – 0
NCKAPPAB – 0
NCKAPPAD – 0
NCKAPPAS – 0

645
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
NCOVD – 0
NCOVS – 0
NDELTAVSAT – 0
NDELTAVSATCV – 0
NDROUT – 0
NDSUB – 0
NDVT0 – 0
NDVT1 – 0
NDVT1SS – 0
NDVTB – 0
NDVTSHIFT – 0
NDVTSHIFTR – 0
NEGIDL – 0
NEGISL – 0
NEIGBINV – 0
NEMOBT – 0
NESATII – 0
NETA0 – 0
NETA0R – 0
NETAMOB – 0
NEU – 0
NEUR – 0
NF Number of fingers – 1
Number of fins per finger (real number enables
NFIN – 1
optimization)
NFINMAX Maximum NFIN for which this model should be used. – 100
NFINMIN Minimum NFIN for which this model should be used. – 0
Parameter for Poly Gate Doping, for metal gate please
NGATE – 0
set NGATE = 0
NGCON number of gate contact (1 or 2 sided) – 1
NI0SUB Intrinsic carrier constant at 300.15K, m-3 – 1.1e+16
NIGBACC parameter for Igb in accumulation – 1
NIGBINV parameter for Igb in inversion – 3
NIGT – 0
NIIT – 0
NJD Drain junction emission coefficient – 0

646
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
NJS Source junction emission coefficient – 1
NJTS Non-ideality factor for JTSS – 20
NJTSD Non-ideality factor for JTSD – 0
NJTSSW Non-ideality factor for JTSSWS – 20
NJTSSWD Non-ideality factor for JTSSWD – 0
NJTSSWG Non-ideality factor for JTSSWGS – 20
NJTSSWGD Non-ideality factor for JTSSWGD – 0
NK0 – 0
NK01 – 0
NK0SI – 0
NK0SI1 – 0
NK1 – 0
NK11 – 0
NK1RSCE – 0
NK1SAT – 0
NK1SAT1 – 0
NK1SI – 0
NK1SI1 – 0
NKSATIV – 0
NKSATIVR – 0
NKT1 – 0
NLII – 0
NLPE0 – 0
NLPEB – 0
NMEXP – 0
NMEXPR – 0
NNBODY – 0
NNGATE – 0
NNIGBACC – 0
NNIGBINV – 0
NNTGEN – 0
NNTOX – 0
NOIA – 6.25e+39
NOIB – 3.125e+24
NOIC – 8.75e+07
NPCLM – 0

647
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
NPCLMCV – 0
NPCLMG – 0
NPCLMR – 0
NPDIBL1 – 0
NPDIBL1R – 0
NPDIBL2 – 0
NPDIBL2R – 0
NPGIDL – 0
NPGISL – 0
NPHIBE – 0
NPHIG – 0
NPHIN – 0
NPIGCD – 0
NPOXEDGE – 0
NPRT – 0
NPRWGD – 0
NPRWGS – 0
NPSAT – 0
NPSATCV – 0
NPTWG – 0
NPTWGR – 0
NPTWGT – 0
NPVAG – 0
NQMFACTOR – 0
NQMTCENCV – 0
NQMTCENCVA – 0
NQMTCENIV – 0
NQSMOD – 0
NRD Number of source diffusion squares – 0
NRDSW – 0
NRDW – 0
NRS Number of source diffusion squares – 0
NRSW – 0
NSD Source/drain active doping concentration in m-3 – 2e+26
NSDE Source/drain active doping concentration at Leff edge – 2e+25

648
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
Number of segments for NQSMOD=3 (3,5 and 10
NSEG – 4
supported)
NSII0 – 0
NSII1 – 0
NSII2 – 0
NSIID – 0
NSTTHETASAT – 0
NTGEN Thermal Generation Current Parameter – 1
NTGIDL – 0
NTII – 0
NTNOI – 1
NTOX Exponent for Tox ratio – 1
NTSS – 0
NU0 – 0
NU0R – 0
NUA – 0
NUA1 – 0
NUA1R – 0
NUAR – 0
NUC – 0
NUC1 – 0
NUC1R – 0
NUCR – 0
NUCS – 0
NUCSTE – 0
NUD – 0
NUD1 – 0
NUD1R – 0
NUDR – 0
NUP – 0
NUTE – 0
NUTER – 0
NUTL – 0
NUTLR – 0
NVSAT – 0
NVSAT1 – 0

649
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
NVSAT1R – 0
NVSATCV – 0
NVSATR – 0
Subthreshold Swing factor multiplied by Vtm. If
NVTM – 0
defined by user, will overwrite nVtm in the code
NWR – 0
NXRCRG1 – 0
NXRCRG2 – 0
PA1 – 0
PA11 – 0
PA2 – 0
PA21 – 0
PAGIDL – 0
PAGISL – 0
PAIGBACC – 0
PAIGBACC1 – 0
PAIGBINV – 0
PAIGBINV1 – 0
PAIGC – 0
PAIGC1 – 0
PAIGD – 0
PAIGD1 – 0
PAIGEN – 0
PAIGS – 0
PAIGS1 – 0
PALPHA0 – 0
PALPHA1 – 0
PALPHAII0 – 0
PALPHAII1 – 0
PAT – 0
PATCV – 0
PATR – 0
PBD Drain-side bulk junction built-in potential – 0
PBETA0 – 0
PBETAII0 – 0
PBETAII1 – 0

650
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
PBETAII2 – 0
PBGIDL – 0
PBGISL – 0
PBIGBACC – 0
PBIGBINV – 0
PBIGC – 0
PBIGD – 0
PBIGEN – 0
PBIGS – 0
PBS Source-side bulk junction built-in potential – 1
Built-in potential for Drain-side sidewall junction
PBSWD – 0
capacitance
Built-in potential for Drain-side gate sidewall junction
PBSWGD – 0
capacitance
Built-in potential for Source-side gate sidewall
PBSWGS – 0
junction capacitance
Built-in potential for Source-side sidewall junction
PBSWS – 1
capacitance
PCDSC – 0
PCDSCD – 0
PCDSCDR – 0
PCFD – 0
PCFS – 0
PCGBL – 0
PCGDL – 0
PCGIDL – 0
PCGISL – 0
PCGSL – 0
PCIGBACC – 0
PCIGBINV – 0
PCIGC – 0
PCIGD – 0
PCIGS – 0
PCIT – 0
PCITR – 0
PCKAPPAB – 0
PCKAPPAD – 0

651
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
PCKAPPAS – 0
PCLM – 0.013
PCLMCV CLM parameter for Short Channel CV – 0
PCLMG – 0
PCLMR Reverse Model PCLM parameter – 0
PCOVD – 0
PCOVS – 0
Drain to substrate PN junction perimeter
PDEJ – 0
(BULKMOD=1)
PDELTAVSAT – 0
PDELTAVSATCV – 0
Perimeter of drain to substrate overlap region through
PDEO – 0
oxide
PDIBL1 DIBL Output Conductance parameter - forward mode – 1.3
PDIBL1R DIBL Output Conductance parameter - reverse mode – 0
PDIBL2 DIBL Output Conductance parameter – 0.0002
PDIBL2R DIBL Output Conductance parameter - reverse mode – 0
PDROUT – 0
PDSUB – 0
PDVT0 – 0
PDVT1 – 0
PDVT1SS – 0
PDVTB – 0
PDVTSHIFT – 0
PDVTSHIFTR – 0
PEGIDL – 0
PEGISL – 0
PEIGBINV – 0
PEMOBT – 0
PESATII – 0
PETA0 – 0
PETA0R – 0
PETAMOB – 0
PEU – 0
PEUR – 0
PGIDL parameter for body-bias effect on GIDL – 0

652
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
PGISL parameter for body-bias effect on GISL – 1
Body effect voltage parameter, V, after binnig should
PHIBE – 0.7
be from [0.2:1.2]
PHIG Gate workfunction, eV – 4.61
PHIGL Length dependence of Gate workfunction, eV/m – 0
PHIGN1 NFIN dependence of Gate workfunction – 0
PHIGN2 NFIN dependence of Gate workfunction – 100000
Nonuniform vertical doping effect on surface
PHIN – 0.05
potential, V
PIGCD parameter for Igc partition – 1
PIGT – 0
PIIT – 0
PK0 – 0
PK01 – 0
PK0SI – 0
PK0SI1 – 0
PK1 – 0
PK11 – 0
PK1RSCE – 0
PK1SAT – 0
PK1SAT1 – 0
PK1SI – 0
PK1SI1 – 0
PKSATIV – 0
PKSATIVR – 0
PKT1 – 0
PLII – 0
PLPE0 – 0
PLPEB – 0
PMEXP – 0
PMEXPR – 0
PNBODY – 0
PNGATE – 0
PNIGBACC – 0
PNIGBINV – 0
PNTGEN – 0

653
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
PNTOX – 0
POXEDGE Factor for the gate edge Tox – 1
PPCLM – 0
PPCLMCV – 0
PPCLMG – 0
PPCLMR – 0
PPDIBL1 – 0
PPDIBL1R – 0
PPDIBL2 – 0
PPDIBL2R – 0
PPGIDL – 0
PPGISL – 0
PPHIBE – 0
PPHIG – 0
PPHIN – 0
PPIGCD – 0
PPOXEDGE – 0
PPRT – 0
PPRWGD – 0
PPRWGS – 0
PPSAT – 0
PPSATCV – 0
PPTWG – 0
PPTWGR – 0
PPTWGT – 0
PPVAG – 0
PQM Slope of normalized Tcen in inversion – 0.66
PQMACC Slope of normalized Tcen in accumulation – 0.66
PQMFACTOR – 0
PQMTCENCV – 0
PQMTCENCVA – 0
PQMTCENIV – 0
PRDDR Drain side quasi-saturation parameter – 0
PRDSW – 0
PRDW – 0
PRSDEND – 0

654
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
PRSDR Source side quasi-saturation parameter – 1
PRSW – 0
PRT – 0.001
Gate bias dependence of drain extension resistance,
PRWGD – 0
Units:1/V
Gate bias dependence of source extension resistance,
PRWGS – 0
Units:1/V
Velocity saturation exponent, after binnig should be
PSAT – 2
from [2.0:inf)
PSATCV Velocity saturation exponent for C-V – 0
Source to substrate PN junction perimeter
PSEJ – 0
(BULKMOD=1)
Perimeter of source to substrate overlap region
PSEO – 0
through oxide
PSII0 – 0
PSII1 – 0
PSII2 – 0
PSIID – 0
PSTTHETASAT – 0
PTGIDL – 0
PTII – 0
PTSS – 0
PTWG Gmsat degradation parameter - forward mode – 0
PTWGR Gmsat degradation parameter - reverse mode – 0
PTWGT – 0.004
PU0 – 0
PU0R – 0
PUA – 0
PUA1 – 0
PUA1R – 0
PUAR – 0
PUC – 0
PUC1 – 0
PUC1R – 0
PUCR – 0
PUCS – 0
PUCSTE – 0

655
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
PUD – 0
PUD1 – 0
PUD1R – 0
PUDR – 0
PUP – 0
PUTE – 0
PUTER – 0
PUTL – 0
PUTLR – 0
PVAG – 1
PVSAT – 0
PVSAT1 – 0
PVSAT1R – 0
PVSATCV – 0
PVSATR – 0
PWR – 0
PXRCRG1 – 0
PXRCRG2 – 0
Knee-Point for Tcen in inversion (Charge normalized
QM0 – 0.001
to Cox)
Knee-Point for Tcen in accumulation (Charge
QM0ACC – 0.001
normalized to Cox)
QMFACTOR Prefactor + switch for QM Vth correction – 0
Prefactor + switch for QM Width and Toxeff
QMTCENCV – 0
correction for CV
Prefactor + switch for QM Width and Toxeff
QMTCENCVA – 0
correction for CV (accumulation region)
QMTCENIV Prefactor + switch for QM Width correction for IV – 0
RDDR Drain side drift resistance parameter - forward mode – 0
RDDRR Drain side drift resistance parameter - reverse mode – 0
RDSMOD Resistance model selector – 0
RDSW – 100
RDSWMIN – 0
RDW – 50
RDWMIN – 0
RGATEMOD Gate electrode resistance on/off seector – 0
RGEOA Fitting parameter for RGEOMOD=1 – 1

656
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
RGEOB Fitting parameter for RGEOMOD=1 – 0
RGEOC Fitting parameter for RGEOMOD=1 – 0
RGEOD Fitting parameter for RGEOMOD=1 – 0
RGEOE Fitting parameter for RGEOMOD=1 – 0
RGEOMOD Geometry-dependent source/drain resistance – 0
RGEXT Effective gate electrode external resistance – 0
RGFIN Effective gate electrode per finger per fin resistance – 0.001
RHOC – 1e-12
RHORSD – 1
RNOIA Thermal noise coefficient – 0.577
RNOIB Thermal noise coefficient – 0.37
RNOIC Thermal noise coefficient for TNOIMOD=2 – 0.395
RSDR Source side drift resistance parameter - forward mode – 0
RSDRR Source side drift resistance parameter - reverse mode – 0
RSHD Drain-side sheet resistance – 0
RSHS Source-side sheet resistance – 0
RSW – 50
RSWMIN – 0
RTH0 Thermal resistance – 0.01
SCALEN – 100000
SDTERM – 0
Self heating and T node switcher — NOT USED IN
SHMOD – 0
XYCE
SII0 Vgs dependent parameter of Iii, 1/V – 0.5
SII1 1st Vgs dependent parameter of Iii, 1/V – 0.1
SII2 2nd Vgs dependent parameter of Iii – 0
SIID 3rd Vds dependent parameter of Iii, 1/V – 0
SJD Constant for drain-side two-step second junction – 0
SJS Constant for source-side two-step second junction – 0
Constant for drain-side sidewall two-step second
SJSWD – 0
junction
Constant for source-side gate sidewall two-step
SJSWGD – 0
second junction
Constant for source-side gate sidewall two-step
SJSWGS – 0
second junction
Constant for source-side sidewall two-step second
SJSWS – 0
junction

657
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
TBGASUB Bandgap Temperature Coefficient (eV / degrees) – 0.000702
TBGBSUB Bandgap Temperature Coefficient (degrees) – 1108
TCJ Temperature coefficient for CJS/CJD – 0
TCJSW Temperature coefficient for CJSWS/CJSWD – 0
TCJSWG Temperature coefficient for CJSWGS/CJSWGD – 0
TEMPMOD – 0
TETA0 Temperature dependence of DIBL coefficient, 1/K – 0
Temperature dependence of Reverse-mode DIBL
TETA0R – 0
coefficient, 1/K
TFIN Body (Fin) thickness – 1.5e-08
TGATE Gate height on top of the hard mask – 3e-08
TGIDL GIDL/GISL Temperature Dependence – -0.003
DIBL length dependence. If defined by user, will
THETADIBL – 0
overwrite Theta_DIBL in the code
Vth roll-off length dependence. If defined by user, will
THETASCE – 0
overwrite Theta_SCE in the code
Subthreshold Swing length dependence. If defined by
THETASW – 0
user, will overwrite Theta_SW in the code
Impact Ionization Temperature Dependence,
TII – 0
IIMOD=2
TMASK Height of hard mask on top of the fin – 3e-08
TMEXP – 0
TMEXPR – 0
TNJTS Temperature coefficient for NJTS – 0
TNJTSD Temperature coefficient for NJTSD – 0
TNJTSSW Temperature coefficient for NJTSSW – 0
TNJTSSWD NTemperature coefficient for NJTSSWD – 0
TNJTSSWG Temperature coefficient for NJTSSWG – 0
TNJTSSWGD Temperature coefficient for NJTSSWGD – 0
TNOIA Thermal noise parameter – 1.5
TNOIB Thermal noise parameter – 3.5
TNOIC Thermal noise parameter for TNOIMOD=2 – 3.5
0: charge based, 1: holistic thermal noise based on
TNOIMOD – 0
BSIM4 noise model
TNOM Temperature at which the model is extracted (degrees) – 27
oxide thickness for gate current model in meters,
TOXG – 0
Introduced in BSIM-CMG106.1.0
TOXP physical oxide thickness in meters – 1.2e-09

658
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
TOXREF Target tox value [m] – 1.2e-09
TPB Temperature coefficient for PBS/PBD – 0
TPBSW Temperature coefficient for PBSWS/PBSWD – 0
TPBSWG Temperature coefficient for PBSWGS/PBSWGD – 0
TRDDR – 0
TRSDR – 0
Thickness of the silicide on top of the raised
TSILI – 1e-08
source/drain
TSS SSwing Temperature Coefficient (/ degrees) – 0
TYPE – 0
U0 – 0.03
U0MULT Variability in carrier mobility – 1
U0N1 NFIN dependence of U0 – 0
U0N1R – 0
U0N2 NFIN dependence of U0 – 100000
U0N2R – 0
U0R – 0
UA – 0.3
UA1 – 0.001032
UA1R – 0
UAR – 0
Body effect for mobility degradation parameter -
UC – 0
BULKMOD=1
UC1 – 5.6e-11
UC1R – 0
UCR – 0
UCS – 1
UCSTE – -0.004775
UD – 0
UD1 – 0
UD1R – 0
UDR – 0
UP – 0
UPR – 0
UTE – 0
UTER – 0

659
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
UTL – -0.0015
UTLR – 0
VSAT – 85000
Velocity Saturation parameter for I_on degradation -
VSAT1 – 0
forward mode
VSAT1N1 NFIN dependence of VSAT1 – 0
VSAT1N2 NFIN dependence of VSAT1 – 0
Velocity Saturation parameter for I_on degradation -
VSAT1R – 0
reverse mode
VSAT1RN1 NFIN dependence of VSAT1R – 0
VSAT1RN2 NFIN dependence of VSAT1R – 0
VSATCV Velocity Saturation parameter for CV – 0
VSATN1 NFIN dependence of VSAT – 0
VSATN2 NFIN dependence of VSAT – 100000
VSATR – 0
VSATRN1 NFIN dependence of VSATR – 0
VSATRN2 NFIN dependence of VSATR – 0
Bottom drain junction trap-assisted current voltage
VTSD – 0
dependent parameter
Bottom source junction trap-assisted current voltage
VTSS – 10
dependent parameter
Unit length trap-assisted current voltage dependent
VTSSWD – 0
parameter for sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGD – 0
parameter for gate-edge sidewall drain junction
Unit length trap-assisted current voltage dependent
VTSSWGS – 10
parameter for gate-edge sidewall source junction
Unit length trap-assisted current voltage dependent
VTSSWS – 10
parameter for sidewall source junction
WR – 1
WTH0 Width dependence coefficient for Rth and Cth – 0
XJBVD Fitting parameter for drain diode breakdown current – 0
XJBVS Fitting parameter for source diode breakdown current – 1
XL L offset for channel length due to mask/etch effect – 0
XRCRG1 – 12
XRCRG2 – 1
XTID Drain junction current temperature exponent – 0
XTIS Source junction current temperature exponent – 3

660
 Table 2-132. BSIM-CMG FINFET v108.0.0 Device Model Parameters
Parameter Description Units Default
XTSD Power dependence of JTSD on temperature – 0
XTSS Power dependence of JTSS on temperature – 0.02
XTSSWD Power dependence of JTSSWD on temperature – 0
XTSSWGD Power dependence of JTSSWGD on temperature – 0
XTSSWGS Power dependence of JTSSWGS on temperature – 0.02
XTSSWS Power dependence of JTSSWS on temperature – 0.02

Table 2-133. MOSFET level 108 Output Variables

Parameter Description Units Default
WEFF WEFF – none
LEFF LEFF – none
WEFFCV WEFFCV – none
LEFFCV LEFFCV – none
IDS IDS – none
IDEFF IDEFF – none
ISEFF ISEFF – none
IGTOT IGTOT – none
IDSGEN IDSGEN – none
III III – none
IGIDL IGIDL – none
IGISL IGISL – none
IJSB IJSB – none
IJDB IJDB – none
ISUB ISUB – none
BETA BETA – none
VTH VTH – none
VDSSAT VDSSAT – none
VFB VFB – none
GM GM – none
GDS GDS – none
GMBS GMBS – none
QGI QGI – none
QDI QDI – none
QSI QSI – none
QBI QBI – none

661
 Table 2-133. MOSFET level 108 Output Variables
Parameter Description Units Default
QG QG – none
QD QD – none
QS QS – none
QB QB – none
CGGI CGGI – none
CGSI CGSI – none
CGDI CGDI – none
CGEI CGEI – none
CDGI CDGI – none
CDDI CDDI – none
CDSI CDSI – none
CDEI CDEI – none
CSGI CSGI – none
CSDI CSDI – none
CSSI CSSI – none
CSEI CSEI – none
CEGI CEGI – none
CEDI CEDI – none
CESI CESI – none
CEEI CEEI – none
CGG CGG – none
CGS CGS – none
CGD CGD – none
CGE CGE – none
CDG CDG – none
CDD CDD – none
CDS CDS – none
CDE CDE – none
CSG CSG – none
CSD CSD – none
CSS CSS – none
CSE CSE – none
CEG CEG – none
CED CED – none
CES CES – none
CEE CEE – none

662
 Table 2-133. MOSFET level 108 Output Variables
Parameter Description Units Default
CGSEXT CGSEXT – none
CGDEXT CGDEXT – none
CGBOV CGBOV – none
CJST CJST – none
CJDT CJDT – none
RSGEO RSGEO – none
RDGEO RDGEO – none
CFGEO CFGEO – none
TDEVICE TDEVICE – none

663
2.3.20.15. Levels 2000 and 2001 MOSFET Tables (MVS version 2.0.0)

Xyce includes the MIT Virtual Source (MVS) MOSFET model version 2.0.0 in both ETSOI and HEMT
variants. The code in Xyce was generated from the MIT Verilog-A input. Model parameters for the MVS
model are given in 2-134 and 2-135. The MVS model does not have instance parameters. Details of the
model are documented MVS Nanotransistor Model 2.0.0 manual, available from the NEEDS web site at
https://nanohub.org/publications/74/1.
NOTE: Unlike all other MOSFET models in Xyce, the MVS model takes only 3 nodes, the drain, gate and
source. It takes no substrate node.

Table 2-134. MVS ETSOI 2.0.0 Device Model Parameters

Parameter Description Units Default
B – 6.8e-09
BETA – 1.55
CINS – 0.0317
DELTA – 0.12
DLG – 1.05e-08
DQM0 – 4.6e-09
ENERGY_DIFF_VOLT – 0.153
EPS – 13.6
KSEE – 0.1
LGDR – 8e-08
ML – 0.89
MT – 0.19
MU_EFF – 1
N0 – 1.35
ND – 0
NU – 0.7
RS0 – 0.00016
THETA – 2.5
TJUN – 300
TYPE – 1
VERSION – 2
W – 1e-06

Table 2-135. MVS HEMT 2.0.0 Device Model Parameters

Parameter Description Units Default
B – 6.8e-09

664
 Table 2-135. MVS HEMT 2.0.0 Device Model Parameters
Parameter Description Units Default
BETA – 1.55
CINS – 0.0317
DELTA – 0.12
DLG – 1.05e-08
DQM0 – 4.6e-09
ENERGY_DIFF_VOLT – 0.153
EPS – 13.6
KSEE – 0.1
LGDR – 8e-08
MEFF – 0.041
MU_EFF – 1
N0 – 1.35
NACC – 2.25e+16
ND – 0
NP_MASS – 9
RC0 – 0.00016
THETA – 2.5
TJUN – 300
TYPE – 1
VERSION – 2
W – 1e-06

665
2.3.20.16. Level 2002 MOSFET Tables (MVSG_CMC version 1.1.0)

Xyce includes the MIT Virtual Source GaN HEMT High-Voltage (MVSG_CMC) MOSFET model version
1.1.0. The code in Xyce was generated from the MIT Verilog-A input. Model parameters for the MVS
model are given in 2-136 and 2-137, and its output variables in 2-138. More information about this model
may be obtained from the CMC standard models page at https://si2.org/standard-models.

Table 2-136. MVSG-HV HEMT MODEL Device Instance Parameters

Parameter Description Units Default
DTEMP Device temperature offset from ambient K 0
L Effective gate length m 2.5e-07
NGF Number of Fingers — 2
W Width per Finger m 0.00018

Table 2-137. MVSG-HV HEMT MODEL Device Model Parameters

Parameter Description Units Default
AF Flicker noise exponent — 2
ALPHA Weak to strong inversion transition factor — 3.5
ALPHAFP1 FP weak to strong inversion transition factor — 0.01
ALPHAFP2 FP weak to strong inversion transition factor — 0.01
ALPHAFP3 FP weak to strong inversion transition factor — 0.01
ALPHAFP4 FP weak to strong inversion transition factor — 0.01
ALPHAGD G-D high injection smoothing parameter — 10
ALPHAGS G-S high injection smoothing parameter — 10
ALPHARD DAR weak to strong inversion transition factor — 3.5
ALPHARS SAR weak to strong inversion transition factor — 3.5
ALPHAT1 Trap coefficient 1 on bias stress — 0.0001
ALPHAT2 Trap coefficient 2 on bias stress V 21
BETA Linear to saturation parameter — 1.5
BETAFP1 FP linear to saturation parameter — 1
BETAFP2 FP linear to saturation parameter — 1
BETAFP3 FP linear to saturation parameter — 1
BETAFP4 FP linear to saturation parameter — 1
BETARD DAR linear to saturation parameter — 1
BETARECD G-D linear to saturation parameter — 0.25
BETARECS G-S linear to saturation parameter — 2
BETARS SAR linear to saturation parameter — 1
CBFP1 Body to drain (under FP) cap/width — 0

666
 Table 2-137. MVSG-HV HEMT MODEL Device Model Parameters
Parameter Description Units Default
CBFP2 Body to drain (under FP) cap/width — 0
CBFP3 Body to drain (under FP) cap/width — 0
CBFP4 Body to drain (under FP) cap/width — 0
CCFP1 Source or gate to drain (under FP) cap/width — 9e-11
CCFP2 Source or gate to drain (under FP) cap/width — 3e-11
CCFP3 Source or gate to drain (under FP) cap/width — 9e-11
CCFP4 Source or gate to drain (under FP) cap/width — 9e-11
CFP1S FP (source-side) to source cap/width — 1e-19
CFP2S FP (source-side) to source cap/width — 1e-19
CFP3S FP (source-side) to source cap/width — 1e-19
CFP4S FP (source-side) to source cap/width — 1e-19
CG Gate cap/area — 0.004
CGFP1 FP gate-cap/area — 0.0002
CGFP2 FP gate-cap/area — 0.0001
CGFP3 FP gate-cap/area — 0.0002
CGFP4 FP gate-cap/area — 0.0002
CGRD DAR gate-cap/area — 0.0043
CGRS SAR gate-cap/area — 0.005
COFDM Gate - Drain outer fringing cap/width — 1e-10
COFDSM Source - Drain outer fringing cap/width — 1e-10
COFDSUBM Sub - Drain outer fringing cap/width — 0
COFGSUBM Sub - Gate outer fringing cap/width — 0
COFSM Gate - Source outer fringing cap/width — 1e-09
COFSSUBM Sub - Source outer fringing cap/width — 0
CTH Thermal capacitance — 0.0001
CTRAP DC-block capacitor F 0.001
DELTA1 DIBL Coefficient 1 — 0.016
DELTA1FP1 FP DIBL Coefficient — 0
DELTA1FP2 FP DIBL Coefficient — 0
DELTA1FP3 FP DIBL Coefficient — 0
DELTA1FP4 FP DIBL Coefficient — 0
DELTA1RD DAR DIBL Coefficient — 0.35
DELTA1RS SAR DIBL Coefficient — 0.1
DELTA2 DIBL Coefficient 2 — 0
DIBSAT DIBL saturation Voltage V 10
EPSILON Mobility dependence on temperature — 2.3

667
 Table 2-137. MVSG-HV HEMT MODEL Device Model Parameters
Parameter Description Units Default
FFE Flicker noise exponent for frequency — 1.2
FLAGFP1 Flag parameter: GFP=1 or SFP=0 — 1
FLAGFP1S Flag parameter: cfp1s select=1 or cfp1s not select=0 — 1
FLAGFP2 Flag parameter: GFP=1 or SFP=0 — 0
FLAGFP2S Flag parameter: cfp1s select=1 or cfp1s not select=0 — 1
FLAGFP3 Flag parameter: GFP=1 or SFP=0 — 0
FLAGFP3S Flag parameter: cfp1s select=1 or cfp1s not select=0 — 1
FLAGFP4 Flag parameter: GFP=1 or SFP=0 — 0
FLAGFP4S Flag parameter: cfp1s select=1 or cfp1s not select=0 — 1
Flag parameter for resistor: resistor is chosen if
FLAGRES — 0
flagres=1 or implicit transitor is chosen if flagres=0
G-D fractional change in ideality factor due to high
FRACD — 0.5
injection
G-S fractional change in ideality factor due to high
FRACS — 0.5
injection
GMDISP Flag parameter for gm-dispersion 0=off, 1=on — 0
IGMOD Flag parameter for gate leakage 0=off, 1=on — 1
IJD G-D reverse leakage current normalized to width — 1e-12
IJS G-S reverse leakage current normalized to width — 1e-12
IRECD G-D reverse leakage current normalized to width — 2e-05
IRECS G-S reverse leakage current normalized to width — 1e-18
G-D fitting parameter to turn on the breakdown of
KBDGATED — 0
G-D diode
G-S fitting parameter to turn on the breakdown of G-S
KBDGATES — 0
diode
KF Flicker noise coefficient — 0.0001
LAMBDA CLM parameter — 0
LGD Drain access region (DAR) length parameter m 4.85e-06
LGFP1 FP Length m 0
LGFP2 FP Length m 0
LGFP3 FP Length m 0
LGFP4 FP Length m 0
LGS Source access region (SAR) length parameter m 3e-06
LMAX Maximum length for use of this model m 100
LMIN Minimum length for use of this model m 0
MINC Minimum capacitance F 0
Minimum length of access or FP regions for modeling
MINL m 1e-09
them as transistors

668
 Table 2-137. MVSG-HV HEMT MODEL Device Model Parameters
Parameter Description Units Default
MINR Minimum resistance ˙ 0.001
MTHETA Scattering: mobility reduction parameter with Vg — 0
MTHETAFP1 FP scattering: mobility reduction parameter with Vg — 0
MTHETAFP2 FP scattering: mobility reduction parameter with Vg — 0
MTHETAFP3 FP scattering: mobility reduction parameter with Vg — 0
MTHETAFP4 FP scattering: mobility reduction parameter with Vg — 0
MTHETARD DAR scattering: mobility reduction parameter with Vg — 0
MTHETARS SAR scattering: mobility reduction parameter with Vg — 0
MU0 Low-field mobility — 0.135
MU0FP1 FP low-field mobility — 0.2
MU0FP2 FP low-field mobility — 0.2
MU0FP3 FP low-field mobility — 0.2
MU0FP4 FP low-field mobility — 0.2
MU0RD DAR low-field mobility — 0.1
MU0RS SAR low-field mobility — 0.1
ND Punchthrough factor for subth slope — 0
NDFP1 FP punchthrough factor for subth slope — 0
NDFP2 FP punchthrough factor for subth slope — 0
NDFP3 FP punchthrough factor for subth slope — 0
NDFP4 FP punchthrough factor for subth slope — 0
NDRD DAR punchthrough factor for subth slope — 3.8
NDRS SAR punchthrough factor for subth slope — 0
NOISEMOD Select knob for noise model 0=off, 1=on — 0
G-D fitting parameter for breakdown: Something like
PBDGD — 4
1/eta*Vt
G-S fitting parameter for breakdown: Something like
PBDGS — 4
1/eta*Vt
PG_PARAM1 Something like 1/eta — 0.82
PG_PARAMD G-D something like 1/eta*Vt — 1
PG_PARAMS G-S something like 1/eta*Vt — 1
PGSRECD G-D something like 1/eta for reverse recombination — 0.8
PGSRECS G-S something like 1/eta for reverse recombination — 0.5
RCD Drain contact resistance * Width — 0.0008
RCS Source contact resistance * Width — 0.0008
RCT1 Linear Rsh and Rc temperature coefficient 1/K 0
RCT2 Quadratic Rsh and Rc temperature coefficient — 0

669
 Table 2-137. MVSG-HV HEMT MODEL Device Model Parameters
Parameter Description Units Default
RGSP Gate resistance * Width — 0
RSH 2-DEG Sheet Resistance — 150
RTH Thermal resistance — 25
SFP1 FP Sub-threshold slope — 3.2
SFP2 FP Sub-threshold slope — 3.2
SFP3 FP Sub-threshold slope — 3.2
SFP4 FP Sub-threshold slope — 3.2
SHD G-D shot noise parameter — 3
SHS G-S shot noise parameter — 3
SRD DAR Sub-threshold slope — 0.3
SRS SAR Sub-threshold slope — 0.1
SS Sub-threshold slope — 0.12
TAUGMRF gm-dispersion time constant s 0.001
TAUT Trap time constant s 3e-05
TEMPT Temperature coefficient for trapping 1/K 0.0001
TNOM Reference temperature for the model — 27
TRAPSELECT Select knob for charge trapping 0=off, 1=on — 0
TYPE nFET=1 pFET=-1 — 1
VBDGD G-D soft breakdown voltage of G-D diode V 600
VBDGS G-S soft breakdown voltage of G-S diode V 600
VERSION Version number — 1
VGSATD G-D high injection effect V 1
VGSATQD G-D mimics depletion saturation V 0.8
VGSATQS G-S mimics depletion saturation V 2
VGSATS G-S high injection effect V 1
VJG Gate diode cut in voltage V 1.1
VTHETA Scattering: velocity reduction parameter with Vg — 0
VTHETAFP1 FP scattering: velocity reduction parameter with Vg — 0
VTHETAFP2 FP scattering: velocity reduction parameter with Vg — 0
VTHETAFP3 FP scattering: velocity reduction parameter with Vg — 0
VTHETAFP4 FP scattering: velocity reduction parameter with Vg — 0
VTHETARD DAR scattering: velocity reduction parameter with Vg — 0
VTHETARS SAR scattering: velocity reduction parameter with Vg — 0
VTO Threshold voltage V -2.72
VTOFP1 FP threshold voltage V -44.5
VTOFP2 FP threshold voltage V -74.5

670
 Table 2-137. MVSG-HV HEMT MODEL Device Model Parameters
Parameter Description Units Default
VTOFP3 FP threshold voltage V -44.5
VTOFP4 FP threshold voltage V -44.5
VTORD DAR threshold voltage V -650
VTORS SAR threshold voltage V -650
VTTRAP Trapping stress threshold voltage V 230
VTZETA vto dependence on temperature V/K -0.0004
VX0 Source injection velocity — 300000
VX0FP1 FP source injection velocity — 120000
VX0FP2 FP source injection velocity — 120000
VX0FP3 FP source injection velocity — 120000
VX0FP4 FP source injection velocity — 120000
VX0RD DAR source injection velocity — 100000
VX0RS SAR source injection velocity — 100000
VZETA vx0 dependence on temperature 1/K 150000
WMAX Maximum width for use of this model m 100
WMIN Minimum width for use of this model m 0

Table 2-138. MOSFET level 2002 Output Variables

Parameter Description Units Default
vgisi internal gate-source voltage V none
vdisi internal drain-source voltage V none
vti internal threshold voltage including DIBL V none
vdsati internal drain-source saturation voltage V none
pdc total power dissipation from the device W none
idisi drain-to-source current in intrinsic transistor A none
igs gate-source gate-leakage current A none
igd gate-drain gate-leakage current A none
qgi intrinsic gate charge C none
qdi intrinsic drain charge C none
qsi intrinsic source charge C none
qbi intrinsic body charge C none
gmi intrinsic transconductance A/V none
gdsi intrinsic output-conductance A/V none
gmbsi intrinsic body-transconductance A/V none
cggi intrinsic gate-gate capacitance F none

671
 Table 2-138. MOSFET level 2002 Output Variables
Parameter Description Units Default
cgdi intrinsic gate-drain capacitance F none
cgsi intrinsic gate-source capacitance F none
cgbi intrinsic gate-body capacitance F none
cdgi intrinsic drain-gate capacitance F none
cddi intrinsic drain-drain capacitance F none
cdsi intrinsic drain-source capacitance F none
cdbi intrinsic drain-body capacitance F none
csgi intrinsic source-gate capacitance F none
csdi intrinsic source-drain capacitance F none
cssi intrinsic source-source capacitance F none
csbi intrinsic source-body capacitance F none
cbgi intrinsic body-gate capacitance F none
cbdi intrinsic body-drain capacitance F none
cbsi intrinsic body-source capacitance F none
cbbi intrinsic body-body capacitance F none
cgs gate-to-source fringing capacitance F none
cgd gate-to-drain fringing capacitance F none
t_total_k actual device temperature in Kelvin K none
t_total_c actual device temperature in Celsius deg C none
t_delta_sh change in device temperature caused by self-heating K none
rs resistance of source access region Ohm none
rd resistance of drain access region Ohm none

672
2.3.20.17. Level 260 MOSFET Tables (EKV version 2.6)

Xyce includes the EKV MOSFET model, version 2.6 as the level 260 MOSFET device.
Official documentation of this model may be found at
https://www.epfl.ch/labs/iclab/wp-content/uploads/2019/02/ekv_v262.pdf.
We have implemented EKV 2.6 directly from the Verilog-A source published by its authors at
https://github.com/ekv26/model. While it is a faithful implementation of the model provided there,
we have had anecdotal evidence that other simulators have different implementations that contain additional
parameters and possibly a different extrinsic model. Model cards containing parameters extracted from
other simulators may not result in Xyce simulations that match those other simulators. Watch carefully for
any warnings from Xyce regarding unrecognized model parameters, as these are a strong indication that the
model card is not extracted using the exact version of EKV provided by Xyce.
Tables of EKV MOSFET 2.6 parameters are in tables 2-139 and 2-140.
Note that in the tables the device claims that the default TNOM and TEMP parameter values is 1e21. This is
merely an artifact of an unusual way the authors have defined those parameters in the Verilog-A source. In
fact, if not given TNOM defaults to 25 ◦ C, and if not given TEMP defaults to the ambient temperature of the
simulation.

Table 2-139. EKV MOSFET version 2.6 Device Instance Parameters

Parameter Description Units Default
AD – 0
AS – 0
L – 1e-05
M – 1
NS – 1
PD – 0
PS – 0
W – 1e-05

Table 2-140. EKV MOSFET version 2.6 Device Model Parameters

Parameter Description Units Default
AD – 0
AF – 1
AGAMMA – 1e-06
AKP – 1e-06
AS – 0
AVTO – 1e-06
BEX – -1.5
COX – 0.002

673
 Table 2-140. EKV MOSFET version 2.6 Device Model Parameters
Parameter Description Units Default
DL – -1e-08
DW – -1e-08
E0 – 1e+08
GAMMA – 0.7
HDIF – 5e-07
IBA – 5e+08
IBB – 4e+08
IBBT – 0.0009
IBN – 1
KF – 0
KP – 0.00015
L – 1e-05
LAMBDA – 0.8
LETA – 0.3
LK – 4e-07
LMAX – 100
LMIN – 0
M – 1
NOISE – 1
NS – 1
PD – 0
PHI – 0.5
PS – 0
Q0 – 0.00023
RSH – 0
TCV – 0.001
TEMP – 1e+21
THETA – 0
TNOM – 1e+21
TP_CJ – 0
TP_CJSW – 0
TP_CJSWG – 0
TP_NJTS – 0
TP_NJTSSW – 0
TP_NJTSSWG – 0
TP_PB – 0

674
 Table 2-140. EKV MOSFET version 2.6 Device Model Parameters
Parameter Description Units Default
TP_PBSW – 0
TP_PBSWG – 0
TP_XTI – 3
TRISE – 0
TYPE – 1
UCEX – 0.8
UCRIT – 2e+06
VTO – 0.5
W – 1e-05
WETA – 0.2
WMAX – 100
WMIN – 0
XD_BV – 10
XD_CJ – 1e-09
XD_CJSW – 1e-12
XD_CJSWG – 1e-12
XD_GMIN – 0
XD_JS – 1e-09
XD_JSW – 1e-12
XD_JSWG – 1e-12
XD_MJ – 0.9
XD_MJSW – 0.7
XD_MJSWG – 0.7
XD_N – 1
XD_NJTS – 1
XD_NJTSSW – 1
XD_NJTSSWG – 1
XD_PB – 0.8
XD_PBSW – 0.6
XD_PBSWG – 0.6
XD_VTS – 0
XD_VTSSW – 0
XD_VTSSWG – 0
XD_XJBV – 0
XJ – 3e-07

675
2.3.20.18. Level 301 MOSFET Tables (EKV version 3.0.1)

Xyce includes the EKV MOSFET model, version 3.0.1 [19][31][32]. Full documentation for the EKV3
model is available on the Xyce internal web site; the documentation for the EKV3 model may be freely
redistributed. Instance and model parameters for the EKV model are given in tables 2-141 and 2-142.
The EKV3 model is developed by the EKV Team of the Electronics Laboratory-TUC (Technical University
of Crete). It is included in Xyce under license from Technical University of Crete. The official web site of
the EKV model is http://ekv.epfl.ch/.
Due to licensing restrictions, the EKV3 MOSFET is not available in open-source versions of Xyce.
The license for EKV3 authorizes Sandia National Laboratories to distribute EKV3 only in binary
versions of code.

Table 2-141. EKV3 MOSFET Device Instance Parameters

Parameter Description Units Default
AD DRAIN’S AREA – 0
AS SOURCE’S AREA – 0
L GATE’S LENGTH – 1e-05
M NUMBER OF DEVICES IN PARALLEL – 1
NF NUMBER OF FINGERS – 1
PD DRAIN’S PERIMETER – 0
PS SOURCE’S PERIMETER – 0
SA STI PARAMETER; DISTANCE FROM STI – 0
SB STI PARAMETER; DISTANCE FROM STI – 0
SD STI PARAMETER; DISTANCE BETWEEN GATES – 0
W GATE’S WIDTH – 1e-05

Table 2-142. EKV3 MOSFET Device Model Parameters

Parameter Description Units Default
ACLM – 0.83
AF – 1
AGAM – 0
MATCHING PARAMETER FOR BODY FACTOR
AGAMMA – 0
(GAMMA)
AGIDL – 0
AKP MATCHING PARAMETER FOR MOBILITY (KP) – 0
AQMA – 0.5
AQMI – 0.4
AVT – 0

676
 Table 2-142. EKV3 MOSFET Device Model Parameters
Parameter Description Units Default
MATCHING PARAMETER FOR THRESHOLD
AVTO – 0
VOLTAGE (VTO)
BEX – -1.5
BGIDL – 2.3e+09
BVD – 10
BVS – 10
CGBO – 0
CGDO – 0
CGIDL – 0.5
CGSO – 0
CJD – 0
CJF – 0
CJS – 0
CJSWD – 0
CJSWGD – 0
CJSWGS – 0
CJSWS – 0
COX – 0.012
DDITS – 0.3
DELTA – 2
DFR – 0.001
DGAMMAEDGE – 0
DL – -1e-08
DLC – 0
DPHIEDGE – 0
DW – -1e-08
DWC – 0
E0 – 1e+10
E1 – 3.1e+08
EB – 2.9e+10
EF – 2
EGIDL – 0.8
ETA – 0.5
ETAD – 1
ETAQM – 0.75
FLR – 0

677
 Table 2-142. EKV3 MOSFET Device Model Parameters
Parameter Description Units Default
FPROUT – 1e+06
GAMMA – 0.3
GAMMAG – 4.1
GAMMAGOV – 10
GAMMAOV – 1.6
GC – 1
GMIN – 0
HDIF – 0
IBA – 0
IBB – 3e+08
IBBT – 0.0008
IBN – 1
INFO_LEVEL – 0
JSD – 0
JSS – 0
JSSWD – 0
JSSWGD – 0
JSSWGS – 0
JSSWS – 0
JTSD – 0
JTSS – 0
JTSSWD – 0
JTSSWGD – 0
JTSSWGS – 0
JTSSWS – 0
KA – 0
KB – 0
KETAD – 0
KF – 0
KG – 0
KGAMMA – 0
KGFN – 0
KJF – 0
KKP – 0
KP – 0.0005
KRGL1 – 0

678
 Table 2-142. EKV3 MOSFET Device Model Parameters
Parameter Description Units Default
KUCRIT – 0
KVTO – 0
LA – 1
LAMBDA – 0.5
LB – 1
LDIF – 0
LDPHIEDGE – 0
LDW – 0
LETA – 0.5
LETA0 – 0
LETA2 – 0
LGAM – 1
LKKP – 0
LKVTO – 0
LL – 0
LLN – 1
LLODKKP – 1
LLODKVTO – 1
LNWR – 0
LODKETAD – 1
LODKGAMMA – 1
LOV – 2e-08
LOVIG – 2e-08
LQWR – 0
LR – 5e-08
LVT – 1
LWR – 0
MJD – 0.9
MJS – 0.9
MJSWD – 0.7
MJSWGD – 0.7
MJSWGS – 0.7
MJSWS – 0.7
N0 – 1
NCS – 1
NFVTA – 0

679
 Table 2-142. EKV3 MOSFET Device Model Parameters
Parameter Description Units Default
NFVTB – 10000
NJD – 1
NJS – 1
NJTSD – 1
NJTSS – 1
NJTSSWD – 1
NJTSSWGD – 1
NJTSSWGS – 1
NJTSSWS – 1
NLR – 0.01
NQS_NOI – 1
NWR – 0.005
PBD – 0.8
PBS – 0.8
PBSWD – 0.6
PBSWGD – 0.6
PBSWGS – 0.6
PBSWS – 0.6
PDITS – 0
PDITSD – 1
PDITSL – 0
PHIF FERMI BULK POTENTIAL – 0.45
PKKP – 0
PKVTO – 0
QLR – 0.0005
QOFF – 0
QWR – 0.0003
RBN – 0
RBWSH – 0.003
RD – 0
RDBN – 0
RDBWSH – 0.001
RDSBSH – 1000
RDX – -1
RGSH – 3
RINGTYPE – 1

680
 Table 2-142. EKV3 MOSFET Device Model Parameters
Parameter Description Units Default
RLX EXTERNAL SERIES RESISTANCE – -1
RS – 0
RSBN – 0
RSBWSH – 0.001
RSH – 0
RSX – -1
SAREF – 0
SBREF – 0
SCALE – 1
SIGMAD – 1
SIGN SIGN = 1 FOR NMOS; SIGN = -1 FOR PMOS – 1
TCJ – 0
TCJSW – 0
TCJSWG – 0
TCV – 0.0006
TCVL – 0
TCVW – 0
TCVWL – 0
TE0EX – 0.5
TE1EX – 0.5
TETA – -0.0009
TYPE OF GATE: -1 ENHANCEMENT TYPE; 1
TG – -1
DEPLETION TYPE
TH_NOI – 0
THC – 0
TKKP – 0
TLAMBDA – 0
TNJTSD – 0
TNJTSS – 0
TNJTSSWD – 0
TNJTSSWGD – 0
TNJTSSWGS – 0
TNJTSSWS – 0
TNOM – 27
TPB – 0
TPBSW – 0

681
 Table 2-142. EKV3 MOSFET Device Model Parameters
Parameter Description Units Default
TPBSWG – 0
TR – 0
TR2 – 0
UCEX – 1.5
UCRIT – 5e+06
VBI – 0
VFBOV – 0
VFR – 0
VOV – 1
VTO THRESHOLD VOLTAGE – 0.3
VTSD – 0
VTSS – 0
VTSSWD – 0
VTSSWGD – 0
VTSSWGS – 0
VTSSWS – 0
WDL – 0
WDPHIEDGE – 0
WE0 – 0
WE1 – 0
WEDGE – 0
WETA – 0.2
WETAD – 0
WGAM – 1
WKKP – 0
WKP1 – 1e-06
WKP2 – 0
WKP3 – 1
WKVTO – 0
WLAMBDA – 0
WLDGAMMAEDGE – 0
WLDPHIEDGE – 0
WLOD – 0
WLODKKP – 1
WLODKVTO – 1
WLR – 0

682
 Table 2-142. EKV3 MOSFET Device Model Parameters
Parameter Description Units Default
WNLR – 0
WQLR – 0
WR – 9e-08
WRLX – 0
WUCEX – 0
WUCRIT – 0
WVT – 1
XB – 3.1
XJ – 2e-08
XJBVD – 0
XJBVS – 0
XL – 0
XTID – 3
XTIS – 3
XTSD – 0
XTSS – 0
XTSSWD – 0
XTSSWGD – 0
XTSSWGS – 0
XTSSWS – 0
XW – 0
ZC – 1e-06

683
2.3.20.19. Level 10240 MOSFET Tables (L_UTSOI Version 102.4.0)

Select Xyce binaries include the L_UTSOI MOSFET model as the level 10240 MOSFET. This model’s
parameters and output variables are listed in tables 2-143, 2-144, and 2-145

Table 2-143. L_UTSOI MOSFET Device Instance Parameters

Parameter Description Units Default
ADRAIN Drain region area m2 1e-12
ASOURCE Source region area m2 1e-12
DELVTO Threshold voltage shift parameter V 0
FACTUO Low-field mobility pre-factor — 1
L Channel length m 1e-06
M multiplicity factor — 1
MULT Number of devices in parallel — 1
NF Number of fingers — 1
PDRAIN Drain region perimeter m 1e-06
PSOURCE Source region perimeter m 1e-06
SA Distance between OD-edge and poly at source side m 0
SB Distance between OD-edge and poly at drain side m 0
SD Distance between neighboring fingers m 0
W Channel width m 1e-06

Table 2-144. L_UTSOI MOSFET Device Model Parameters

Parameter Description Units Default
A1 Impact ionization pre-factor — 1
A1L Length dependence of A1 — 0
A1O Geometry independent impact ionization pre-factor — 1
A1W Width dependence of A1 — 0
A2 Impact ionization exponent at TR — 10
A2O Impact ionization exponent at TR — 10
A3 Saturation voltage dependence of impact-ionization — 1
A3L Length dependence of A3 — 0
Geometry independent saturation-voltage dependence
A3O — 1
of II
A3W Width dependence of A3 — 0
AGIDL GIDL pre-factor A/V3 0
AGIDLD GIDL pre-factor at drain side A/V3 0
AGIDLDO GIDL geometry independent pre-factor at drain side A/V3 0

684
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
AGIDLDW GIDL pre-factor for a width of WEN at drain side A/V3 0
AGIDLO GIDL geometry independent pre-factor A/V3 0
AGIDLW GIDL pre-factor for a width of WEN A/V3 0
ALP CLM pre-factor — 0
ALP1 CLM enhancement factor above threshold V 0
Length dependence of CLM enhancement factor
ALP1L1 V 0
above threshold ALP1
ALP1L2 Second order length dependence of ALP1 — 0
ALP1LEXP Exponent for length dependence of ALP1 — 0.5
Exponent for second order length dependence of
ALP1LEXP2 — 1.5
ALP1
ALP1W Width dependence of ALP1 — 0
ALPB Back bias dependence of channel length modulation — 0
ALPBO Back bias dependence of channel length modulation — 0
ALPL1 Length dependence of CLM pre-factor ALP — 0
ALPL2 Second order length dependence of ALP — 0
ALPLEXP Exponent for length dependence of ALP — 1
ALPLEXP2 Exponent for second order length dependence of ALP — 2
ALPW Width dependence of ALP — 0
AREAQ Effective channel area for intrinsic CV m2 1e-12
AX Linear/saturation transition exponent — 8
AXL Length dependence of AX — 0
AXL2 Second order length dependence of AX — 0
AXLEXP Exponent for length dependence of AX — 1
AXLEXP2 Exponent for second order length dependence of AX — 1.5
Geometry independent linear/saturation transition
AXO — 8
exponent
BETN Front channel aspect ratio times zero-field mobility m2 /(Vs) 0.05
Back channel over front channel zero-field mobility
BETNB — 1
ratio
Back channel over front channel zero-field mobility
BETNBO — 1
ratio
BETW1 First width dependence modulation of BETN — 0
BETW2 Second width dependence modulation of BETN — 0
BGIDL GIDL probability factor at TR V 41
BGIDLD GIDL probability factor at TR at drain side V 41
BGIDLDO GIDL probability factor at TR at drain side V 41

685
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
BGIDLO GIDL probability factor at TR V 41
CF DIBL parameter — 0
CFB DIBL back to front interface asymmetry factor — 1
CFBO DIBL back to front interface asymmetry factor — 1
CFD Drain voltage dependence parameter of DIBL V 0.2
CFDL DIBL modulation coefficient due to Leff variation — 0
CFDLB Back bias dependence of DIBL modulation — 0
CFDLBO Back bias dependence of DIBL modulation — 0
CFDLL DIBL modulation coefficient due to Leff variation — 0
CFDLW Width dependence of CFDL — 0
CFDO Drain voltage dependence parameter of DIBL V 0.2
CFL Length dependence of DIBL-parameter — 0
CFLEXP Exponent for length dependence of CF — 2
CFR Outer fringe capacitance per side F 0
CFRD Outer fringe capacitance at drain side F 0
CFRDO Corner related outer fringe capacitance at drain side F 0
Outer fringe capacitance per side for a width of WEN
CFRDW F 0
at drain side
CFRO Corner related outer fringe capacitance F 0
CFRW Outer fringe capacitance per side for a width of WEN F 0
CFW Width dependence of CF — 0
CGBOV Oxide capacitance for gate-substrate overlap F 0
Length dependent gate-substrate overlap capacitance
CGBOVL F 0
part for a length of LEN
Geometry independent gate-substrate overlap
CGBOVO F 0
capacitance part
CGIDL Substrate bias dependence of GIDL V−1 0
CGIDLD Substrate bias dependence of GIDL at drain side V−1 0
CGIDLDO Substrate bias dependence of GIDL at drain side V−1 0
CGIDLO Substrate bias dependence of GIDL V−1 0
CHIB Tunneling barrier height V 3.1
CHIBO Tunneling barrier height V 3.1
CIC Long channel back interface coupling coefficient — 1
CICF Long channel front interface coupling coefficient — 1
CICFO Long channel front interface coupling coefficient — 1
CICO Long channel back interface coupling coefficient — 1
COV Overlap capacitance per side F 0

686
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
COVD Overlap capacitance at drain side F 0
Overlap capacitance modulation coefficient due to
COVDL — 0
Leff variation
COVDLB Overlap capacitance modulation with back bias — 0
COVDLBO Overlap capacitance modulation with back bias — 0
Overlap capacitance modulation coefficient due to
COVDLO — 0
Leff variation
COVDLW Width dependence of COVDL — 0
CS Remote Coulomb scattering parameter at TR — 0
Field dependence of Coulomb scattering at back
CSBI — 0
interface
Field dependence of Coulomb scattering at back
CSBIO — 0
interface
CSD Drain-source capacitance F 1.04e-18
CSDBP Drain/source to substrate perimeter capacitance — 0
CSDBPO Drain/source to substrate perimeter capacitance — 0
CSDO Drain-source capacitance correction factor — 1
Field dependence of Coulomb scattering at front
CSFI — 0
interface
Field dependence of Coulomb scattering at front
CSFIO — 0
interface
CSL Length dependence of CS — 0
CSLEXP Exponent describing length dependence of CS — 1
CSLW Area dependence of CS — 0
CSO Remote Coulomb scattering parameter at TR — 0
CSTHR Remote Coulomb scattering threshold level — 2
Remote Coulomb scattering threshold asymmetry
CSTHRB — 1
parameter
Remote Coulomb scattering threshold asymmetry
CSTHRBO — 1
parameter
CSTHRO Remote Coulomb scattering threshold level — 2
CSW Width dependence of CS — 0
CT Interface states factor — 0
CTH Thermal capacitance — 1e-11
CTHO Geometry independent thermal capacitance — 1e-12
CTO Interface states factor — 0
DGIDL High longitudinal field parameter of GIDL V−1 0

687
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
High longitudinal field parameter of GIDL at drain
DGIDLD V−1 0
side
High longitudinal field parameter of GIDL at drain
DGIDLDL V−1 0
side
High field geometry independent parameter of GIDL
DGIDLDO V−1 0
at drain side
DGIDLL High longitudinal field parameter of GIDL V−1 0
DGIDLO High field geometry independent parameter of GIDL V−1 0
Effective channel length additional offset for charge
DLQ m 0
model
DVFBOV Overlap capacitance flat-band voltage adjustment V 0
DVFBOVO Overlap capacitance flat-band voltage adjustment V 0
Effective channel width additional offset for charge
DWQ m 0
model
EF Frequency coefficient of flicker noise — 1
EFO Frequency coefficient of flicker noise — 1
FBET1 First length dependence modulation of BETN — 0
FBET1W Width dependence of FBET1 — 0
FBET2 Second length dependence modulation of BETN — 0
FETA Effective field parameter — 1
FETAO Effective field parameter — 1
FIF Inner fringe capacitance prefactor — 0
FIFW Inner fringe capacitance prefactor for a width of WEN — 0
FNOVINV Extra gate to overlap current pre-factor in inversion A 0
Extra gate to overlap current pre-factor in inversion at
FNOVINVD A 0
drain side
Extra gate to overlap current pre-factor for a width of
FNOVINVDW A 0
WEN in inversion at drain side
Extra gate to overlap current pre-factor for a width of
FNOVINVW A 0
WEN in inversion
FNT Thermal noise coefficient — 1
FNTEXC Excess noise coefficient — 0
FNTEXCL Length dependence coefficient of excess noise — 0
FNTEXCLEXP Length dependence exponent of excess noise — 2
FNTO Thermal noise coefficient — 1
Short channel effect adjustment factor for charge
FSCEAC — 0
model
Short channel effect adjustment factor for charge
FSCEACO — 0
model

688
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
First neighbour thermal coupling factor for
FTHO — 0
multifinger devices
GC2CH Gate to channel current slope factor — 0.375
GC2CHO Gate to channel current slope factor — 0.375
Gate current slope factor for overlap regions in
GC2OVACC — 0.375
accumulation mode
Gate current slope factor for overlap regions in
GC2OVACCO — 0.375
accumulation mode
Gate current slope factor for overlap regions in
GC2OVINV — 0.375
inversion mode
Gate current slope factor for overlap regions in
GC2OVINVO — 0.375
inversion mode
GC3CH Gate to channel current curvature factor — 0.063
GC3CHO Gate to channel current curvature factor — 0.063
Gate current curvature factor for overlap regions in
GC3OVACC — 0.063
accumulation mode
Gate current curvature factor for overlap regions in
GC3OVACCO — 0.063
accumulation mode
Gate current curvature factor for overlap regions in
GC3OVINV — 0.063
inversion mode
Gate current curvature factor for overlap regions in
GC3OVINVO — 0.063
inversion mode
GCDOV High drain voltage dependence of overlap gate current V−1 0
GCDOVL High drain voltage dependence of overlap gate current V−1 0
GCO Gate tunneling energy adjustment in inversion — 0
GCOO Gate tunneling energy adjustment in inversion — 0
Extra gate current slope factor for overlap regions in
GCOVINVFN — 0.2
inversion mode
Extra gate current slope factor for overlap regions in
GCOVINVFNO — 0.2
inversion mode
Threshold of high drain voltage effect on overlap gate
GCVDOV V 1
current
Threshold of high drain voltage effect on overlap gate
GCVDOVO V 1
current
IGINV Gate to channel current pre-factor A 0
Gate to channel current pre-factor for a channel area
IGINVLW A 0
of WEN.LEN
IGOVACC Gate to overlap current pre-factor in accumulation A 0
Gate to overlap current pre-factor in accumulation at
IGOVACCD A 0
drain side

689
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
Gate to overlap current pre-factor for a width of WEN
IGOVACCDW A 0
in accumulation at drain side
Gate to overlap current pre-factor for a width of WEN
IGOVACCW A 0
in accumulation
IGOVINV Gate to overlap current pre-factor in inversion A 0
Gate to overlap current pre-factor in inversion at drain
IGOVINVD A 0
side
Gate to overlap current pre-factor for a width of WEN
IGOVINVDW A 0
in inversion at drain side
Gate to overlap current pre-factor for a width of WEN
IGOVINVW A 0
in inversion
KUO Mobility degradation/enhancement coefficient m 0
Saturation velocity degradation/enhancement
KVSAT — 0
coefficient
KVTHO Threshold shift parameter Vm 0
Characteristic length of lateral thermal coupling for
LAMBTHO m 1e-07
multifinger devices
LAP Effective channel length reduction per side m 0
LKUO Length dependence of KUO — 0
LKVTHO Length dependence of KVTHO — 0
LLODKUO Length parameter for UO stress effect — 0
LLODVTH Length parameter for VTH-stress effect — 0
LODETAO Eta0 shift modification factor for stress effect — 1
LOVDO Overlap length for gate/drain hdd overlap capacitance m 0
Overlap length for gate/source-drain hdd overlap
LOVO m 0
capacitance
LP1 First characteristic length of BETN scaling m 1e-08
LP1W Width dependence of LP1 — 0
LP2 Second characteristic length of BETN scaling m 1e-08
LVARL Length dependence of LPS — 0
Geometry independent difference between physical
LVARO m 0
and drawn gate lengths
LVARW Width dependence of LPS — 0
MUE Front channel mobility reduction coefficient at TR — 0
MUEO Front channel mobility reduction coefficient at TR — 0
Thin film doping (n-type=negative value,
NCH — 0
p-type=positive value)
Thin film doping (n-type=negative value,
NCHO — 0
p-type=positive value)

690
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
NFA First coefficient of flicker noise — 8e+22
NFALW First coefficient of flicker noise — 8e+22
NFAW Long channel first coefficient of flicker noise — 0
NFB Second coefficient of flicker noise — 3e+07
NFBLW Second coefficient of flicker noise — 3e+07
NFC Third coefficient of flicker noise V−1 0
NFCLW Third coefficient of flicker noise V−1 0
NFE Flicker noise front transverse field effect coefficient — 0
NFEB Flicker noise back transverse field effect coefficient — 0
NFEBO Flicker noise back transverse field effect coefficient — 0
NFEO Flicker noise front transverse field effect coefficient — 0
Gate tunneling slope adjustment in subthreshold
NIGINV — 0
regime
Gate tunneling slope adjustment in subthreshold
NIGINVO — 0
regime
NOV Effective doping of overlap-ldd regions — 1e+20
NOVD Effective doping of overlap-ldd regions ar drain side — 1e+20
NOVDO Effective doping of overlap-ldd regions at drain side — 1e+20
NOVO Effective doping of overlap-ldd regions — 1e+20
NSDAC Source/Drain effective doping level for AC model — 1e+22
NSDACO Source/Drain effective doping level for AC model — 1e+22
NSDDC Source/Drain effective doping level for DC model — 1e+22
NSDDCO Source/Drain effective doping level for DC model — 1e+22
Substrate doping (n-type=negative value,
NSUB — 3e+18
p-type=positive value)
Substrate doping (n-type=negative value,
NSUBO — 3e+18
p-type=positive value)
PKUO Cross-term dependence of KUO — 0
PKVTHO Cross-term dependence of KVTHO — 0
PNCE Narrow channel effect on body factor — 0
Narrow channel effect on body factor for a width of
PNCEW — 0
WEN
PSCE Short channel effect coefficient — 0
PSCEB Short channel back to front interface asymmetry factor — 1
PSCEBO Short channel back to front interface asymmetry factor — 1
Back bias dependence of short channel effect
PSCEDLB — 0
modulation

691
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
Back bias dependence of short channel effect
PSCEDLBO — 0
modulation
PSCEL Length dependence of PSCE — 0
PSCELEXP Exponent describing length dependence of PSCE — 2
PSCEW Width dependence of PSCE — 0
Quantum correction factor (no correction=0, full
QMC — 1
correction=1)
RS Source/Drain series resistance at TR ˙ 30
RSB Back bias dependence of RS — 0
RSBO Back bias dependence of RS — 0
RSG Transverse electric field dependence of RS — 0
RSGO Transverse electric field dependence of RS — 0
RSIG Source/Drain extension resistance coefficient — 0
RSIGO Source/Drain extension resistance coefficient — 0
Source/Drain series resistance for channel width
RSW1 ˙ 30
WEN at TR
Higher-order width scaling of source/drain series
RSW2 — 0
resistance
RTH Thermal resistance — 10000
RTHL Length dependence of RTH — 1.5
RTHLW Area dependence of RTH — 4.5
RTHO Geometry independent thermal resistance — 100000
RTHW Width dependence of RTH — 3
Reference distance between OD-edge and poly from
SAREF m 1e-06
one side
Reference distance between OD-edge and poly from
SBREF m 1e-06
other side
STA2 Temperature dependence of A2 — 0
STA2O Temperature dependence of A2 — 0
STBET Temperature dependence of BETN — 1.5
STBETL Length dependence of STBET — 0
STBETLW Area dependence of STBET — 0
Geometry independent temperature dependence of
STBETO — 1.5
BETN
STBETW Width dependence of STBET — 0
STBGIDL Temperature dependence of BGIDL V/K 0
STBGIDLD Temperature dependence of BGIDL at drain side V/K 0
STBGIDLDO Temperature dependence of BGIDL at drain side V/K 0

692
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
STBGIDLO Temperature dependence of BGIDL V/K 0
Temperature dependence of CF, with same scaling as
STCF 1/K 0
CF
Temperature dependence of CF, with same scaling as
STCFL 1/K 0
CF
STCS Temperature dependence of CS — 0
STCSL Length dependence of STCS — 0
STCSLW Area dependence of STCS — 0
STCSO Temperature dependence of CS — 0
STCSW Width dependence of STCS — 0
STETAO Eta0 shift factor related to VTHO change m 0
STIG Temperature dependence of all gate currents — 0
Temperature dependence of extra gate to overlap
STIGFN — 0
current
Temperature dependence of extra gate to overlap
STIGFNO — 0
current
STIGO Temperature dependence of all gate currents — 0
STMUE Temperature dependence of MUE — 0
STMUEO Temperature dependence of MUE — 0
STRALPHA Asymmetry parameter — 3
STRDCFL DIBL variation parameter — 0
STRDVFBO Threshold shift parameter V 0
STRLAMBDA Relaxation characteristic length m 1e-07
STRRUO Mobility degradation/enhancement coefficient — 0
Saturation velocity degradation/enhancement
STRRVSAT — 0
coefficient
STRS Temperature dependence of RS — 0
STRSO Temperature dependence of RS — 0
STRTH Temperature dependence of RTH — 0
STRTHO Temperature dependence of RTH — 0
Temperature dependence of mobility
STRTRUO — 0
degradation/enhancement coefficient
STRWDVFBO Width dependence of threshold shift parameter — 0
STTHECS Temperature dependence of THECS — 0
STTHECSO Temperature dependence of THECS — 0
STTHEMU Temperature dependence of THEMU — 0
STTHEMUO Temperature dependence of THEMU — 0
STTHESAT Temperature dependence of THESAT — -0.1

693
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
STTHESATL Length dependence of STTHESAT — 0
STTHESATLW Area dependence of STTHESAT — 0
Geometry independent temperature dependence of
STTHESATO — -0.1
THESAT
STTHESATW Width dependence of STTHESAT — 0
STVFB Temperature dependence of VFB and VFBB V/K 0
STVFBL Length dependence of STVFB — 0
STVFBLW Area dependence of STVFB — 0
Geometry-independent temperature dependence of
STVFBO V/K 0
VFB and VFBB
STVFBW Width dependence of STVFB — 0
STXCOR Temperature dependence of XCOR — 0
STXCORO Temperature dependence of XCOR — 0
SWGIDL Flag for GIDL current (without=0, with=1) — 0
SWIGATE Flag for gate current (without=0, with=1) — 0
Flag for induced gate noise model (without=0,
SWIGN — 1
with=1)
SWIMPACT Flag for impact ionization current (without=0, with=1) — 0
Flag for source/drain junction asymmetry (without=0,
SWJUNASYM — 0
with=1)
SWSCALE Parameter set mode (local=0, global=1) — 0
SWSHE Flag for self heating effect (without=0, with=1) — 0
Stress model selection flag: 0=disable, 1=classical STI
SWSTRESS — 1
stress model, 2=strained channel stress model
Flag for substrate depletion model (without=0,
SWSUBDEP — 0
with=1)
TBOX Buried oxide thickness m 1e-07
TBOXO Buried oxide thickness m 1e-07
THECS Remote Coulomb scattering exponent at TR — 1.5
THECSO Remote Coulomb scattering exponent at TR — 1.5
THEMU Front channel mobility reduction exponent at TR — 1.5
THEMUO Front channel mobility reduction exponent at TR — 1.5
THERSG Transverse electric field dependence exponent of RS — 2
THERSGO Transverse electric field dependence exponent of RS — 2
THESAT Velocity saturation parameter at TR V−1 0
THESATB Back gate bias dependence of velocity saturation — 0
THESATBO Back gate bias dependence of velocity saturation — 0
THESATG Front gate bias dependence of velocity saturation — 0

694
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
THESATGO Front gate bias dependence of velocity saturation — 0
THESATL Length dependence of THESAT — 0
THESATLEXP Exponent for length dependence of THESAT — 1
THESATLW Area dependence of THESAT — 0
Geometry independent velocity saturation parameter
THESATO — 0
at TR
THESATW Width dependence of THESAT — 0
TKUO Temperature dependence of KUO — 0
TMAX Maximum self-heating temperature elevation C 150
TOXE Front gate equivalent oxide thickness m 2e-09
TOXEO Front gate equivalent oxide thickness m 2e-09
TOXP Front gate physical oxide thickness m 2e-09
TOXPO Front gate physical oxide thickness m 2e-09
TR Nominal temperature C 21
TSI Silicon or SiGe film thickness m 1e-08
TSIO Silicon or SiGe film thickness m 1e-08
TYPE Channel type parameter, +1=NMOS -1=PMOS — 1
UO Front channel zero-field mobility at TR m2 /(Vs) 0.05
VERSION Model version — 102.4
VFB Flat-band voltage of the front gate at TR V 0
VFBB Flat-band voltage of the back gate at TR V 0
Geometry-independent back gate flat-band voltage at
VFBBO V 0
TR
VFBL Length dependence of VFB V 0
VFBL2 Second order length dependence of VFB — 0
VFBLBO Roll-off back to front interface asymmetry factor — 0
VFBLEXP Exponent describing length dependence of VFB — 2
VFBLEXP2 Exponent of second order length dependence of VFB — 2
VFBLW Area dependence of VFB V 0
Geometry-independent front gate flat-band voltage at
VFBO V 0
TR
VFBW Width dependence of VFB V 0
VP CLM logarithm dependence factor V 0.05
Transverse electric field dependence of CLM
VPG — 0
logarithm factor
Transverse electric field dependence of CLM
VPGO — 0
logarithm factor

695
 Table 2-144. L_UTSOI MOSFET Device Model Parameters
Parameter Description Units Default
VPO CLM logarithm dependence factor V 0.05
WBET Characteristic width of BETN scaling m 1e-08
WKUO Width dependence of KUO — 0
WKVTHO Width dependence of KVTHO — 0
WLOD Width parameter m 0
WLODKUO Width parameter for UO stress effect — 0
WLODVTH Width parameter for VTH-stress effect — 0
WOT Effective reduction of channel width per side m 0
WVARL Length dependence of WOD — 0
Geometry-independent difference between physical
WVARO m 0
and drawn field-oxide opening
WVARW Width dependence of WOD — 0
XCOR Front channel non-universality factor — 0
XCORB Asymmetry term of non-universality factor — 1
XCORBO Asymmetry term of non-universality factor — 1
XCORL Length dependence of XCOR — 0
XCORLEXP Exponent describing length dependence of XCOR — 1
XCORLW Area dependence of XCOR — 0
XCORO Geometry-independent part of non-universality factor — 0
XCORW Width dependence of XCOR — 0
XGE Fraction of Ge content in the channel — 0
XGEO Fraction of Ge content in the channel — 0

Table 2-145. MOSFET level 10240 Output Variables

Parameter Description Units Default
type Flag for channel type none
vds Internal drain-source DC voltage (NMOS convention) V none
vsb Internal source-bulk DC voltage (NMOS convention) V none
vgs Internal gate-source DC voltage (NMOS convention) V none
vth Threshold voltage V none
Effective gate drive voltage, including back bias, drain
vth_drive V none
bias effects and self-heating
vdsat Drain saturation voltage at the given bias V none
vdsat_marg Vds voltage margin V none
id Total DC drain current flowing into drain terminal A none

696
 Table 2-145. MOSFET level 10240 Output Variables
Parameter Description Units Default
ig Total DC gate current flowing into gate terminal A none
is Total DC source current flowing into source terminal A none
ib Total DC bulk current flowing into bulk terminal A none
DC channel current, excluding tunnel, GISL and
ids A none
GIDL currents
igidl DC Gate Induced Drain Leakage current A none
igisl DC Gate Induced Source Leakage current A none
igs DC gate-source leakage current A none
igd DC gate-drain leakage current A none
idb DC drain-bulk current A none
isb DC source-bulk current A none
gm Internal DC transconductance A/V none
gmb Internal DC bulk transconductance A/V none
gds Internal DC output conductance A/V none
Internal AC gate capacitance, including overlap
cgg F none
capacitances
Internal AC gate-drain transcapacitance, including
cgd F none
overlap capacitances
Internal AC gate-source transcapacitance, including
cgs F none
overlap capacitances
cgb Internal AC gate-bulk transcapacitance F none
cdd Internal AC drain capacitance F none
cdg Internal AC drain-gate transcapacitance F none
cds Internal AC drain-source transcapacitance F none
cdb Internal AC drain-bulk transcapacitance F none
cbb Internal AC bulk capacitance F none
cbg Internal AC bulk-gate transcapacitance F none
cbs Internal AC bulk-source transcapacitance F none
cbd Internal AC bulk-drain transcapacitance F none
css Internal AC source capacitance F none
csg Internal AC source-gate transcapacitance F none
csb Internal AC source-bulk transcapacitance F none
csd Internal AC source-drain transcapacitance F none
tk MOSFET device temperature K none
MOSFET device temperature increase due to
dtsh K none
self-heating
self_gain Internal L-UTSOI model self gain none

697
 Table 2-145. MOSFET level 10240 Output Variables
Parameter Description Units Default
rout AC output resistance Ohm none
beff Gain factor in saturation A/V2 none
ft Unity gain frequency at the given bias Hz none
rgate MOS gate resistance (intrinsic input resistance) Ohm none
gmoverid Gm over Id 1/V none
vearly Equivalent Early voltage V none

698
2.3.21. Lossy Transmission Line (LTRA)

Symbol

Instance Form O<name> <A port (+) node> <A port (-) node>
+ <B port (+) node> <B port (-) node> [model name]

Model Form .MODEL <model name> LTRA R=<value> L=<value> C=<value>

+ G=<value> LEN=<value> [model parameters]

Examples Oline1 inp inn outp outn cable1

Oline2 inp inn outp outn cable1

Comments The lossy transmission line, or LTRA, device is a two port (A and B), bi-directional
device. The (+) and (-) nodes define the polarity of a positive voltage at a port.
R, L, C, and G are the resistance, inductance, capacitance, and conductance of the
transmission line per unit length, respectively. LEN is the total length of the
transmission line. Supported configurations for the LTRA are RLC, RC, LC (lossless)
and RG.
The lossy transmission line, or LTRA, device does not work with AC analysis at this
time. LTRA models will need to be replaced with lumped transmission line models
(YTRANSLINE) when used in AC analysis. The LTRA models do work correctly in
harmonic balance simulation.

Model Parameters

Table 2-146. Lossy Transmission Line Device Model Parameters

Parameter Description Units Default
ABS Abs. rate of change of deriv. for bkpt – 1
C Capacitance per unit length F/m 0
COMPACTABS special abstol for straight line checking – 1e-12
COMPACTREL special reltol for straight line checking – 0.001
do complex time step control using local truncation logical
COMPLEXSTEPCONTROL false
error estimation (T/F)
G Conductance per unit length ˙ −1 m −1 0
L Inductance per unit length Hm−1 0
LEN length of line m 0
logical
LININTERP use linear interpolation false
(T/F)

699
 Table 2-146. Lossy Transmission Line Device Model Parameters
Parameter Description Units Default
use linear interpolation if quadratic results look logical
MIXEDINTERP false
unacceptable (T/F)
don’t limit timestep size based on the time constant of logical
NOSTEPLIMIT false
the line (T/F)
logical
QUADINTERP use quadratic interpolation true
(T/F)
R Resistance per unit length ˙/m 0
REL Rel. rate of change of deriv. for bkpt – 1
limit timestep size based on the time constant of the logical
STEPLIMIT true
line (T/F)
don’t limit timestep to keep impulse response logical
TRUNCDONTCUT false
calculation errors low (T/F)
logical
TRUNCNR use N-R iterations for step calculation in LTRAtrunc false
(T/F)

By default time step limiting is on in the LTRA. This means that simulation step sizes will be reduced if
required by the LTRA to preserve accuracy. This can be disabled by setting NOSTEPLIMIT=1 and
TRUNCDONTCUT=1 on the .MODEL line.
The option most worth experimenting with for increasing the speed of simulation is REL. The default value
of 1 is usually safe from the point of view of accuracy but occasionally increases computation time. A value
greater than 2 eliminates all breakpoints and may be worth trying depending on the nature of the rest of the
circuit, keeping in mind that it might not be safe from the viewpoint of accuracy. Breakpoints may be
entirely eliminated if the circuit does not exhibit any sharp discontinuities. Values between 0 and 1 are
usually not required but may be used for setting many breakpoints.
COMPACTREL and COMPACTABS are tolerances that control when the device should attempt to compact past
history. This can significantly speed up the simulation, and reduce memory usage, but can negatively
impact accuracy and in some cases may cause problems with the nonlinear solver. In general this capability
should be used with linear type signals, such as square-wave-like voltages. In order to activate this
capability the general device option TRYTOCOMPACT=1 must be set, if it is not no history compaction will be
performed and the COMPACT options will be ignored.
Example:
.OPTIONS DEVICE TRYTOCOMPACT=1

References See references [33] and [34] for more information about the model.

700
2.3.22. Voltage- or Current-controlled Switch

Instance Form S<name> <(+) switch node> <(-) switch node>

+ <(+) control node> <(-) control node>
+ <model name> [ON] [OFF]

W<name> <(+) switch node> <(-) switch node>

+ <control node voltage source>
+ <model name> [ON] [OFF]

Model Form .MODEL <model name> VSWITCH [model parameters]

.MODEL <model name> ISWITCH [model parameters]

Examples S1 21 23 12 10 SMOD1
SSET 15 10 1 13 SRELAY
W1 1 2 VCLOCK SWITCHMOD1
W2 3 0 VRAMP SM1 ON

Comments The voltage- or current-controlled switch is a particular type of controlled resistor.

This model is designed to help reduce numerical issues. See Special Considerations
below.
The resistance between the <(+) switch node> and the <(-) switch node> is
dependent on either the voltage between the <(+) control node> and the <(-)
control node> or the current through the control node voltage source. The resistance
changes in a continuous manner between the RON and ROFF model parameters.
No resistance is inserted between the control nodes. It is up to the user to make sure
that these nodes are not floating.
Even though evaluating the switch model is computationally inexpensive, for transient
analysis Xyce steps through the transition section using small time-steps in order to
calculate the waveform accurately. Thus, a circuit with many switch transitions can
result in lengthy run times.
The ON and OFF parameters are used to specify the initial state of the switch at the
first step of the operating point calculation; this does not force the switch to be in that
state, it only gives the operating point solver an initial state to work with. If it is known
that the switch should be in a particular state in the operating point it could help
convergence to specify one of these keywords.
The power dissipated in the switch is calculated with 𝐼 · Δ𝑉 where the voltage drop is
calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− . This will essentially
be the power dissipated in either RON or ROFF, since the switch is a particular type of
controlled resistor.
Note: The voltage- and current-controlled switches specified in this manner are
converted at parse time into equivalent “generic” switches.

701
Model Parameters

Table 2-147. Controlled Switch Device Model Parameters

Parameter Description Units Default
IOFF Off current A 0
ION On current A 0.001
OFF Off control value – 0
ON On control value – 1
ROFF Off resistance ˙ 1e+06
RON On resistance ˙ 1
VOFF Off voltage V 0
VON On voltage V 1

Special Considerations
• Due to numerical limitations, Xyce can only manage a dynamic range of approximately 12 decades.
Thus, it is recommended the user limit the ratio ROFF/RON to less than 1012 . This soft limitation is
not enforced by the code, and larger ratios might converge for some problems.
• Do not set RON to 0.0, as the code computes the “on” conductance as the inverse of RON. Using 0.0
will cause the simulation to fail when this invalid division results in an infinite conductance. Use a
very small, but non-zero, on resistance instead.
• Furthermore, it is a good idea to limit the narrowness of the transition region. This is because in the
transition region, the switch has gain and the narrower the region, the higher the gain and the more
potential for numerical problems. The smallest value recommended for ∥VON − VOFF∥ or
∥ION − IOFF∥ is 1 × 10−12 . This recommendation is not a restriction, and you might find for some
problems that narrower transition regions might work well.

Controlled switch equations The equations in this section use the following variables:
𝑅𝑠 = switch resistance
𝑉𝑐 = voltage across control nodes
𝐼𝑐 = current through control node voltage source √ 
𝐿 𝑚 = log-mean of resistor values = ln RON · ROFF
𝐿 𝑟 = log - ratio of resistor values = ln (RON/ROFF)
𝑉𝑑 = difference of control voltages = VON − VOFF
𝐼 𝑑 = difference of control currents = ION − IOFF

Switch Resistance To compute the switch resistance, Xyce first calculates the “switch state” 𝑆 as
𝑆 = (𝑉𝑐 − VOFF)/𝑉𝑑 or 𝑆 = (𝐼𝑐 − IOFF)/𝐼 𝑑 . The switch resistance is then:

 RON,


 𝑆 ≥ 1.0
𝑅𝑠 = ROFF, 𝑆 ≤ 0.0
 exp 𝐿 𝑚 + 0.75𝐿 𝑟 (2𝑆 − 1) − 0.25𝐿 𝑟 (2𝑆 − 1) 3
, 0 < 𝑆 < 1



702
2.3.23. Generic Switch

Instance Form S<name> <(+) switch node> <(-) switch node> <model name> [ON]
[OFF] <control = expression >

Model Form .MODEL <model name> VSWITCH [model parameters]

.MODEL <model name> ISWITCH [model parameters]
.MODEL <model name> SWITCH [model parameters]

Examples S1 1 2 SWI OFF CONTROL={I(VMON)}

SW2 1 2 SWV OFF CONTROL={V(3)-V(4)}
S3 1 2 SW OFF CONTROL={if(time>0.001,1,0)}

Comments The generic switch is similar to the voltage- or current-controlled switch except that
the control variable is anything that can be writen as an expression. The examples
show how a voltage- or current-controlled switch can be implemented with the generic
switch. Also shown is a relay that turns on when a certain time is reached. Model
parameters are given in Table 2-147.
The voltage- and current-controlled switch syntaxes are converted at parse time to their
equivalent generic device, and so all three variants in fact use the same code internally.
The power dissipated in the generic switch is calculated with 𝐼 · Δ𝑉 where the voltage
drop is calculated as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− . This will
essentially be the power dissipated in either RON or ROFF, since the generic switch is a
particular type of controlled resistor.

703
2.3.24. Linear device

The linear (YLIN) device allows an S-, Y-, or Z-parameter model to be used to define an N-port device. It is
mostly commonly used as part of a Harmonic Balance (HB) analysis.

Instance Form YLIN <name> <(+) node> <(-) node> [model name]

Model Form .MODEL <model name> LIN [model parameters]

Examples YLIN YLIN1 1 0 2 0 YLIN_MOD1

.MODEL YLIN_MOD1 LIN TSTONEFILE=yparams.y2p

Parameters and
Options model name
Name of the model defined in a .MODEL line.

Comments At present, the YLIN device is only supported in the frequency domain for HB
analyses.

Model Parameters

Table 2-148. LIN Device Model Parameters

Parameter Description Units Default
Touchstone file contains frequency-domain logical
ISC_FD false
short-circuit current data (T/F)
TSTONEFILE Touchstone File Name – ”
Interpolation method. Supported methods:
• 1 (linear)
INTERPOLATION – 1
• 2 (akima spline)

The Touchstone file name must be specified. The YLIN device accepts both Touchstone 1 and Touchstone 2
formatted input files [7].
For coupling with EM codes, such as EIGER, the YLIN device also accepts a non-standard version of the
Touchstone input files. If the ISC_FD model parameter is set to true then each row of “network data” in the
input file also contains additional columns with the “per-port frequency-domain short-circuit currents”.
There are then two such additional columns for each port. The format (RI, MA or DB) of those additional
columns will be as specified by the Option line in the Touchstone file. In this non-standard case, only the
“Full” matrix format is supported for Touchstone 2 input files.

704
2.3.25. Lossless (Ideal) Transmission Line

Symbol

Instance Form T<name> <port 1 (+) node> <port 1 (-) node>

+ <port 2 (+) node> <port 2 (-) node>
+ Z0=<value> [TD=<value>] [F=<value> [NL=<value>]]

Examples Tline inp inn outp outn Z0=50 TD=1us

Tline2 inp inn outp outn Z0=50 F=1meg NL=1.0

Comments The lossless transmission line device is a two port (A and B), bi-directional delay line.
The (+) and (-) nodes define the polarity of a positive voltage at a port.
Z0 is the characteristic impedance. For user convenience, ZO (“Zee Oh”) is an allowed
synonym for Z0 (“Zee Zero”).
The transmission line’s length is specified by either TD (a delay in seconds) or by the
combination of F and NL (a frequency in Hz and the relative wavelength at F). NL
defaults to 0.25 (F is the quarter-wave frequency). If F is given, the time delay is
computed as 𝑁𝐹𝐿 .
While both TD and F are optional, at least one of them must be given. It is an instance
line error if both are given.
Lead currents for the two terminals (1 and 2) of the lossless transmission device
(e.g.,for the T device line2) are accessed via I1(Tline2) and I2(Tline2). The
polarity conventions are that positive current flows into the positive node of the
specified terminal, and negative current flows out of the positive node of the specified
terminal.
Power for the lossless transmission line is calculated as 𝐼1 · Δ𝑉1 + 𝐼2 · Δ𝑉2 , where the
voltage drops (Δ𝑉1 and Δ𝑉2 ) are the voltage drops between the positive and negative
terminals of each port (e.g., Δ𝑉 = (𝑉+ − 𝑉− )). The sign conventions for the lead
currents 𝐼1 and 𝐼2 were given in the previous paragraph. This definition can be viewed
as the instantaneous sum of the power flowing into terminal 1 and the power flowing
into terminal 2. This definition for power for the lossless transmission line may differ
from commercial simulators, such as HSPICE.
The lossless transmission line device does not work with AC analysis at this time.
Lossless transmission line models will need to be replaced with lumped transmission
line models (YTRANSLINE) when used in AC analysis. The lossless transmission
line does work correctly in harmonic balance simulation.

Instance Parameters

705
 Table 2-149. Ideal Transmission Line Device Instance Parameters
Parameter Description Units Default
F Frequency Hz 0
NL Length in wavelengths – 0.25
TD Time delay s 0
Z0 Characteristic Impedance ˙ 0
ZO Characteristic Impedance ˙ 0

706
2.3.26. Lumped Transmission Line

Symbol

Instance Form ytransline <name> <Input port> <Output port> testLine

+ len=<value> lumps=<value>

Model Form .model testLine transline r=<value> l=<value>

+ c=<value> [model parameters]

Examples ytransline line1 inn out testLine len=12.0 lumps=1440

Comments The lumped transmission line, device is a two port bi-directional device. The
specification is patterned, loosely, from the netlist specification for the LTRA device.
R, L, and C are the resistance, inductance, and capacitance of the transmission line per
unit length, respectively. LEN is the total length of the transmission line, and LUMPS is
the number of lumped elements used to discretize the line. Supported configurations
for this device are RLC and LC.
Unlike the LTRA device, which is based on an analytic solution, this device is based
on assembling chains of linear R,L and C devices to approximate the solution to the
Telegraph equations. It is the functional equivalent of building a transmission line in
the netlist using subcircuits of linear elements. The advantage of using this approach is
that it automates the mechanics of this process, and thus is less prone to error. It can be
used with all analysis types, including harmonic balance (HB).
The model is based on the assumption that the segments of the line are evenly spaced.
The number of segments is specified by the parameter LUMPS and the larger this
number, the more accurate the calculation.

Device Parameters

Table 2-150. Lumped Transmission Line Device Instance Parameters

Parameter Description Units Default
LEN length of line m 0
LUMPS – 1

Model Parameters

707
 Table 2-151. Lumped Transmission Line Device Model Parameters
Parameter Description Units Default
C Capacitance per unit length F/m 0
ELEV – 2
G Conductance per unit length ˙ −1 m−1 0
L Inductance per unit length Hm−1 0
R Resistance per unit length ˙/m 0

708
2.3.27. Ideal Delay

An ideal delay device, operating in a manner similar to a voltage-controlled voltage source, is provided by
the YDELAY device.

Instance Form YDELAY <name> <positive node> <negative node>

+ <positive control node> <negative control node>
+ TD=<time delay>
+ [EXTRAPOLATION=<true|false>] [BPENABLED=<true|false>]
+ [LINEARINTERP=<true|false>]

Examples YDELAY delay1 2 0 1 0 TD=10N

R1 2 0 1
YDELAY delay1 3 0 2 0 TD=10N LINEARINTERP=true
R2 3 0 1
YDELAY delay1 4 0 3 0 TD=10N BPENABLED=FALSE
R4 4 0 1
YDELAY delay1 5 0 4 0 TD=10N EXTRAPOLATION=false
R5 5 0 1

Comments The voltage between the positive and negative control nodes is reproduced at the
positive and negative output nodes delayed by a time equal to the specified TD
parameter.

The device is equivalent in connectivity to a voltage-controlled voltage source — the

device puts no load on the control nodes, and its output must be connected to a valid
circuit.

Unlike the transmission line, no impedance matching is required, and reflections due
to impedance mismatch do not occur.

These devices may be chained to create outputs at different delays, but each instance
must have its output connected to a valid closed circuit. The examples above are
chained correctly so that each of the output nodes is delayed by 10 nanoseconds from
the previous stage.

The device functions by storing a history of its input at each accepted time point. At
each new time step, interpolation is performed on this history to determine what the
signal would have been at a time TD in the past. At each step, the device checks its
history to determine if the previous three saved steps include a discontinuity in the
input. If so, the device assures that Xyce will correctly resolve the same discontinuity
when it appears on the output.

With no special options specified, three-point quadratic interpolation is used except

after a discontinuity, when linear interpolation is performed. If Xyce has advanced the
time by more than TD and no discontinuity has occured, then this interpolation is
actually extrapolation.

709
 When LINEARINTERP=true is specified, the history interpolation used is always
linear interpolation.
When EXTRAPOLATION=false, Xyce will never attempt extrapolation when it has
taken a time step larger than TD. In this case, the current, unconverged value of the
solution is used as the third interpolation point and the interpolation is recomputed at
every step of the nonlinear solve.
When BPENABLED=false, the device will not set a simulation breakpoint to force the
time integrator to stop exactly TD seconds after a detected discontinuity on the input.
It will still force a maximum time step on the time integrator after such a discontinuity,
and other techniques will be applied to assure the discontinuity is resolved. This
option may result in Xyce rejecting a lot more time steps and slower simulation than
when it is left at its default.

2.3.27.1. Delay device instance parameters

The instance parameters for the delay device are shown in Table 2-152.

Table 2-152. Delay element Device Instance Parameters

Parameter Description Units Default
BPENABLED Can this device set discontinuity breakpoints? – true
EXTRAPOLATION Can this device use extrapolation on history? – true
Should this device use only linear interpolation on
LINEARINTERP – false
history?
TD Time delay s 0

710
2.3.28. Behavioral Digital Devices

Instance Form U<name> <type>(<num inputs>) [digital power node]

+ [digital ground node] <input node>* <output node>*
+ <model name> [device parameters]

Model Form .MODEL <model name> DIG [model parameters]

Examples UMYAND AND(2) DPWR DGND in1 in2 out DMOD IC=TRUE
UTHEINV INV DPWR DGND in out DMOD
.model DMOD DIG (
+ CLO=1e-12 CHI=1e-12
+ S0RLO=5 S0RHI=5 S0TSW=5e-9
+ S0VLO=-1 S0VHI=1.8
+ S1RLO=200 S1RHI=5 S1TSW=5e-9
+ S1VLO=1 S1VHI=3
+ RLOAD=1000
+ CLOAD=1e-12
+ DELAY=20ns )

Parameters and
Options type
Type of digital device. Supported devices are: INV, BUF, AND, NAND, OR,
NOR, XOR, NXOR, DFF, JKFF, TFF, DLTCH and ADD. (Note: NOT is an
allowed synonym for INV, but will be deprecated in future Xyce releases.)
The following gates have a fixed number of inputs. INV and BUF have only
one input and one output node. XOR and NXOR have two inputs and one
output. ADD has three inputs (in1, in2, carryIn) and two outputs (sumOut
and carryOut). DFF has four inputs (PREB, CLRB, Clock and Data) and two
¯ TFF has two inputs (T and CLK) and two outputs (𝑄 and
outputs (𝑄 and 𝑄).
¯ The TFF uses “positive” (“rising”) edge clocking. The JKFF has five
𝑄).
inputs (PREB, CLRB, Clock, J and K) and two outputs (𝑄 and 𝑄). ¯ The JKFF
uses “negative” (“falling”) edge clocking. DLTCH has four inputs (PREB,
CLRB, Enable and Data) and two outputs (𝑄 and 𝑄). ¯
The AND, NAND, OR and NOR gates have one output but a variable
number of inputs. There is no limit on the number of inputs for AND,
NAND, OR and NOR gates, but there must be at least two inputs.
num inputs
For AND, NAND, OR and NOR gates, with N inputs, the syntax is (N), as
shown for the MYAND example given above, where AND(2) is specified.
The inclusion of (N) is mandatory for gates with a variable number of inputs,
and both the left and right parentheses must be used to enclose N.
This parameter is optional, and typically omitted, for gates with a fixed
number of inputs, such as INV, BUF, XOR, NXOR, DFF, JKFF, TFF,

711
 DLTCH and ADD. This is illustrated by the THEINV example given above,
where the device type is INV rather than INV(1).
digital power node
Dominant node to be connected to the output node(s) to establish high output
state. This node is connected to the output by a resistor and capacitor in
parallel, whose values are set by the model. This node must be specified on
the instance line.
digital ground node
This node serves two purposes, and must be specified on the instance line. It
is the dominant node to be connected to the output node(s) to establish low
output state. This node is connected to the output by a resistor and capacitor
in parallel, whose values are set by the model. This node is also connected to
the input node by a resistor and capacitor in parallel, whose values are set by
the model. Determination of the input state is based on the voltage drop
between the input node and this node.
input nodes, output nodes
Input and output nodes that connect to the circuit.
model name
Name of the model defined in a .MODEL line.
device parameters
Parameter listed in Table 2-153 may be provided as <parameter>=<value>
specifications as needed. For devices with more than one output, multiple
output initial states may be provided as Boolean values in either a comma
separated list (e.g. IC=TRUE,FALSE for a device with two outputs) or
individually (e.g. IC1=TRUE IC2=FALSE or IC2=FALSE). Finally, the IC
specification must use TRUE and FALSE rather than T and F.

Device Parameters

Table 2-153. Behavioral Digital Device Instance Parameters

Parameter Description Units Default
logical
IC1 Vector of initial values for output(s) false
(T/F)
IC2 – false

Model Parameters

Table 2-154. Behavioral Digital Device Model Parameters

Parameter Description Units Default
CHI Capacitance between output node and high reference F 1e-06
CLO Capacitance between output node and low reference F 1e-06
CLOAD Capacitance between input node and input reference F 1e-06

712
 Table 2-154. Behavioral Digital Device Model Parameters
Parameter Description Units Default
DELAY Delay time of device s 1e-08
RLOAD Resistance between input node and input reference ˙ 1000
Low state resitance between output node and high
S0RHI ˙ 100
reference
Low state resistance between output node and low
S0RLO ˙ 100
reference
S0TSW Switching time transition to low state s 1e-08
S0VHI Maximum voltage to switch to low state V 1.7
S0VLO Minimum voltage to switch to low state V -1.5
High state resistance between output node and high
S1RHI ˙ 100
reference
High state resistance between output node and low
S1RLO ˙ 100
reference
S1TSW Switching time transition to high state s 1e-08
S1VHI Maximum voltage to switch to high state V 7
S1VLO Minimum voltage to switch to high state V 0.9

Model Description The input interface model consists of the input node connected with a resistor and
capacitor in parallel to the digital ground node. The values of these are: RLOAD and CLOAD.

The logical state of any input node is determined by comparing the voltage relative to the reference to the
range for the low and high state. The range for the low state is S0VLO to S0VHI. Similarly, the range for
the high state is S1VLO to S1VHI. The state of an input node will remain fixed as long as its voltage stays
within the range for its current state. That input node will transition to the other state only when its state
goes outside the voltage range of its current state.

The output interface model is more complex than the input model, but shares the same basic configuration
of a resistor and capacitor in parallel to simulate loading. For the output case, there are such parallel RC
connections to two nodes, the digital ground node and the digital power node. Both of these nodes must be
specified on the instance line.

The capacitance to the high node is specified by CHI, and the capacitance to the low node is CLO. The
resistors in parallel with these capacitors are variable, and have values that depend on the state. In the low
state (S0), the resistance values are: S0RLO and S0RHI. In the high state (S1) ,the resistance values are:
S1RLO and S1RHI. Transition to the high state occurs exponentially over a time of S1TSW, and to the
low state S0TSW.

The device’s delay is given by the model parameter DELAY. Any input changes that affect the device’s
outputs are propagated after this delay.

As a note, the model parameters VREF, VLO and VHI are used by the now deprecated Y-type digital
device, but are ignored by the U device. A warning message is emitted if any of these three parameters are
used in the model card for a U device.

713
Another caveat is that closely spaced input transitions to the Xyce digital behavioral models may not be
accurately reflected in the output states. In particular, input-state changes spaced by more than DELAY
seconds have independent effects on the output states. However, two input-state changes (S1 and S2) that
occur within DELAY seconds (e.g., at time=t1 and time=t1+0.5*DELAY) have the effect of masking the effects
of S1 on the device’s output states, and only the effects of S2 are propagated to the device’s output states.

DCOP Calculations for Flip-Flops and Latches The behavior of the digital devices during the DC
Operating Point (DCOP) calculations can be controlled via the IC1 and IC2 instance parameters and the
DIGINITSTATE device option. See 2.1.25 for more details on the syntax for device options. Also, this
section applies to the Y-Type Behavioral Digital Devices discussed in 2.3.29.

The IC1 instance parameter is supported for all gate types. The IC2 instance parameter is supported for all
gate types that have two outputs. These instance parameters allow the outputs of individual gates to be set to
known states (either TRUE (1) or FALSE (0)) during the DCOP calculation, irregardless of their input
state(s). There are two caveats. First, the IC1 and IC2 settings at a given gate will override the global effects
of the DIGINITSTATE option, discussed below, at that gate. Second, IC1 and IC2 do not support the X, or
“undetermined”, state discussed below.

The DIGINITSTATE option only applies to the DLTCH, DFF, JKFF and TFF devices. It was added for
improved compatibility with PSpice. It sets the initial state of all flip-flops and latches in the circuit:
0=clear, 1=set, 2=X. At present, the use of the DIGINITSTATE option during the DCOP is the only place
that Xyce supports the X, or “undetermined”, state. The X state is modeled in Xyce by having the DLTCH,
DFF, JKFF and TFF outputs simultaneously “pulled-up” and “pulled-down”. That approach typically
produces an output level, for the X state, that is approximately halfway between the voltage levels for TRUE
and FALSE (e.g., halfway between V_HI and V_LO). As mentioned above, the IC1 and IC2 instance
parameters take precedence at a given gate.

Xyce also supports a default DIGINITSTATE, whose value is 3. For this default value, for the DFF, JKFF,
TFF and DLTCH devices, Xyce enforces 𝑄 and 𝑄¯ being different at DCOP, if both PREB and CLRB are TRUE
. The behavior of the DFF, JKFF and DLTCH devices at the DCOP for DIGINITSTATE=3 is shown in
Tables 2-155, 2-157 and 2-156. In these three tables, the 𝑋 state denotes the “Don’t Care” condition,
where the input state can be 0, 1 or the “undetermined” state. The first row in each truth-table (annotated
with ∗) is “unstable”, and will change to a state with 𝑄 and 𝑄¯ being different once both PREB and CLRB are
not both in the FALSE state.

The behavior of the TFF device at the DCOP, for the default DIGINITSTATE of 3, is simpler, and is not
shown as a table. The design decision was to have 𝑄 and 𝑄¯ be different, with the 𝑄 value equal to the state
of the 𝑇 input.

Table 2-155. DFF Truth-Table for DIGINITSTATE=3

PREB CLRB CLOCK DATA 𝑄 𝑄¯ (Qbar)
0 0 X X 1* 1*
0 1 X X 1 0
1 0 X X 0 1
1 1 X 0 0 1
1 1 X 1 1 0

714
 Table 2-156. DLTCH Truth-Table for DIGINITSTATE=3
PREB CLRB ENABLE DATA 𝑄 𝑄¯ (Qbar)
0 0 X X 1* 1*
0 1 X X 1 0
1 0 X X 0 1
1 1 X 0 0 1
1 1 X 1 1 0

Table 2-157. JKFF Truth-Table for DIGINITSTATE=3

PREB CLRB CLOCK J K 𝑄 𝑄¯ (Qbar)
0 0 X X X 1* 1*
0 1 X X X 1 0
1 0 X X X 0 1
1 1 X 0 X 0 1
1 1 X 1 X 1 0

715
2.3.29. Y-Type Behavioral Digital Devices (Deprecated)

Instance Form Y<type> <name> [low output node] [high output node]
+ [input reference node] <input node>* <output node>*
+ <model name> [device parameters]

Model Form .MODEL <model name> DIG [model parameters]

Examples YAND MYAND in1 in2 out DMOD IC=TRUE

YNOT THENOT in out DMOD
YNOR ANOR2 vlo vhi vref in1 in2 out DDEF
.model DMOD DIG (
+ CLO=1e-12 CHI=1e-12
+ S0RLO=5 S0RHI=5 S0TSW=5e-9
+ S0VLO=-1 S0VHI=1.8
+ S1RLO=200 S1RHI=5 S1TSW=5e-9
+ S1VLO=1 S1VHI=3
+ RLOAD=1000
+ CLOAD=1e-12
+ VREF=0 VLO=0 VHI=3
+ DELAY=20ns )
.MODEL DDEF DIG

Parameters and
Options type
Type of digital device. Supported devices are: NOT, BUF, AND, NAND, OR,
NOR, XOR, NXOR, DFF, JKFF, TFF, DLTCH and ADD. (Note: INV is now
the preferred synonym for NOT. The NOT device type will be deprecated in
future Xyce releases.) For Y-type digital devices, all devices have two input
nodes and one output node, except for NOT, DFF and ADD. NOT has one
input and one output. ADD has three inputs (in1, in2, carryIn) and two
outputs (sumOut and carryOut). DFF has four inputs (PREB, CLRB, Clock
and Data) and two outputs (𝑄 and 𝑄).¯ TFF has two inputs (T and Clock) and
¯ The TFF uses “positive” (“rising”) edge clocking.
two outputs (𝑄 and 𝑄).
The JKFF has five inputs (PREB, CLRB, Clock, J and K) and two outputs (𝑄
¯ The JKFF uses “negative” (“falling”) edge clocking. DLTCH has
and 𝑄).
four inputs (PREB, CLRB, Enable and Data) and two outputs (𝑄 and 𝑄). ¯
name
Name of the device instance. This must be present, and when combined with
the Y<type>, must be unique in the netlist. In the examples, MYAND,
THENOT and ANOR2 have been used as names for the three devices.
low output node
Dominant node to be connected to the output node(s) to establish low output
state. This node is connected to the output by a resistor and capacitor in

716
 parallel, whose values are set by the model. If specified by the model, this
node must be omitted from the instance line and a fixed voltage VLO is used
instead.
high output node
Dominant node to be connected to the output node(s) to establish high output
state. This node is connected to the output by a resistor and capacitor in
parallel, whose values are set by the model. If specified by the model, this
node must be omitted from the instance line and a fixed voltage VHI is used
instead.
input reference node
This node is connected to the input node by a resistor and capacitor in
parallel, whose values are set by the model. Determination if the input state
is based on the voltge drop between the input node and this node. If specified
by the model, this node must be omitted from the instance line and a fixed
voltage VREF is used instead.
input nodes, output nodes
Nodes that connect to the circuit.
model name
Name of the model defined in a .MODEL line.
device parameters
Parameter listed in Table 2-158 may be provided as <parameter>=<value>
specifications as needed. For devices with more than one output, multiple
output initial states may be provided as Boolean values in either a comma
separated list (e.g. IC=TRUE,FALSE for a device with two outputs) or
individually (e.g. IC1=TRUE IC2=FALSE or IC2=FALSE). Finally, the IC
specification must use TRUE and FALSE rather than T and F.

Device Parameters

Table 2-158. Behavioral Digital Device Instance Parameters

Parameter Description Units Default
logical
IC1 Vector of initial values for output(s) false
(T/F)
IC2 – false

Model Parameters

Table 2-159. Behavioral Digital Device Model Parameters

Parameter Description Units Default
CHI Capacitance between output node and high reference F 1e-06
CLO Capacitance between output node and low reference F 1e-06
CLOAD Capacitance between input node and input reference F 1e-06

717
 Table 2-159. Behavioral Digital Device Model Parameters
Parameter Description Units Default
DELAY Delay time of device s 1e-08
RLOAD Resistance between input node and input reference ˙ 1000
Low state resitance between output node and high
S0RHI ˙ 100
reference
Low state resistance between output node and low
S0RLO ˙ 100
reference
S0TSW Switching time transition to low state s 1e-08
S0VHI Maximum voltage to switch to low state V 1.7
S0VLO Minimum voltage to switch to low state V -1.5
High state resistance between output node and high
S1RHI ˙ 100
reference
High state resistance between output node and low
S1RLO ˙ 100
reference
S1TSW Switching time transition to high state s 1e-08
S1VHI Maximum voltage to switch to high state V 7
S1VLO Minimum voltage to switch to high state V 0.9
VHI Internal high state supply voltage V 0
VLO Internal low state supply voltage V 0
VREF Internal reference voltage for inputs V 0

Model Description The input interface model consists of the input node connected with a resistor and
capacitor in parallel to the digital ground node. The values of these are: RLOAD and CLOAD.

The logical state of any input node is determined by comparing the voltage relative to the reference to the
range for the low and high state. The range for the low state is S0VLO to S0VHI. Similarly, the range for
the high state is S1VLO to S1VHI. The state of an input node will remain fixed as long as its voltage stays
within the voltage range for its current state. That input node will transition to the other state only when its
state goes outside the range of its current state.

The output interface model is more complex than the input model, but shares the same basic configuration
of a resistor and capacitor in parallel to simulate loading. For the output case, there are such connections to
two nodes, the digital ground node and the digital power node. Both of these nodes must be specified on the
instance line.

The capacitance to the high node is specified by CHI, and the capacitance to the low node is CLO. The
resistors in parallel with these capacitors are variable, and have values that depend on the state. In the low
state (S0), the resistance values are: S0RLO and S0RHI. In the high state (S1) ,the resistance values are:
S1RLO and S1RHI. Transition to the high state occurs exponentially over a time of S1TSW, and to the
low state S0TSW.

The device’s delay is given by the model parameter DELAY. Any input changes that affect the device’s
outputs are propagated after this delay.

718
Another caveat is that closely spaced input transitions to the Xyce digital behavioral models may not be
accurately reflected in the output states. In particular, input-state changes spaced by more than DELAY
seconds have independent effects on the output states. However, two input-state changes (S1 and S2) that
occur within DELAY seconds (e.g., at time=t1 and time=t1+0.5*DELAY) have the effect of masking the effects
of S1 on the device’s output states, and only the effects of S2 are propagated to the device’s output states.

DCOP Calculations for Flip-Flops and Latches The behavior of the digital devices during the DC
Operating Point (DCOP) calculations can be controlled via the IC1 and IC2 instance parameters and the
DIGINITSTATE device option. See 2.3.28 and 2.1.25 for more details on these instance parameters and
device option.

Converting Y-Type Digital Devices to U-Type Digital Devices Xyce is migrating the digital
behavioral devices to U devices. The goal is increased compatibility with PSpice netlists. This subsection
gives four examples of how to convert an existing Xyce netlist using Y-type digital devices to the
corresponding U device syntaxes. The conversion process depends on whether the device has a fixed
number of inputs or a variable number of inputs. In all cases, the the model parameters VREF, VLO and
VHI should be omitted from the U device model card. For U devices, the nodes vlo and vhi are always
specified on the instance line.

Example 1: Fixed number of inputs, Y-device model card contains VREF, VLO and VHI. Assume
VREF=VLO.

YNOT THENOT in out DMOD

.model DMOD DIG (
+ CLO=1e-12 CHI=1e-12
+ S0RLO=5 S0RHI=5 S0TSW=5e-9
+ S0VLO=-1 S0VHI=1.8
+ S1RLO=200 S1RHI=5 S1TSW=5e-9
+ S1VLO=1 S1VHI=3
+ RLOAD=1000
+ CLOAD=1e-12
+ VREF=0 VLO=0 VHI=3
+ DELAY=20ns )

* Digital power node. Assume digital ground node = GND

V1 DPWR 0 3V
UTHENOT INV DPWR 0 in out DMOD1
.model DMOD1 DIG (
+ CLO=1e-12 CHI=1e-12
+ S0RLO=5 S0RHI=5 S0TSW=5e-9
+ S0VLO=-1 S0VHI=1.8
+ S1RLO=200 S1RHI=5 S1TSW=5e-9
+ S1VLO=1 S1VHI=3
+ RLOAD=1000
+ CLOAD=1e-12
+ DELAY=20ns )

719
Example 2: Fixed number of inputs, Y-device instance line contains vlo, vhi and vref. Assume vref=vlo.

YNOT THENOT vlo vhi vref in out DMOD1

UTHENOT INV vhi vlo in out DMOD1

Example 3: Variable number of inputs, Y-device model card contains VREF, VLO and VHI. Assume
VREF=VLO.

YAND MYAND in1 in2 out DMOD

UMYAND AND(2) DPWR 0 in1 in2 out DMOD1

Example 4: Variable number of inputs, Y-device instance line contains vlo, vhi and vref. Assume
vref=vlo.

YAND MYAND vlo vhi vref in1 in2 out DMOD1

UMYAND AND(2) vhi vlo in1 in2 out DMOD1

720
2.3.30. Accelerated mass

Simulation of electromechanical devices or magnetically driven machines may require that Xyce simulate
the movement of an accelerated mass, that is, to solve the second order initial value problem

𝑑2𝑥
= 𝑎(𝑡)
𝑑𝑡
𝑥(0) = 𝑥 0
𝑥¤0 = 𝑣0

where 𝑥 is the position of the object, 𝑥¤ its velocity, and 𝑎(𝑡) the acceleration. In Xyce, this simulation
capability is provided by the accelerated mass device.

Instance Form YACC <name> <acceleration node> <velocity node> <position node>
+ [v0=<initial velocity>] [x0=<initial position>]

Examples * Simulate a projectile thrown upward against gravity

V1 acc 0 -9.8
R1 acc 0 1
YACC acc1 acc vel pos v0=10 x0=0
.print tran v(pos)
.tran 1u 10s
.end

* Simulate a damped, forced harmonic oscillator

* assuming K, c, mass, amplitude and frequency
* are defined in .param statements
B1 acc 0 V={(-K * v(pos) - c*v(vel))/mass
+ + amplitude*sin(frequency*TIME)}
R1 acc 0 1
YACC acc2 acc vel pos v0=0 x0=0.4
.print tran v(pos)
.tran 1u 10s
.end

Comments When used as in the examples, Xyce will emit warning messages about the pos and
vel nodes not having a DC path to ground. This is normal and should be ignored. The
position and velocity nodes should not be connected to any real circuit elements. Their
values may, however, be used in behavioral sources; this is done in the second example.

721
2.3.31. Power Grid

The Power Grid devices are a family of device models that can be used to model steady-state power flow in
electric power grids. They include device models for branches, bus shunts, transformers and generator
buses.
Power flow in electric power grids can be modeled as a complex-valued voltage-current problem with
standard admittance-matrix techinques. This approach solves the system of equations 𝐼 = 𝑌𝑉, and is termed
IV format in this document. However, it is more typically modeled as a power-flow problem that solves the
system of equations 𝑆 = 𝑃 + 𝑗𝑄 = 𝑉 𝐼 ∗ , where 𝑆 is the complex power flow, 𝑉 and 𝐼 are complex-valued
quantities, and 𝐼 ∗ is the complex conjugate of 𝐼. The complex power flow can then be solved in either
rectangular or polar coordinates. These two solution formats are termed PQ Rectangular (aka, PQR format)
and PQ Polar (aka, PQP format) in this document. The variables for each solution format are described in
more detail in the device descriptions given below.
In all three formulations,an Equivalent Real Formulation (ERF) [35] must be used for compatibility with
the existing solver libraries in Xyce. More details on these equations are given below after the individual
device descriptions.

PowerGridBranch

Instance Form Y<type> <name> <input node1> <output node1>

+ <input node2> <output node2> [device parameters]

Examples YPowerGridBranch pg1_2 VR1 VR2 VI1 VI2 AT=IV R=0.05 B=0.1 X=0.05
YPGBR pg1_2a VR1 VR2 VI1 VI2 AT=IV R=0.05 B=0.1 X=0.05
YPowerGridBranch pg1_2b VR1 VR2 VI1 VI2 AT=PQR R=0.05 B=0.1 X=0.05
YPGBR pg1_2c VR1 VR2 VI1 VI2 AT=PQR R=0.05 B=0.1 X=0.05
YPowerGridBranch pg1_2d Th1 Th2 VM1 VM2 AT=PQP R=0.05 B=0.1 X=0.05
YPGBR pg1_2e Th1 Th2 VM1 VM2 AT=PQP R=0.05 B=0.1 X=0.05

Parameters and
Options type
The device type has a verbose (PowerBranchBranch) and a shortened
(PGBR) form. Their usage may be mixed within a netlist.
name
Name of the device instance. This must be present, and unique amongst the
PowerGridBranch devices in the netlist.
input node
There are two input nodes, <input node1> and <input node2>, whose
definitions depend on the AnalysisType (AT) specified. Both nodes must be
specified. This device can be viewed as a generalized 4-port resistor, using
the Equivalent Real Form (ERF) described below in the equation
subsections. For IV and PQR formats, <input node1> is the real part (VR)
of the voltage at terminal 1 while <input node2> is the imaginary part (VI)
of the voltage at terminal 1. For PQP format, <input node1> is the angle (Θ

722
 or Th) of the voltage at terminal 1 while <input node2> is the magnitude
(VM or |𝑉 |) of the voltage at terminal 1. Finally, by analogy to other Xyce
devices, node 1 can be considered as the positive terminal for this device,
while node 2 is the negative terminal.
output node
There are two output nodes, <output node1> and <output node2>,
whose definitions depend on the AnalysisType (AT) specified. Both nodes
must be specified. This device can be viewed as a generalized 4-port resistor,
using the ERF described below in the equation subsections. For IV and PQR
formats, <output node1> is the real part (VR) of the voltage at terminal 2
while <output node2> is the imaginary part (VI) of the voltage at terminal
2. For PQP format, <output node1> is the angle (Θ or Th) of the voltage at
terminal 2 while <output node2> is the magnitude (VM or |𝑉 |) of the
voltage at terminal 2. Finally, by analogy to other Xyce devices, node 2 can
be considered as the negative terminal for this device, while node 1 is the
positive terminal.

AT This device supports all three analysis types (AT), namely IV, PQR and PQP.
The equations for these analysis types are described below. All power grid
devices, of all types, in a Xyce netlist must use the same analysis type. This
constraint is not checked during netlist parsing. Violation of this constraint
may cause unpredictable results.

B Branch susceptance, given in per unit. As discussed in the Equation section

below, the susceptance value given on the branch description lines in IEEE
Common Data Format (CDF) files is split equally between terminals 1 and 2.

R Branch resistance, given in per unit.

X Branch reactance, given in per unit.

PowerGridBranch Device Parameters

Table 2-160. PowerGridBranch Device Instance Parameters

Parameter Description Units Default
AT Analysis Type – ’PQP’
B Branch Shunt Susceptance per unit 0
R Branch Resistance per unit 0
X Branch Reactance per unit 0

PowerGridBusShunt

Instance Form Y<type> <name> <input node1> <output node1>

+ <input node2> <output node2 [device parameters]

723
Examples YPowerGridBusShunt pg1_2 VR1 VR2 VI1 VI2 AT=IV R=0.05 B=0.1 X=0.05
YPGBS pg1_2a VR1 VR2 VI1 VI2 AT=IV R=0.05 B=0.1 X=0.05
YPowerGridBusShunt pg1_2b VR1 VR2 VI1 VI2 AT=PQR R=0.05 B=0.1 X=0.05
YPGBS pg1_2c VR1 VR2 VI1 VI2 AT=PQR R=0.05 B=0.1 X=0.05
YPowerGridBusShunt pg1_2d Th1 Th2 VM1 VM2 AT=PQP R=0.05 B=0.1 X=0.05
YPGBS pg1_2e Th1 Th2 VM1 VM2 AT=PQP R=0.05 B=0.1 X=0.0

Parameters and
Options type
The device type has a verbose (PowerGridBusShunt) and a shortened
(PGBS) form. Their usage may be mixed within a netlist.
name
Name of the device instance. This must be present, and unique amongst the
PowerGridBusShunt devices in the netlist.
input node
There are two input nodes, <input node1> and <input node2>, whose
definitions depend on the AnalysisType (AT) specified. Both nodes must be
specified. This device can be viewed as a generalized 4-port resistor, using
the Equivalent Real Form (ERF) described below in the equation
subsections. For IV and PQR formats, <input node1> is the real part (VR)
of the voltage at terminal 1 while <input node2> is the imaginary part (VI)
of the voltage at terminal 1. For PQP format, <input node1> is the angle (Θ
or Th) of the voltage at terminal 1 while <input node2> is the magnitude
(VM or |𝑉 |) of the voltage at terminal 1. Finally, by analogy to other Xyce
devices, node 1 can be considered as the positive terminal for this device,
while node 2 is the negative terminal.
output node
There are two output nodes, <output node1> and <output node2>,
whose definitions depend on the AnalysisType (AT) specified. Both nodes
must be specified. This device can be viewed as a generalized 4-port resistor,
using the ERF described below in the equation subsections. For IV and PQR
formats, <output node1> is the real part (VR) of the voltage at terminal 2
while <output node2> is the imaginary part (VI) of the voltage at terminal
2. For PQP format, <output node1> is the angle (Θ or Th) of the voltage at
terminal 2 while <output node2> is the magnitude (VM or |𝑉 |) of the
voltage at terminal 2. Finally, by analogy to other Xyce devices, node 2 can
be considered as the negative terminal for this device, while node 1 is the
positive terminal.
AT This device supports all three analysis types (AT), namely IV, PQR and PQP.
The equations for these analysis types are described below. All power grid
devices, of all types, in a Xyce netlist must use the same analysis type. This
constraint is not checked during netlist parsing. Violation of this constraint
may cause unpredictable results.
B Shunt susceptance, given in per unit.

724
 G Shunt conductance, given in per unit.

Bus Shunt Device Parameters

Table 2-161. PowerGridBusShunt Device Instance Parameters

Parameter Description Units Default
AT Analysis Type – ’PQP’
B Shunt Susceptance per unit 0
G Shunt Conductance per unit 0

PowerGridTransformer

Instance Form Y<type> <name> <input node1> <output node1>

+ <input node2> <output node2> [control node] [device parameters]

Examples YPowerGridTransformer pg1_2 VR1 VR2 VI1 VI2 AT=IV R=0.05 X=0.05
+ TR=0.9 PS=0.1
YPGTR pg1_2a VR1 VR2 VI1 VI2 AT=IV R=0.05 X=0.05 TR=0.9 PS={18*PI/180}
YPowerGridTransformer pg1_2b VR1 VR2 VI1 VI2 AT=PQR R=0.05 X=0.05
+ TR=0.9 PS=0.1
YPGTR pg1_2c VR1 VR2 VI1 VI2 AT=PQR R=0.05 B=0.1 X=0.05 TR=0.9 PS=0.1
YPowerGridTransformer pg1_2d Th1 Th2 VM1 VM2 AT=PQP
+ R=0.05 X=0.05 PS={18*PI/180}
YPGTR pg1_2e Th1 Th2 VM1 VM2 AT=PQP R=0.05 X=0.0 TR=0.9 PS=0.1
YPGTR pg1_2f Th1 Th2 VM1 VM2 N AT=PQP R=0.05 X=0.0 TT=VT PS=0.1
YPGTR pg1_2g Th1 Th2 VM1 VM2 Phi AT=PQP R=0.05 X=0.0 TT=PS TR=0.9

Parameters and
Options type
The device type has a verbose (PowerGridTransformer) and a shortened
(PGTR) form. Their usage may be mixed within a netlist.
name
Name of the device instance. This must be present, and unique amongst the
PowerGridTransformer devices in the netlist.
input node
There are two input nodes, <input node1> and <input node2>, whose
definitions depend on the AnalysisType (AT) specified. Both nodes must be
specified. This device can be viewed as a generalized 4-port resistor, using
the Equivalent Real Form (ERF) described below in the equation
subsections. For IV and PQR formats, <input node1> is the real part (VR)
of the voltage at terminal 1 while <input node2> is the imaginary part (VI)
of the voltage at terminal 1. For PQP format, <input node1> is the angle (Θ
or Th) of the voltage at terminal 1 while <input node2> is the magnitude

725
 (VM or |𝑉 |) of the voltage at terminal 1. Finally, by analogy to other Xyce
devices, node 1 can be considered as the positive terminal for this device,
while node 2 is the negative terminal.
output node
There are two output nodes, <output node1> and <output node2>,
whose definitions depend on the AnalysisType (AT) specified. Both nodes
must be specified. This device can be viewed as a generalized 4-port resistor,
using the ERF described below in the equation subsections. For IV and PQR
formats, <output node1> is the real part (VR) of the voltage at terminal 2
while <output node2> is the imaginary part (VI) of the voltage at terminal
2. For PQP format, <output node1> is the angle (Θ or Th) of the voltage at
terminal 2 while <output node2> is the magnitude (VM or |𝑉 |) of the
voltage at terminal 2. Finally, by analogy to other Xyce devices, node 2 can
be considered as the negative terminal for this device, while node 1 is the
positive terminal.
control input
This is an optional node. However, it must be specified on the instance line if
the transformer type (TT) is set to either 2 or 3. It does not exist, and must not
be specified on the instance line, for the default of TT=1. The use of the
control input node is covered under the definition of the TT instance
parameter.
AT This device supports all three analysis types (AT), namely IV, PQR and PQP.
The equations for these analysis types are described below. All power grid
devices, of all types, in a Xyce netlist must use the same analysis type. This
constraint is not checked during netlist parsing. Violation of this constraint
may cause unpredictable results.
PS Phase shift given in radians. As illustrated above, PS={18*PI/180} is a
convenient syntax for converting between decimal degrees and radians on a
Xyce instance line. This instance parameter is ignored if TT=3, since the
phase shift is set by the optional control node in that case.
R Resistance, given in per unit.
TR Turns ratio, given in per unit. This instance parameter is ignored if TT=2,
since this value is set by the optional control node in that case..
X Reactance, given in per unit.
TT This is the “Transformer Type”. It allows the user to implement tap-changing
or phase-shifting transformers, by attaching an appropriate control-circuit to
the control input node. The allowed values for TT are FT, VT or PS, with
default value of FT. Any other values will cause a netlist parsing error. A
transformer type of FT has a fixed turns-ratio, and is a four-terminal device
with two input nodes (<input node1> and <input node2>) and two
output nodes (<output node1> and <output node2>). Let the effective
complex turns ratio be 𝑟 = 𝑚 + 𝑗 𝑝 = 𝑛 ∗ (𝑐𝑜𝑠(𝜙) + 𝑗 ∗ 𝑠𝑖𝑛(𝜙)). The
transformer type of VT exposes the 𝑛 variable as the control input node,
and hence can operate with a variable turns-ratio. The transformer type of PS

726
 exposes the 𝜙 variable as the control input node, and hence can act as a
phase shifter. The instantaneous value of 𝑛 (or 𝜙) can be set to the voltage
applied to the control input node. There will be no current draw into (or
out of) the control input node. This device model does not yet support
simultaneously varying both 𝑛 and 𝜙.

Transformer Device Parameters

Table 2-162. PowerGridTransformer Device Instance Parameters

Parameter Description Units Default
AT Analysis Type – ’PQP’
PS Phase Shift rad 0
R Resistance per unit 0
TR Transformer Turns Ratio per unit 1
TT Transformer Type – ’FT’
X Reactance per unit 0

PowerGridGenBus

Instance Form Y<type> <name> <input node1> <output node1>

+ <input node2> <output node2> [device parameters]

Examples YPowerGridGenBus GenBus1 Th1 0 VM1 0 AT=PQP VM=1.045 P=0.4

YPGGB GenBus2 Th2 GND VM2 GND AT=PQP VM=1.045 P=0.4

Parameters and
Options type
The device type has a verbose (PowerGridGenBus) and a shortened (PGGB)
form. Their usage may be mixed within a netlist.
name
Name of the device instance. This must be present, and unique amongst the
PowerGridGenBus devices in the netlist.
input node
There are two input nodes, <input node1> and <input node2>, whose
definitions depend on the AnalysisType (AT) specified. Both nodes must be
specified. This device can be viewed as a generalized 4-port resistor, using
the Equivalent Real Form (ERF) described below in the equation
subsections. For IV and PQR formats, <input node1> is the real part (VR)
of the voltage at terminal 1 while <input node2> is the imaginary part (VI)
of the voltage at terminal 1. For PQP format, <input node1> is the angle (Θ
or Th) of the voltage at terminal 1 while <input node2> is the magnitude
(VM or |𝑉 |) of the voltage at terminal 1. Finally, by analogy to other Xyce

727
 devices, node 1 can be considered as the positive terminal for this device,
while node 2 is the negative terminal.
output node
There are two output nodes, <output node1> and <output node2>,
whose definitions depend on the AnalysisType (AT) specified. Both nodes
must be specified. This device can be viewed as a generalized 4-port resistor,
using the ERF described below in the equation subsections. For IV and PQR
formats, <output node1> is the real part (VR) of the voltage at terminal 2
while <output node2> is the imaginary part (VI) of the voltage at terminal
2. For PQP format, <output node1> is the angle (Θ or Th) of the voltage at
terminal 2 while <output node2> is the magnitude (VM or |𝑉 |) of the
voltage at terminal 2. Finally, by analogy to other Xyce devices, node 2 can
be considered as the negative terminal for this device, while node 1 is the
positive terminal.

AT This device currently only supports the PQP analysis type (AT). The
equations for the PQP analysis type are described below. All power grid
devices, of all types, in a Xyce netlist must use the same analysis type. This
constraint is not checked during netlist parsing. Violation of this constraint
may cause unpredictable results.

P Generator Output Power, given in per unit. As noted below, positive real
power (P) and positive reactive power (Q) flow out of the positive (<input
node1> and <input node2>) terminals into the power grid. This is
opposite from the normal convention for voltage and current sources in Xyce
and SPICE.
QLED
This is the Q-Limit Enforcement Delay. It is only used if either QMAX or QMIN
is specified. The Q-Limits are not enforced for the first QLED Newton
iterations of the DC Operating Point (DCOP) calculation. This may be useful
if a given generator bus has, for example, a very small value of QMIN [36]. If
QMAX or QMIN is specified and QLED is omitted then the default QLED value of
0 is used.
QMAX
The upper limit on the reactive power (Q) flow into the power grid, given in
per unit. If this parameter is omitted on the instance line then no upper limit
on the reactive power flow is enforced. It is recommended that either both
QMAX and QMIN be specified or that both be omitted.
QMIN
The lower limit on the reactive power (Q) flow into the power grid, given in
per unit. If this parameter is omitted on the instance line then no lower limit
on the reactive power flow is enforced. It is recommended that either both
QMAX and QMIN be specified or that both be omitted.

VM Fixed voltage magnitude, given in per unit.

Generator Bus Device Parameters

728
 Table 2-163. PowerGridGenBus Device Instance Parameters
Parameter Description Units Default
AT Analysis Type – ’PQP’
P Generator Output Power per unit 1
QLED Q-Limit Enforcement Delay – 0
QMAX Reactive Power Max Limit per unit 1
QMIN Reactive Power Min Limit per unit 0
VM Voltage Magnitude per unit 1

Branch Current and Power Accessors This version of the Power Grid devices does not support the
branch current accessor, I(), or the power accessors, P() or W().

Compatibility with .STEP .STEP should work with all of the instance parameters for the power grid
devices. The two exceptions are the Analysis Type (AT) for all of the power grid devices and the
Transformer Type (TT) for the Transformer device. Those two parameters must be constant for all steps.

Model Limitations and Caveats The following features are not supported by this release of the Power
Grid device models.

• The Generator Bus device model only supports the PQ Polar format. So, reactive power (QMAX and
QMIN) limits in the Generator Bus device model are also only supported for that format.

• Magnetizing susceptance for transformers.

• Certain instance parameters, or combinations of instance parameters, will cause errors during netlist
parsing. In particular, either B, R or X must be non-zero for the Branch device. Either B or G must be
non-zero for the Bus Shunt device. Either R or X must be non-zero for the Transformer device. TR
must not be zero for the Transformer device. VM must be positive for the Generator Bus device.

Equivalent Real Form An Equivalent Real Form (ERF) must be used to make the complex-valued
voltage-current and power-flow equations compatible with the real-valued solvers used by Xyce. The
equations given below use a K1 ERF [35], which solves the complex-valued system of equations 𝐼 = 𝑌𝑉 as
follows. Let 𝑌 = (𝑔 + 𝑗 𝑏), 𝑉 = (𝑉𝑅 + 𝑗𝑉𝐼 ) and 𝐼 = (𝐼 𝑅 + 𝑗 𝐼 𝐼 ). Then the equivalent set of real-valued
equations is:

    
𝐼𝑅 𝑔 −𝑏 𝑉𝑅
= (2.25)
𝐼𝐼 𝑔 𝑏 𝑉𝐼

729
Y Matrices for Power Grid Branch and Bus Shunt The Y-Matrix for the PowerGridBranch device
can be expressed as follows where 𝐴 = (𝑅 + 𝑗 𝑋) −1 ,
𝑅 is the branch resistance, 𝑋 is the branch reactance
and 𝐵 is the branch shunt susceptance given on the device’s instance line:

     
𝑌11 𝑌12 𝑔11 + 𝑗 𝑏 11 𝑔12 + 𝑗 𝑏 12 𝐴 −𝐴 + 0.5 𝑗 ∗ 𝐵
= = (2.26)
𝑌21 𝑌22 𝑔21 + 𝑗 𝑏 21 𝑔22 + 𝑗 𝑏 22 −𝐴 + 0.5 𝑗 ∗ 𝐵 𝐴
The Y-Matrix for the PowerGridBusShunt device can be expressed as follows where 𝐺 is the bus shunt

Figure 2-5. Lumped Π Model for PowerGridBranch.

conductance and 𝐵 is the bus shunt susceptance given on the device’s instance line:

     
𝑌11 𝑌12 𝑔11 + 𝑗 𝑏 11 𝑔12 + 𝑗 𝑏 12 𝐺 + 𝑗 𝐵 −𝐺 − 𝑗 𝐵
= = (2.27)
𝑌21 𝑌22 𝑔21 + 𝑗 𝑏 21 𝑔22 + 𝑗 𝑏 22 −𝐺 − 𝑗 𝐵 𝐺 + 𝑗 𝐵

Figure 2-6. Equivalent Circuit for PowerGridBusShunt.

Equations Common to Power Grid Branch and Bus Shunt The PowerGridBranch and
PowerGridBusShunt devices use the same basic equations to model voltage and current flow or voltage
and power flow. The differences are in the Y-Matrices described above. There are three options for the
equations used, namely I=YV, PQ Polar and PQ Rectangular.
For the I=YV format, the device equations for the PowerGridBranch and PowerGridBusShunt devices
are as follows, where the 𝑔𝑖 𝑗 and 𝑏 𝑖 𝑗 terms are given above. Also, 𝑉𝑅1 and 𝑉𝐼1 are the real and imaginary
parts of the voltage at terminal 1. 𝐼 𝑅1 and 𝐼 𝐼1 are the real and imaginary parts of the current at terminal 1.


 𝐼 𝑅1   𝑔11
  𝑔12 −𝑏 11 −𝑏 12 


 𝑉𝑅1 

 𝐼 𝑅2   𝑔21
= 𝑔22 −𝑏 21 −𝑏 22   𝑉𝑅2 
    (2.28)

 𝐼 𝐼1   𝑏 11
  𝑏 12 𝑔11 𝑔12 


 𝑉𝐼1 


 𝐼 𝐼2   𝑏 21
  𝑏 22 𝑔21 𝑔22 


 𝑉𝐼2 


730
For the PQ Rectangular format, the device equations are nonlinear [36].
2 2
𝑃1 = 𝑔11 (𝑉𝑅1 + 𝑉𝐼1 ) + 𝑉𝑅1 (𝑔12 ∗ 𝑉𝑅2 − 𝑏 12 ∗ 𝑉𝐼2 ) + 𝑉𝐼1 (𝑏 12 ∗ 𝑉𝑅2 + 𝑔12 ∗ 𝑉𝐼2 ) (2.29)
2 2
𝑃2 = 𝑔22 (𝑉𝑅2 + 𝑉𝐼2 ) + 𝑉𝑅2 (𝑔21 ∗ 𝑉𝑅1 − 𝑏 21 ∗ 𝑉𝐼1 ) + 𝑉𝐼2 (𝑏 21 ∗ 𝑉𝑅1 + 𝑔21 ∗ 𝑉𝐼1 ) (2.30)
2 2
𝑄 1 = −𝑏 11 (𝑉𝑅1 + 𝑉𝐼1 ) + 𝑉𝐼1 (𝑔12 ∗ 𝑉𝑅2 − 𝑏 12 ∗ 𝑉𝐼2 ) + 𝑉𝑅1 (𝑏 12 ∗ 𝑉𝑅2 + 𝑔12 ∗ 𝑉𝐼2 ) (2.31)
2 2
𝑄 2 = −𝑏 22 (𝑉𝑅2 + 𝑉𝐼2 ) + 𝑉𝐼2 (𝑔21 ∗ 𝑉𝑅1 − 𝑏 21 ∗ 𝑉𝐼1 ) + 𝑉𝑅2 (𝑏 21 ∗ 𝑉𝑅1 + 𝑔21 ∗ 𝑉𝐼1 ) (2.32)

For the PQ Polar format, the device equations are also nonlinear [36]. Define |𝑉1 | as the voltage magnitude
at terminal 1 and Θ1 as the voltage angle at terminal 1.

𝑃1 = 𝑔11 ∗ |𝑉1 | 2 + |𝑉1 | ∗ |𝑉2 | ∗ (𝑔12 ∗ 𝑐𝑜𝑠(Θ1 − Θ2 ) + 𝑏 12 ∗ 𝑠𝑖𝑛(Θ1 − Θ2 )) (2.33)

2
𝑃2 = 𝑔22 ∗ |𝑉2 | + |𝑉2 | ∗ |𝑉1 | ∗ (𝑔21 ∗ 𝑐𝑜𝑠(Θ2 − Θ1 ) + 𝑏 21 ∗ 𝑠𝑖𝑛(Θ2 − Θ1 )) (2.34)
2
𝑄 1 = −𝑏 11 ∗ |𝑉1 | + |𝑉1 | ∗ |𝑉2 | ∗ (𝑔12 ∗ 𝑠𝑖𝑛(Θ1 − Θ2 ) − 𝑏 12 ∗ 𝑐𝑜𝑠(Θ1 − Θ2 )) (2.35)
2
𝑄 2 = −𝑏 22 ∗ |𝑉2 | + |𝑉2 | ∗ |𝑉1 | ∗ (𝑔21 ∗ 𝑠𝑖𝑛(Θ2 − Θ1 ) − 𝑏 21 ∗ 𝑐𝑜𝑠(Θ2 − Θ1 )) (2.36)

Equations for Power Grid Transformer The equations for the PowerGridTransformer device are
similar to those used by the PowerGridBranch and PowerGridBusShunt devices. The circuit diagram for the
PowerGridTransformer is shown below.

For I=YV and PQ Rectangular formats, the equations are the same as for the PowerGridBranch and
PowerBusBusShunt devices. However, the following Y-Matrix is used where where 𝐴 = (𝑅 + 𝑗 𝑋) −1 , 𝑅 is
the resistance, 𝑋 is the reactance, 𝑛 is the turns ratio (which is the TR instance parameter) and 𝜙 is the phase
shift in radians (which is the PS instance parameter).

For the I=YV and PQ Rectangular formats, the Y matrix is not symmetric and is given by the following
[37]. Let the effective complex turns ratio be 𝑟 = 𝑚 + 𝑗 𝑝 = 𝑛 ∗ (𝑐𝑜𝑠(𝜙) + 𝑗 ∗ 𝑠𝑖𝑛(𝜙)):

𝐴 ∗ (𝑚 2 + 𝑝 2 ) −1 −𝐴 ∗ (𝑚 − 𝑗 𝑝) −1
     
𝑌11 𝑌12 𝑔11 + 𝑗 𝑏 11 𝑔12 + 𝑗 𝑏 12
= = (2.37)
𝑌21 𝑌22 𝑔21 + 𝑗 𝑏 21 𝑔22 + 𝑗 𝑏 22 −𝐴 ∗ (𝑚 + 𝑗 𝑝) −1 𝐴

The voltage-current and power flow equations for the I=YV and PQ Rectangular formats are then the same
as for the PowerGridBranch and PowerGridBusShunt devices, with the modified Y-matrix parameters
given above.

For the PQ Polar format, the Y matrix is not symmetric and is given by [36]:

𝐴 ∗ 𝑛 −2 −𝐴 ∗ 𝑛 −1
     
𝑌11 𝑌12 𝑔11 + 𝑗 𝑏 11 𝑔12 + 𝑗 𝑏 12
= = (2.38)
𝑌21 𝑌22 𝑔21 + 𝑗 𝑏 21 𝑔22 + 𝑗 𝑏 22 −𝐴 ∗ 𝑛 −1 𝐴
The power flow equation for PQ Polar format are then:

𝑃1 = 𝑔11 ∗ |𝑉1 | 2 + |𝑉1 | ∗ |𝑉2 | ∗ (𝑔12 ∗ 𝑐𝑜𝑠(Θ1 − Θ2 − 𝜙) + 𝑏 12 ∗ 𝑠𝑖𝑛(Θ1 − Θ2 − 𝜙)) (2.39)

2
𝑃2 = 𝑔22 ∗ |𝑉2 | + |𝑉2 | ∗ |𝑉1 | ∗ (𝑔21 ∗ 𝑐𝑜𝑠(Θ2 − Θ1 + 𝜙) + 𝑏 21 ∗ 𝑠𝑖𝑛(Θ2 − Θ1 + 𝜙)) (2.40)
2
𝑄 1 = −𝑏 11 ∗ |𝑉1 | + |𝑉1 | ∗ |𝑉2 | ∗ (𝑔12 ∗ 𝑠𝑖𝑛(Θ1 − Θ2 − 𝜙) − 𝑏 12 ∗ 𝑐𝑜𝑠(Θ1 − Θ2 − 𝜙)) (2.41)
2
𝑄 2 = −𝑏 22 ∗ |𝑉2 | + |𝑉2 | ∗ |𝑉1 | ∗ (𝑔21 ∗ 𝑠𝑖𝑛(Θ2 − Θ1 + 𝜙) − 𝑏 21 ∗ 𝑐𝑜𝑠(Θ2 − Θ1 + 𝜙)) (2.42)

731
 Figure 2-7. Equivalent Circuit for PowerGridTransformer.

Equations for Power Grid Gen Bus The PowerGridGenBus is an active device that functions as an
ideal generator with a fixed power output (𝑃) and voltage magnitude (𝑉 𝑀). Reactive power (QMAX and
QMIN) limits are also supported. The device equations for the PQ Polar format are as follows [36]. The
other solution formulations are not supported in this release. If reactive power limits are not being enforced
then:

𝑃1 = 𝑃 (2.43)
|𝑉1 | = 𝑉 𝑀 (2.44)

If reactive power limits are being enforced then 𝑃1 is still held constant but the behavior of the 𝑉1 terminal
changes between a constant-voltage and a constant-current source. In particular, |𝑉1 | = 𝑉 𝑀 only if QMIN
< 𝑄 1 < QMAX. Otherwise, |𝑉1 | is unconstrained and the appropriate QMIN or QMAX value is enforced at the
𝑉1 terminal instead.
The convention for Power Grids is that positive power is injected into the grid. So, positive real (P) and
reactive power (Q) flow out of the positive terminals (inputNode 1 and inputNode 2). This is reversed
from the normal convention for current direction for voltage and current sources in either Xyce or SPICE.

732
2.3.32. Memristor Device

Symbol

Instance Form ymemristor <name> <(+) node> <(-) node> <model>

Model Form .MODEL <model name> MEMRISTOR level=2 [model parameters]

Examples
ymemristor mr1 n1 n2 mrm2

.model mrm2 memristor level=2 ron=50 roff=1000

+ koff=1.46e-18 kon=-4.68e-22
+ alphaoff=10 alphaon=10 wc=1.0e-12
+ ioff=115e-6 ion=-8.9e-6 xscaling=1.0e9 wt=4

ymemristor mr2 n1 n2 mrm3 xo=0.11

.MODEL mrm3 memristor level=3 a1=0.17 a2=0.17 b=0.05 vp=0.16 vn=0.15

+ ap=4000 an=4000 xp=0.3 xn=0.5 alphap=1 alphan=5 eta=1

ymemristor mr3 n1 n2 mrm4

.model mrm4 memristor level=4

+ fxpdata=fxp_table.csv
+ fxmdata=fxm_table.csv
+ I1=85.37e-6 I2=90.16e-6 V1=0.265 V2=0.265 G0=130.72e-6
+ VP=0.7 VN=1.0 d1=9.87 d2=-4.82
+ C1=1000 C2=1000

Parameters and
Options (+) node
(-) node
Polarity definition for a positive voltage across the memristor. The first node
is defined as positive. Therefore, the voltage across the component is the first
node voltage minus the second node voltage.

Comments The level=2 memristor device is an implementation of the TEAM formulation

described in [38] and [39]. The level=3 memristor device is an implementation of
the Yakopcic formulation described in [40]. The level=4 memristor device is an
implementation of the Piecewise Empirical Model described in [41].

733
 Positive current flows from the (+) node through the device to the (-) node. The
power through the device is calculated with 𝐼 · Δ𝑉 where the voltage drop is calculated
as (𝑉+ − 𝑉− ) and positive current flows from 𝑉+ to 𝑉− .

Device Equations for TEAM Formulation The current voltage relationship for the TEAM formulation
can be linear or nonlinear and this is selectable with the instance parameter IVRELATION. The default is the
linear relationship which is:
 
𝑅𝑂𝐹𝐹 − 𝑅𝑂 𝑁
𝑣(𝑡) = 𝑅𝑂 𝑁 + (𝑥 − 𝑥𝑂 𝑁 ) 𝑖(𝑡) (2.45)
𝑥𝑂𝐹𝐹 − 𝑥𝑂 𝑁

The non-linear relationship is:

𝑣(𝑡) = 𝑅𝑂 𝑁 𝑒 𝜆( 𝑥− 𝑥𝑂𝑁 )/( 𝑥𝑂𝐹𝐹 − 𝑥𝑂 𝑁 ) 𝑖(𝑡) (2.46)

where 𝜆 is defined as:

𝑅𝑂𝐹𝐹
= 𝑒𝜆 (2.47)
𝑅𝑂 𝑁
In the above equations 𝑥 represents a doped layer whose growth determines the overall resistance of the
device. The equation governing the value of 𝑥 is:
  𝛼𝑂𝐹𝐹
𝑖
− 1 𝑓𝑂𝐹𝐹 (𝑥) 0 < 𝑖𝑂𝐹𝐹 < 𝑖

 𝑘
 𝑂𝐹𝐹 𝑖𝑂𝐹𝐹


𝑑𝑥 
= 0 𝑖𝑂 𝑁 < 𝑖 < 𝑖𝑂𝐹𝐹 (2.48)
𝑑𝑡  
𝑖
 𝛼𝑜𝑛
−1 𝑓𝑂 𝑁 (𝑥) 𝑖 < 𝑖𝑂 𝑁 < 0

 𝑘 𝑂 𝑁 𝑖𝑜𝑛



The functions 𝑓𝑂 𝑁 (𝑥) and 𝑓𝑂𝐹𝐹 (𝑥) are window functions designed to keep 𝑥 within the defined limits of
𝑥𝑂 𝑁 and 𝑥𝑂𝐹𝐹 . Four different types of window functions are available and this is selectable with the model
parameter WT. Note that the TEAM memristor device is formulated to work best with the TEAM,
Kvatinsky, window function WT=4 . Other window functions should be used with caution.

Device Equations for Yakopcic Formulation The current voltage relationship for the Yakopcic
memristor device is: [40] 
𝑎 1 𝑥(𝑡) sinh(𝑏𝑉 (𝑡)) 𝑉 (𝑡) ≥ 0
𝐼 (𝑡) = (2.49)
𝑎 2 𝑥(𝑡) sinh(𝑏𝑉 (𝑡)) 𝑉 (𝑡) < 0

𝑉 (𝑡 ) − exp𝑉𝑝


 𝐴 𝑝 exp
𝑉 (𝑡) > 𝑉 𝑝



𝑔(𝑉 (𝑡)) = −𝐴𝑛 exp−𝑉 (𝑡 ) − exp𝑉𝑛 𝑉 (𝑡) < −𝑉𝑛 (2.50)
 0 −𝑉𝑛 ≤ 𝑉 (𝑡) ≤ 𝑉 𝑝



The internal state variable, 𝑥, is governed by the equation:

𝑑𝑥
= 𝑛𝑔(𝑉 (𝑡)) 𝑓 (𝑥(𝑡)) (2.51)
𝑑𝑡

where 𝑓 (𝑥) is defined by:

734
 exp− 𝛼𝑝 ( 𝑥− 𝑥 𝑝 ) 𝑤 𝑝 (𝑥, 𝑥 𝑝 ) 𝑥 ≥ 𝑥 𝑝

𝑓 (𝑥) = (2.52)
1 𝑥 ≤ 𝑥𝑝

exp 𝛼𝑛 ( 𝑥+𝑥𝑛 −1 𝑤 𝑛 (𝑥, 𝑥 𝑛 ) 𝑥 ≤ 1 − 𝑥 𝑛


𝑓 (𝑥) = (2.53)
1 𝑥 > 1 − 𝑥𝑛
𝑥𝑝 − 𝑥
𝑤 𝑝 (𝑥, 𝑥 𝑝 ) = +1 (2.54)
1 − 𝑥𝑝
𝑥
𝑤 𝑛 (𝑥, 𝑥 𝑛 ) = (2.55)
1 − 𝑥𝑛

Note, the quantities, 𝑥 𝑝 , 𝑥 𝑛 , 𝛼 𝑝 , 𝛼𝑛 , 𝐴 𝑝 , 𝐴𝑛 , 𝑎 1 , 𝑎 2 and𝑏 are model parameters that can be specified in the
device’s model block.

Device Equations for the PEM Formulation The PEM memristor device is similar to the TEAM and
Yakopcic formulations in that an internal state variable, 𝑥, is used to capture the device’s response to its
history.
The I-V relationship is
𝐼 = 𝑥 ℎ(𝑉) (2.56)

and ℎ(𝑉) is defined by:

ℎ(𝑉) = 𝐼1 ∗ 𝑒𝑥 𝑝(𝑉/𝑉1 ) − 𝐼2 ∗ 𝑒𝑥 𝑝(−𝑉/𝑉2 ) + 𝐺 0𝑉 − (𝐼1 − 𝐼2 ) (2.57)

where 𝐼1 , 𝐼2 , 𝑉1 , 𝑉2 and 𝐺 0 are model parameters.

The internal variable, 𝑥, is defined by:
𝑑𝑥
= 𝐺 (𝑉) 𝑓 (𝑥) (2.58)
𝑑𝑡
with  
𝑑1 [ 𝑉 (𝑡 ) −𝑉𝑝 ]
exp −1



 𝐶 1 𝑉 > 𝑉𝑝

  
𝐺 (𝑉) = 𝐶2 exp𝑑2 [𝑉 (𝑡 ) −𝑉𝑛 ] −1 𝑉 < −𝑉𝑛 (2.59)


 0 −𝑉𝑛 ≤ 𝑉 (𝑡) ≤ 𝑉 𝑝



Finally, the function 𝑓 (𝑥) is defined by a user supplied set set data which is used with linear interpolation to
find the current value of 𝑓 (𝑥). Separate data sets are used for forward bias and reverse bias.
 +
𝐹 dataset 𝑉 > 0
𝑓 (𝑥) = (2.60)
𝐹 − dataset 𝑉 < 0

Device Parameters for TEAM Formulation

Table 2-164. MemristorTEAM Device Instance Parameters

Parameter Description Units Default
IVRELATION IV relationship to use, 0 is linear, 1 is nonlinear – 0

735
Model Parameters for TEAM Formulation

Table 2-165. MemristorTEAM Device Model Parameters

Parameter Description Units Default
ALPHAOFF Modeling Coefficient – 3
ALPHAON Modeling Coefficient – 3
AOFF Window Function Parameter (window 4) m 3e-09
AON Window Function Parameter (window 4) m 0
D Window Function Parameter (windows 1, 2 and 3) – 0.000115
IOFF Current scale in off state ˙ 0.000115
ION Current scale in On state A 8.9e-06
J Window Function Parameter (window 3) – 0.000115
KOFF Modeling Coefficient m/s 8e-13
KON Modeling Coefficient m/s -8e-13
P Window Function Parameter (windows 1, 2 and 3) – 0.000115
ROFF Resistence in off state ˙ 1000
RON Resistence in on state ˙ 50
WC Window Function Parameter (window 4) m 1.07e-12
Type of windowing function: 0-None, 1-Jogelkar,
WT – 0
2-Biolek, 3-Prodromakis, 4-Kvatinsky
XOFF Modeling Coefficient m 3e-09
XON Modeling Coefficient m 0
Scaling for x variable. For example 1e9 if x will be in
XSCALING – 1
units of nanometers.

Device Parameters for Yakopcic Formulation

Table 2-166. MemristorYakopcic Device Instance Parameters

Parameter Description Units Default
XO Initial value for internal variable x – 0

Model Parameters for Yakopcic Formulation

Table 2-167. MemristorYakopcic Device Model Parameters

Parameter Description Units Default
A1 Dielectric layer thickness parameter [dimensionless] – 1
A2 Dielectric layer thickness parameter [dimensionless] – 1
ALPHAN State variable motion. – 1

736
 Table 2-167. MemristorYakopcic Device Model Parameters
Parameter Description Units Default
ALPHAP State variable motion. – 1
AN Negative Voltage Threshold Magnitude Parameter – 1
AP Positive Voltage Threshold Magnitude Parameter – 1
Curvature in I-V relation. Relates to how much
B conduction in the device is Ohmic and versus tunnel – 1
barrier.
ETA State variable motion relative to voltage. – 1
RTN model in resistance: Base change in resistance
RESDELTA ˙ 0
for RTN
RTN model in resistance: Base change in resistance
RESDELTAGRAD – 0
for RTN scaled by R
RTN model in resistance: Minimum allowed update
RESEPTD s 1e-10
time
RESLAMBDA RTN model: lambda – 0
RESNOISE RTN model in resistance (on/off) – false
RESSEED RTN model in resistance: seed – 0
RESTD RTN model in resistance: Update time s 0
VN Negative Voltage Threshold V -0.01
VP Positive Voltage Threshold V 0.01
RTN model in growth: Base change in growth rate for
XDELTA ˙ 0
RTN
RTN model in growth: Base change in growth for
XDELTAGRAD – 0
RTN scaled by X
XEPTD RTN model in growth: Minimum allowed update time s 1e-10
XLAMBDA RTN growth model: lambda – 0
XN State variable motion. – 1
XNOISE RTN model in growth (on/off) – false
XP State variable motion. – 1
Scaling for x variable. For example 1e9 if x will be in
XSCALING – 1
units of nanometers.
XSEED RTN model in growth: seed – 0
XTD RTN model in growth: Update time s 0

Device Parameters for PEM Formulation

Table 2-168. MemristorPEM Device Instance Parameters

Parameter Description Units Default
XO Initial value for internal variable x – 0

737
Model Parameters for PEM Formulation

Table 2-169. MemristorPEM Device Model Parameters

Parameter Description Units Default
State variable proportionality parameter for forward
C1 – 1
bias
State variable proportionality parameter for negative
C2 – 1
bias
D1 Positive Voltage Threshold Magnitude Parameter – 1
D2 Negative Voltage Threshold Magnitude Parameter – 1
FXMDATA File from which to read x,f-(x) data – ’filem.dat’
FXPDATA File from which to read x,f+(x) data – ’filep.dat’
G0 Conductance factor. – 1
I1 Current Scale factor. A 1
I2 Current Scale factor. A 1
V1 Voltage Scale factor. V 1
V2 Voltage Scale factor. V 1
VN Negative Voltage Threshold V -0.01
VP Positive Voltage Threshold V 0.01

738
2.3.33. Subcircuit

A subcircuit can be introduced into the circuit netlist using the specified nodes to substitute for the
argument nodes in the definition. It provides a building block of circuitry to be defined a single time and
subsequently used multiple times in the overall circuit netlists. See Section 2.1.37 for more information
about subcircuits.

Instance Form X<name> [node]* <subcircuit name> [PARAMS: [<name> = <value>]*]

Examples X12 100 101 200 201 DIFFAMP

XBUFF 13 15 UNITAMP
XFOLLOW IN OUT VCC VEE OUT OPAMP
XFELT 1 2 FILTER PARAMS: CENTER=200kHz
XNANDI 25 28 7 MYPWR MYGND PARAMS: IO_LEVEL=2

Parameters and
Options subcircuit name
The name of the subcircuit’s definition.
PARAMS:
Passed into subcircuits as arguments and into expressions inside the
subcircuit.

Comments There must be an equal number of nodes in the subcircuit call and in its definition.
Subcircuit references may be nested to any level. However, the nesting cannot be
circular. For example, if subcircuit A’s definition includes a call to subcircuit B, then
subcircuit B’s definition cannot include a call to subcircuit A.

739
This page intentionally left blank.

740
2.4. TCAD Devices

Semiconductor device simulation, which is based on a coupled set of partial differential equations (PDE’s)
is supported in Xyce. Such devices can be invoked from the circuit netlist in a manner similar to traditional
SPICE-style analog devices. One dimensional and two dimensional devices are supported, with the
dimensionality determined by the device model level.

1D Device Form YPDE <name> <node> [node] [model name]

+ [device parameters]

2D Device Form YPDE <name> <node> <node> [node][node] [model name]|

+ [device parameters]

Model Form .MODEL <model name> ZOD [model parameters]

Comments All of the PDE parameters are specified on the instance level. The model statement is
used only for specifying if the device is 1D or 2D, via the level parameter. Both the 1D
and the 2D devices can construct evenly-spaced meshes internally, or an unstructured
mesh can be read in from an external mesh file.
The eletrode, doping and material parameters are specified using a special format that
is described in the tables that are referenced in the instance parameter tables.

TCAD Device Parameters Most TCAD device parameters are specified on the instance level. The only
TCAD device model parameter is the level, which specifies whether the model is one or two dimensions.

Table 2-170. PDE Device Model Parameters

Parameter Description Units Default
LEVEL Determines if the device is 1D or 2D 1=1D, 2=2D – 1

Table 2-171. 1D PDE (level 1) Device Instance Parameters

Parameter Description Units Default
logical
AUGER Flag to turn on/off Auger recombination true
(T/F)
BULKMATERIAL Bulk semiconductor material – ’SI’
DOPINGPROFILES See Table 2-176
logical
FERMIDIRAC Use Fermi-Dirac statistics. false
(T/F)
logical
FIELDDEP If true, use field dependent mobility. false
(T/F)
LAYER See Table 2-173

741
 Table 2-171. 1D PDE (level 1) Device Instance Parameters
Parameter Description Units Default
If set to true, then some variables are excluded from logical
MASKVARSTIA false
the time integration error control calculation. (T/F)
Maximum voltage change used by two-level Newton
MAXVOLTDELTA V 0.025
algorithm.
MESHFILE – ’internal.msh’
MOBMODEL Mobility model. – ’ARORA’
NODE See Table 2-174
NX Number of mesh points – 11
REGION See Table 2-177
Flag to turn on/off Shockley-Read-Hall (SRH) logical
SRH true
recombination. (T/F)
logical
THERMIONICEMISSION false
(T/F)
TUNNELING – ’none’
Flag for using old(inaccurate) intrinsic carrier logical
USEOLDNI false
calculation. (T/F)
Flag to apply voltage limiting. This is only relevant for logical
VOLTLIM false
an experimental two-level Newton solver. (T/F)
Doping Parameters
DOPING_FILE File containing doping profile. – ’NOFILE’
Flag for graded junction vs. abrupt junction. – logical
GRADED false
(1/true=graded, 0/false=abrupt) (T/F)
NA Acceptor doping level cm−3 1e+15
ND Donor doping level cm−3 1e+15
NDOPE_FILE File containing doping profile for N-type dopants. – ’NOFILE’
PDOPE_FILE File containing doping profile for P-type dopants. – ’NOFILE’
WJ Junction width, if graded junction enabled. cm 0.0001
Geometry Parameters
ANODE.AREA Anode area (used for two-terminal devices) cm−2 0
AREA Cross sectional area of the device. cm−2 1
BASE.AREA Base area (used for three-terminal (BJT) devices) cm−2 0
Location of base contact (necessary if running with
BASE.LOC cm 0.0005
three terminals).
CATHODE.AREA Cathode area (used for two-terminal devices) cm−2 0
COLLECTOR.AREA Collector area (used for three-terminal (BJT) devices) cm−2 0
EMITTER.AREA Emitter area (used for three-terminal (BJT) devices) cm−2 0
L Device width. (Synonym with W parameter) cm 0.001
W Device width. (Synonym with L parameter) cm 0.001

742
 Table 2-171. 1D PDE (level 1) Device Instance Parameters
Parameter Description Units Default
Temperature Parameters
TEMP Device temperature ◦C 27
Model Output Parameters
This is an output parameter. It is only used if
logical
FIRSTELECTRODEOFFSET OFFSETOUTPUTVOLTAGE=true. (see description false
(T/F)
of that paramaeter
Flag for gnuplot output. 0 - no gnuplot files. 1 -
gnuplot files. gnuplot is an open source plotting
GNUPLOTLEVEL program that is usually installed on Linux systems. – 1
gnuplot files will have the *Gnu.dat suffix, and the
prefix will be thename of the device instance.
This is an output parameter that determines the “zero”
of the potential at output. If
OFFSETOUTPUTVOLTAGE=true (default) it will
adjust the voltages at output so that the minimum
voltage is zero. If true and also
logical
OFFSETOUTPUTVOLTAGE FIRSTELECTRODEOFFSET=true, then the voltage true
(T/F)
of the first electrode is the zero point. If
OFFSETOUTPUTVOLTAGE=false, the output
voltage sets the intrisic Fermi level to zero. Depending
on circumstances each of these may be more or less
convenient for plotting.
OUTPUTINTERVAL Time interval for tecplot output (if tecplot is enabled). s 0
Flag to determine if the results of the nonlinear
Poisson calculation is included in the output files. logical
OUTPUTNLPOISSON false
Normally, this calculation is used to initialize a (T/F)
drift-diffusion calculation and isn’t of interest.
Setting for Tecplot output: 0 - no Tecplot files 1 -
Tecplot files, each output in a separate file. 2 - Tecplot
TECPLOTLEVEL file, each outputappended to a single file. Tecplot files – 1
will have the .dat suffix, and the prefix will be the
name of the device instance
Scaling Parameters
Density scalar; adjust to mitigate convergence
problems.The model will do all of its scaling
C0 cm−3 1e+15
automatically, so it is generally not necessary to
specify it manually.
Fraction of the maximum doping by which density
will be scaled.The model will do all of its scaling logical
DENSITYSCALARFRACTION 0.1
automatically, so it is generally not necessary to (T/F)
specify it manually.
If set the density will be scaled by a fraction of the
maximum doping.The model will do all of its scaling logical
SCALEDENSITYTOMAXDOPING true
automatically, so it is generally not necessary to (T/F)
specify it manually.

743
 Table 2-171. 1D PDE (level 1) Device Instance Parameters
Parameter Description Units Default
Time scalar; adjust to mitigate convergence
problems.The model will do all of its scaling
t0 s 1e-06
automatically, so it is generally not necessary to
specify it manually.
Length scalar; adjust to mitigate convergence
problems.The model will do all of its scaling
X0 cm 1e-07
automatically, so it is generally not necessary to
specify it manually.
Boundary Condition Parameters
Anode voltage boundary condition. Only used if
ANODE.BC device is uncoupled from circuit, and running in diode V 0.5
mode.
Base voltage boundary condition. Only used if device
BASE.BC V 0
is uncoupled from circuit, and running in BJT mode.
Cathode voltage boundary condition. Only used if
CATHODE.BC device is uncoupled from circuit, and running in diode V 0
mode.
Collector voltage boundary condition. Only used if
COLLECTOR.BC device is uncoupled from circuit, and running in BJT V 0
mode.
Emitter voltage boundary condition. Only used if
EMITTER.BC device is uncoupled from circuit, and running in BJT V 0.5
mode.

744
 Table 2-172. 2D PDE (level 2) Device Instance Parameters
Parameter Description Units Default
logical
AUGER Flag to turn on/off Auger recombination true
(T/F)
BULKMATERIAL Material of bulk material. – ’SI’
logical
DISPLCUR If true, displacement current is computed and output false
(T/F)
DOPINGPROFILES See Table 2-176
Maximum voltage change used by two-level Newton
MAXVOLTDELTA V 0.025
algorithm.
This is a required field for a 2D simulation. If the user
specifies meshfile=internal.mesh, the model will
create a Cartesian mesh using the parameters L,W,NX
MESHFILE and NY. If the user specifies anything else (for – ’internal.msh’
example meshfile=diode.msh), the model will attempt
to read in a mesh file of that name. The format is
assumed to be that of the SG Framework.
MOBMODEL Mobility model. – ’ARORA’
NODE See Table 2-175
NX Number of mesh points, x-direction. – 11
NY Number of mesh points, y-direction. – 11
REGION See Table 2-177
Flag to turn on/off Shockley-Read-Hall (SRH) logical
SRH true
recombination. (T/F)
P-type or N-type - this is only relevant if using the
TYPE – ’PNP’
default dopings
Flag for using old (inaccurate) intrinsic carrier logical
USEOLDNI false
calculation. (T/F)
logical
VOLTLIM false
(T/F)
Doping Parameters
Flag for graded junction vs. abrupt junction. – logical
GRADED false
(1/true=graded, 0/false=abrupt) (T/F)
NA Acceptor doping level cm−3 1e+15
ND Donor doping level cm−3 1e+15
WJ Junction width, if graded junction enabled. cm 0.0001
Geometry Parameters
AREA Cross sectional area of the device. cm−2 1
logical
CYL Flag to enable cylindrical geometry false
(T/F)
L Device length cm 0.001
W Device width cm 0.001

745
 Table 2-172. 2D PDE (level 2) Device Instance Parameters
Parameter Description Units Default
Temperature Parameters
TEMP Device temperature ◦C 27
Model Output Parameters
Flag for gnuplot output. 0 - no gnuplot files. 1 -
gnuplot files. gnuplot is an open source plotting
GNUPLOTLEVEL program that is usually installed on Linux systems. – 0
gnuplot files will have the *Gnu.dat suffix, and the
prefix will be thename of the device instance.
INTERPGRIDSIZE – 20
OUTPUTINTERVAL Time interval for tecplot output (if tecplot is enabled). s 0
Flag to determine if the results of the nonlinear
Poisson calculation is included in the output files. logical
OUTPUTNLPOISSON false
Normally, this calculation is used to initialize a (T/F)
drift-diffusion calculation and isn’t of interest.
Setting for Tecplot output: 0 - no Tecplot files 1 -
Tecplot files, each output in a separate file. 2 - Tecplot
TECPLOTLEVEL file, each outputappended to a single file. Tecplot files – 1
will have the .dat suffix, and the prefix will be the
name of the device instance
Flag for volume-averaged text output. 0 - no text files.
TXTDATALEVEL 1 - text files. txtdataplot files will have the *.txt suffix, – 1
and the prefix will be the name of the device instance.
Scaling Parameters
Density scalar; adjust to mitigate convergence
problems.The model will do all of its scaling
C0 cm−3 1e+15
automatically, so it is generally not necessary to
specify it manually.
Time scalar; adjust to mitigate convergence
problems.The model will do all of its scaling
t0 s 1e-06
automatically, so it is generally not necessary to
specify it manually.
Length scalar; adjust to mitigate convergence
problems. The model will do all of its scaling
X0 cm 0.0001
automatically, so it is generally not necessary to
specify it manually.
Boundary Condition Parameters
CONSTBOUNDARY – false

746
Layer Parameters

Table 2-173. LAYER Composite Parameters

Parameter Description Units Default
CON – 1.42248
ConductionBandDOS – 2.89e+19
DIEL – 13.1
ELMOB0 – 2240
ELVSAT – 7.7e+06
EMASS – 0.067
GRADEDWIDTH – 0
HMASS – 0.5
HOMOB0 – 30
HOVSAT – 7.7e+06
MATERIAL – ’gaas’
NAME – ’EMITTER’
NARCO – 0.047
NARVA – 0.047
NDOPE – 0
NI – 1.79e+06
NX – 25
PDOPE – 5e+19
VAL – 0
ValenceBandDOS – 2.66e+19
WIDTH – 1e-06

Electrode Parameters 1D

Table 2-174. NODE Composite Parameters

Parameter Description Units Default
AREA – 0
Carrier density boundary condition type (dirichlet or
BC – ’dirichlet’
neumann)
LOCATION – 0
MATERIAL Contact material – ’neutral’
NAME Electrode name – ’anode’
OXIDEBNDRYFLAG Oxide layer boolean – false
SIDE Side specification (left or right) – ’left’

747
Electrode Parameters 2D

Table 2-175. NODE Composite Parameters

Parameter Description Units Default
Carrier density boundary condition type (dirichlet or
BC – ’dirichlet’
neumann)
END Ending location cm 0
MATERIAL Contact material – ’neutral’
NAME Electrode name – ’anode’
OXCHARGE Oxide charge – 0
OXIDEBNDRYFLAG Oxide layer boolean – false
OXTHICK Oxide thickness cm 0
SIDE Side specification (top, bottom, left or right) – ’top’
START Starting location cm 0

Doping or Region Parameters The DOPINGPROFILES and REGION parameters are synonyms, therefore
their tables of values are identical. The use of both parameters in the same device instance could lead to
unpredictable behavior.

Table 2-176. DOPINGPROFILES Composite Parameters

Parameter Description Units Default
EL2 – 0
User-defined expressions for dopant profiles as
EXPRESSION – ’none’
function of depth
FILE – ’none’
Determines the doping shape (half-gaussian or a full
FLATX – 0
gaussian)
2D ONLY: Determines the doping shape
FLATY – 0
(half-gaussian or a full gaussian)
Functional form of doping region; options are
FUNCTION – ’uniform’
uniform, gaussian, and step.
NAME – ’none’
NMAX Maximum value of impurity concentration cm−3 1e+15
NMAXCHOP cm−3 1e+20
NMIN Minimum value of impurity concentration cm−3 0
SPECIES – ’none’
TYPE ntype or ptype – ’ntype’
XLOC Peak location of the doping in the x-direction cm 0
XMAX cm 0
XMIN cm 0

748
 Table 2-176. DOPINGPROFILES Composite Parameters
Parameter Description Units Default
Distance from nmax to nmin. This is only applicable
XWIDTH cm 0.001
for the function=gaussian case.
2D ONLY: Peak location of the doping in the
YLOC cm 0
y-direction ()
YMAX 2D ONLY: cm 0
YMIN 2D ONLY: cm 0
2D ONLY: Distance from nmax to nmin. This is only
YWIDTH cm 0.001
applicable for the function=gaussian case.

Table 2-177. REGION Composite Parameters

Parameter Description Units Default
EL2 – 0
EXPRESSION – ’none’
FILE – ’none’
Determines the doping shape (half-gaussian or a full
FLATX – 0
gaussian)
2D ONLY: Determines the doping shape
FLATY – 0
(half-gaussian or a full gaussian)
Functional form of doping region; options are
FUNCTION – ’uniform’
uniform, gaussian, and step.
NAME – ’none’
NMAX Maximum value of impurity concentration cm−3 1e+15
NMAXCHOP cm−3 1e+20
NMIN Minimum value of impurity concentration cm−3 0
SPECIES – ’none’
TYPE ntype or ptype – ’ntype’
XLOC Peak location of the doping in the x-direction cm 0
XMAX cm 0
XMIN cm 0
Distance from nmax to nmin. This is only applicable
XWIDTH cm 0.001
for the function=gaussian case.
2D ONLY: Peak location of the doping in the
YLOC cm 0
y-direction ()
YMAX 2D ONLY: cm 0
YMIN 2D ONLY: cm 0
2D ONLY: Distance from nmax to nmin. This is only
YWIDTH cm 0.001
applicable for the function=gaussian case.

749
Flat Parameters

Table 2-178. Description of the flatx, flaty doping parameters

Flatx or Flaty
Description 1D Cross Section
view

0 Gaussian on both sides of the peak (xloc) location.

Gaussian if x>xloc, flat (constant at the peak value) if

+1
x<xloc.

Gaussian if x<xloc, flat (constant at the peak value) if

-1
x>xloc.

750
2.4.1. Physical Models

This section contains information about physical models used in Xyce for TCAD devices. This includes
various mobility models, expressions for calculating the effective mass for electrons and holes, an
expression for intrinsic carrier concentration as a function of temperature, expressions which describe
contacts to metal as well as contacts to metal-oxide-semiconductor devices.

2.4.1.1. Material Models and Parameters

This section describes some of the basic material properties that are available in Xyce. Described here are
the models for effective mass, intrinsic carrier concentration, and the bandgap. This information is needed
for the more complex models described in the mobility section (section 2.4.2) and the boundary condition
section (section 2.4.2.6).

2.4.1.2. Effective Mass

Xyce includes functions which return the effective mass of electrons and holes for a number of
semiconductor materials.

2.4.1.3. Electron Effective Mass

The electron effective mass is calculated as

𝑚 𝑑𝑒 = (𝑚 𝑙∗ 𝑚 ∗2
𝑡 )
1/3
(2.61)

where 𝑚 𝑙 and 𝑚 𝑡 are the effective masses along the longitudinal and transverse directions of the ellipsoidal
energy surface.

2.4.1.4. Hole Effective Mass

The hole effective mass is calculated as

∗3/2
𝑚 𝑑ℎ = (𝑚 𝑙ℎ + 𝑚 ∗3/2
ℎℎ )
2/3
(2.62)

where 𝑚 𝑙ℎ and 𝑚 ℎℎ are the "light" and "heavy" hole masses, respectively.

751
2.4.1.5. Intrinsic Carrier Concentration

The intrinsic carrier concentration in a semiconductor is obtained from the "np" product

𝑛𝑝 = 𝑛2𝑖 = 𝑁𝐶 𝑁𝑉 𝑒𝑥 𝑝(−𝐸 𝑔 /𝑘𝑇) (2.63)

or √︁
𝑛𝑖 = 𝑁𝐶 𝑁𝑉 𝑒 −𝐸𝑔 /2𝑘𝑇 (2.64)

The expression used in Xyce to calculate the intrinsic carrier concentration comes from this and is given
by

𝑚 𝑑𝑒 𝑚 𝑑ℎ 3/4 1/2 3/2 −𝐸𝑔 /2𝑘𝑇

𝑛𝑖 = 4.9 × 1015 ( ) 𝑀𝑐 𝑇 𝑒 (2.65)
𝑚 02

where 𝑀𝑐 is the number of equivalent minima in the conduction band for the semiconductor, 𝑚 𝑑𝑒 is the
density-of-state effective mass for electrons, 𝑚 𝑑ℎ is the density-of-state effective mass for holes, and 𝑚 0 is
the free-electron mass.

Table 2-179. Intrinsic Carrier Concentration Parameters

𝑛𝑖 at room
Semiconductor Symbol 𝑀𝑐1/2
temperature
√
Silicon si 6.00 1.25𝑥1010
Germanium ge 2.00 2.5𝑥1013
Galium Arsenide gaas 1.00 2.0𝑥106

2.4.1.6. Bandgap

The bandgap is a material and temperature-dependent quantity. The bandgap model for semiconductor
materials, is based on Thurmond [42]. This model is given by:

𝑇 2.0
 
𝐸 𝑔 = 𝐸 𝑔0 − 𝐴 ∗ (2.66)
𝑇 + 𝑇𝑜 𝑓 𝑓

where 𝐸 𝑔 is the bandgap (eV) and 𝑇 is the temperature (K). 𝐴, 𝐸 𝑔0 , and 𝑇𝑜 𝑓 𝑓 are all material-dependent
constants. Insulating materials, such as silicon dioxide, are assumed to have constant bandgaps, so their
bandgaps are given by:

𝐸 𝑔 = 𝐸 𝑔0 (2.67)

where 𝐸 𝑔0 is a material-dependent constant. The values for the material-dependent constants used by
equations 2.66 and 2.67 are given in Table 2-180.

752
 Table 2-180. Bandgap constants
Material Symbol 𝐸 𝑔0 (eV) 𝐴 𝑇𝑜 𝑓 𝑓 (K)
Silicon si 1.17 4.73e-4 636.0
Germanium ge 0.7437 4.774e-4 235.0
Galium Arsenide gaas 1.519 5.405e-4 204.0
Silicon Dioxide sio2 9.00 NA NA
Silicon Nitride wdi 4.7 NA NA
Sapphire cu 4.7 NA NA

753
2.4.2. Mobility Models

A number of mobility models are included in Xyce. The analytic, arora, and carrier-carrier scattering
models are considered to be low-field mobility models. The Lombardi surface mobility model is a
transverse-field dependent model which also incorporates the mobility of the bulk silicon.

2.4.2.1. Analytic Mobility

This is a concentration- and temperature-dependent empirical mobility model, based on the work of
Caughey and Thomas [43], which combines the effects of lattice scattering and ionized impurity scattering.
The equation for the mobility of electrons is:

𝑇 𝑛𝑢𝑛
𝜇 𝑛𝑚𝑎𝑥 ( 𝑇𝑟𝑒 𝑓
) − 𝜇 𝑛𝑚𝑖𝑛
𝜇0𝑛 = 𝜇 𝑛𝑚𝑖𝑛 + 𝑟𝑒 𝑓
(2.68)
𝑇 𝑥𝑖𝑛
1 + ( 𝑇𝑟𝑒 𝑓
) (𝑁𝑡𝑜𝑡 𝑎𝑙 /𝑁 𝑛 ) 𝛼𝑛

and the equation for the mobility of holes is:

𝑇 𝑛𝑢 𝑝
𝜇 𝑝𝑚𝑎𝑥 ( 𝑇𝑟𝑒 𝑓
) − 𝜇 𝑝𝑚𝑖𝑛
𝜇0 𝑝 = 𝜇 𝑝𝑚𝑖𝑛 + 𝑟𝑒 𝑓
(2.69)
𝑇 𝑥𝑖 𝑝
1 + ( 𝑇𝑟𝑒 𝑓
) (𝑁𝑡𝑜𝑡 𝑎𝑙 /𝑁 𝑝 ) 𝛼𝑝

where 𝑁𝑡𝑜𝑡 𝑎𝑙 is the local total impurity concentration (in #/𝑐𝑚 3 ), 𝑇𝑟 𝑒 𝑓 is a reference temperature
𝑟𝑒 𝑓 𝑟𝑒 𝑓
(300.15K), and T is the temperature (in degrees K). The parameters 𝑁 𝑛 and 𝑁 𝑝 are reference values for
the doping concentration. The analytic mobility model can be selected by using the statement
"mobmodel=analytic" in the netlist.
The parameters for the analytic mobility model are given in Table 2-181.

754
 Table 2-181. Analytic Mobility Parameters
Parameter Silicon GaAs
𝜇 𝑛𝑚𝑖𝑛 55.24 0.0
𝜇 𝑛𝑚𝑎𝑥 1429.23 8500.0
𝑟𝑒 𝑓
𝑁𝑛 1.072e17 1.69e17
nun -2.3 -1.0
xin -3.8 0.0
𝛼𝑛 0.73 0.436
𝜇 𝑝𝑚𝑖𝑛 49.70 0.0
𝜇 𝑝𝑚𝑎𝑥 479.37 400.0
𝑟𝑒 𝑓
𝑁𝑝 1.606e17 2.75e17
nup -2.2 -2.1
xip -3.7 0.0
𝛼𝑝 0.70 0.395

2.4.2.2. Arora Mobility

This mobility model is also an analytic model which depends on impurity concentration and temperature. It
comes from the work of Arora, et al. [44] and is based on both experimental data and the modified
Brooks-Herring theory of mobility. The equation for the mobility of electrons is:

𝑇 𝑒𝑥𝑛2
𝑇 𝜇 𝑛2 ( 𝑇𝑟𝑒 𝑓
)
𝑒𝑥𝑛1
𝜇0𝑛 = 𝜇 𝑛1 ( ) + (2.70)
𝑇𝑟 𝑒 𝑓 1 + ( 𝐶𝑛( 𝑁𝑡𝑜𝑡
𝑇
𝑎𝑙
) 𝑒𝑥𝑛3
) 𝛼𝑛
𝑇𝑟𝑒 𝑓

and the equation for the mobility of holes is:

𝑇 𝑒𝑥 𝑝2
𝑇 𝜇 𝑝2 ( 𝑇𝑟𝑒 𝑓
)
𝜇0 𝑝 = 𝜇 𝑝1 ( ) 𝑒𝑥 𝑝1 + (2.71)
𝑇𝑟 𝑒 𝑓 1 + ( 𝐶 𝑝 ( 𝑁𝑡𝑜𝑡
𝑇
𝑎𝑙
) 𝑒𝑥 𝑝3
) 𝛼𝑝
𝑇𝑟𝑒 𝑓

where
𝑇
𝛼𝑛 = 𝐴𝑛( ) 𝑒𝑥𝑛4 (2.72)
𝑇𝑟 𝑒 𝑓

and
𝑇
𝛼 𝑝 = 𝐴𝑝( ) 𝑒𝑥 𝑝4 (2.73)
𝑇𝑟 𝑒 𝑓

The Arora mobility model can be selected by including the statement "mobmodel=arora" in the netlist. The
parameters for the arora mobility model are given in Table 2-182.

755
 Table 2-182. Arora Mobility Parameters
Parameter Silicon GaAs
𝜇 𝑛1 88.0 8.5e3
𝜇 𝑛2 1252.0 0.0
Cn 1.26e17 1.26e17
An 0.88 0.0
exn1 -0.57 -0.57
exn2 -2.33 0.0
exn3 2.4 0.0
exn4 -0.146 0.0
𝜇 𝑝1 54.3 4e2
𝜇 𝑝2 407.0 0.0
Cp 2.35e17 2.35e17
Ap 0.88 0.0
exp1 -0.57 0.0
exp2 -2.23 0.0
exp3 2.4 0.0
exp4 -0.146 0.0

2.4.2.3. Carrier-Carrier Scattering Mobility

This mobility model is based on the work of Dorkel and Leturq [45]. It incorporates carrier-carrier
scattering effects, which are important when high concentrations of electrons and holes are present in the
device. This model also takes lattice scattering and ionized impurity scattering into account. One important
difference between the carrier-carrier scattering mobility model and the two previous mobility models
(analytic and arora models) is that the carrier-carrier scattering mobility model depends upon the actual
carrier concentrations in the device. This model is important for modeling breakdown as well as various
radiation effects, which often result in very high carrier densities.
The expressions for the carrier-carrier model are as follows:

𝑇 −𝛼
𝜇 𝐿 = 𝜇 𝐿0 ( ) (2.74)
𝑇𝑟 𝑒 𝑓

where 𝜇 𝐿 is the lattice mobility, which has to do with scattering due to acoustic phonons.

𝐴𝑇 3/2 𝐵𝑇 2 𝐵𝑇 2 −1
𝜇𝐼 = [𝑙𝑛(1 + )− ] (2.75)
𝑁 𝑁 𝑁 + 𝐵𝑇 2

where 𝜇 𝐼 is the impurity mobility which is related to the interactions between the carriers and the ionized
impurities.

756
 2 × 1017𝑇 3/2
𝜇 𝑐𝑐𝑠 = √ [𝑙𝑛(1 + 8.28 × 108𝑇 2 ( 𝑝𝑛) −1/3 )] −1 (2.76)
𝑝𝑛

where 𝜇 𝑐𝑐𝑠 is the carrier-carrier scattering mobility, which is very important when both types of carriers are
at high concentration.

√︄
6𝜇 𝐿 (𝜇 𝐼 + 𝜇 𝑐𝑐𝑠 )
𝑋= (2.77)
𝜇 𝐼 𝜇 𝑐𝑐𝑠

is an intermediate term and

1.025
𝜇 = 𝜇𝐿 [ − 0.025] (2.78)
1 + (𝑋/1.68) 1.43

is the carrier mobility. The carrier-carrier scattering mobility can be selected by including the statement
"mobmobel=carr" in the netlist. The parameters for the carrier-carrier mobility model are given in
Table 2-183.

757
 Table 2-183. Carrier-Carrier Mobility Parameters
Parameter Carrier Silicon GaAs
Al 𝑒− 1430.0 8.50e3
Bl 𝑒− -2.2 0.0
Ai 𝑒− 4.61e17 4.61e17
Bi 𝑒− 1.52e15 1.52e15
Al ℎ+ 495.0 4.0e2
Bl ℎ+ -2.2 0.0
Ai ℎ+ 1.00e17 1.00e17
Bi ℎ+ 6.25e14 6.25e14

2.4.2.4. Lombardi Surface Mobility Model

This mobility model combines expressions for mobility at the semiconductor-oxide interface and in bulk
silicon. It is based on the work of Lombardi et al. [46]. The overall mobility is found using Mathiessen’s
rule:

1 1 1 1
= + + (2.79)
𝜇 𝜇 𝑎𝑐 𝜇 𝑏 𝜇 𝑠𝑟

where 𝜇 𝑎𝑐 is the carrier mobility due to scattering with surface acoustic phonons, 𝜇 𝑏 is the carrier mobility
in bulk silicon, and 𝜇 𝑠𝑟 is the carrier mobility limited by surface roughness scattering.

The Lombardi model is a more physics-based surface mobility model. It is a semi-empirical model for
carrier mobility, and the expressions for the individual scattering mechanisms were extracted from
experimental data taken in appropriate experimental conditions.

The expressions used in this model are given below:

𝑏𝑛 𝑐𝑛𝑁 𝑒𝑥𝑛4
𝜇 𝑎𝑐,𝑛 = + (2.80)
𝐸 ⊥ 𝑇 (𝐸 ⊥ ) 1/3

is the expression for electron mobility for acoustic phonon scattering,

𝑏 𝑝 𝑐 𝑝𝑁 𝑒𝑥 𝑝4
𝜇 𝑎𝑐, 𝑝 = + (2.81)
𝐸 ⊥ 𝑇 (𝐸 ⊥ ) 1/3

is the expression for hole mobility for acoustic phonon scattering,

𝜇 𝑚𝑎𝑥,𝑛 − 𝜇 𝑛0 𝜇 𝑛1
𝜇 𝑏,𝑛 = 𝜇 𝑛0 + − (2.82)
1 + (𝑁/𝑐𝑟𝑛) 𝑒𝑥𝑛1 1 + (𝑐𝑠𝑛/𝑁) 𝑒𝑥𝑛2

758
is the expression for bulk mobility for electrons, where
𝑇 −𝑒𝑥𝑛3
𝜇 𝑚𝑎𝑥,𝑛 = 𝜇 𝑛2 ( ) (2.83)
𝑇𝑟 𝑒 𝑓

and 𝜇 𝑚𝑎𝑥, 𝑝 𝜇 𝑝1
𝜇 𝑏, 𝑝 = 𝜇 𝑝0 𝑒𝑥 𝑝(−𝑝𝑐/𝑁) + − (2.84)
1 + (𝑁/𝑐𝑟 𝑝) 𝑒𝑥 𝑝1 1 + (𝑐𝑠𝑝/𝑁) 𝑒𝑥 𝑝2

is the expression for bulk mobility for holes, where

𝑇
𝜇 𝑚𝑎𝑥, 𝑝 = 𝜇 𝑝2 ( ) −𝑒𝑥 𝑝3 (2.85)
𝑇𝑟 𝑒 𝑓

The expression for electrons for surface roughness scattering is

𝑑𝑛
𝜇 𝑠𝑟 ,𝑛 = ( ) (2.86)
𝐸 ⊥𝑒𝑥𝑛8

and the expression for holes for surface roughness scattering is

𝑑𝑝
𝜇 𝑠𝑟 , 𝑝 = ( ) (2.87)
𝐸 ⊥𝑒𝑥 𝑝8

The parameters for the lombardi surface mobility model are given in Table2-184.

759
 Table 2-184. Lombardi Surface Mobility Parameters
Parameter Silicon GaAs
𝜇 𝑛0 52.2 0.0
𝜇 𝑛1 43.4 0.0
𝜇 𝑛2 1417.0 1e6
crn 9.68e16 9.68e16
csn 3.43e20 0.0
bn 4.75e7 1e10
cn 1.74e5 0.0
dn 5.82e14 1e6
exn1 0.680 0.0
exn2 2.0 0.0
exn3 2.5 0.0
exn4 0.125 0.0
exn8 2.0 0.0
𝜇 𝑝0 44.9 0.0
𝜇 𝑝1 29.0 0.0
𝜇 𝑝2 470.5 1.0
crp 2.23e17 2.23e17
csp 6.1e20 0.0
bp 9.93e6 1e10
cp 8.84e5 0.0
dp 2.05e14 1e6
exp1 0.719 0.0
exp2 2.0 0.0
exp3 2.2 0.0
exp4 0.0317 0.0
exp8 2.0 0.0
pc 9.23e16 0.0

2.4.2.5. Edge Mobilities

Mobility values are calculated along the edge connecting two nodes. In the case of the analytic, arora, and
surface mobility models, the edge mobilities are calculated by taking the average of the mobilities at the
two nodes. Then, the mobility along the edge connecting nodes 1 and 2 is:

𝜇𝑒𝑑𝑔𝑒 = (𝜇[1] + 𝜇[2])/2.0 (2.88)

760
In the case of the carrier-carrier scattering mobility, the edge mobilities were calculated differently. The
electron and hole concentrations were first calculated at the midpoint of the edge using a "product" average
and then these values of "n" and "p" were used in the function to calculate the mobility at the midpoint of
the edge. For example, if n[1] and n[2] are the electron concentrations at nodes 1 and 2, the electron
concentration along the edge is given by:

√︁
𝑛𝑒𝑑𝑔𝑒 = 𝑛[1] ∗ 𝑛[2] (2.89)

Subsequently, the mobility at the midpoint of an edge is found by using the values of electron and hole
concentration at the midpoint of the edge when calling the function which returns the mobility,
calcMob().

𝑐𝑎𝑟𝑟𝑖𝑒𝑟
𝜇 𝑛,𝑒𝑑𝑔𝑒 = 𝑓 (𝑛𝑒𝑑𝑔𝑒 ) (2.90)

This method makes more sense, especially when the electron and hole concentrations vary by several orders
of magnitude. Then it approximates taking the average of the logarithms.

2.4.2.6. Boundary Conditions for Electrode Contacts

This section describes various boundary conditions that need to be applied to the semiconductor boundary.
Xyce is predominantly an analog circuit simulator, and the TCAD (PDE-based) device modeling that has
been implemented in Xyce takes external circuit information as input. This input consists of voltages and
currents which are applied as boundary conditions to the semiconductor domain.

The physical connection from the circuit to the device generally includes a variety of materials, including
metals and oxides. Electrical differences between the semiconductor and the contact material can result in a
potential barrier that must be included in the imposed voltage boundary condition.

There are three general types of contacts between the circuit and the TCAD device that are handled by
Xyce. The first is the "neutral" contact, in which it is simply assumed that the electrode material does not
impose any addition potential barrier to that of the Fermi level differences in the semiconductor. The
second is the Schottky contact, in which the electrode is a specified metal, and a potential barrier is
imposed to account for the workfunction difference between the metal and the semiconductor. The last type
of contact is the metal-oxide-semiconductor contact, in which the workfunction difference, and the voltage
drop across the oxide must be accounted for.

2.4.2.7. Neutral Contacts

A neutral contact refers to the case in which the contact is made to the semiconductor itself, and barrier
heights due to material differences are not considered. This is the simplest type of contact in Xyce, and
problems which use this type of contact are generally easier to solve, compared with other types of contacts.
In this case, the boundary is given by

𝑉𝑏𝑐 = 𝑉𝑐𝑘𝑡 + 𝑉𝑏𝑖 (2.91)

761
where 𝑉𝑐𝑘𝑡 is the potential applied by the circuit and 𝑉𝑏𝑖 is the "built-in" potential of the semiconductor. For
a p-type substrate, the built-in potential is given by

𝑘𝑇 𝑁𝐴
𝑉𝑏𝑖 = − 𝑙𝑛( ) (2.92)
𝑞 𝑛𝑖

and for an n-type substrate, the built-in potential is given by

𝑘𝑇 𝑁𝐷
𝑉𝑏𝑖 = 𝑙𝑛( ) (2.93)
𝑞 𝑛𝑖

Figure 2-8. Neutral Contacts.

𝑉𝑏𝑖 represents the extent of the energy band bending due to the doping of a device. While most of the
dramatic changes will happen away from the contact, near junctions, it is still incorporated into the voltage
boundary condition to maintain a flat potential near the contacts. Figure 2-8 shows the energy band
variation across a PN junction, and the corresponding electrostatic potential. This variation is due to the
internal physics of the device, and needs to be there even in the event of zero applied voltage. This is
partially enforced by the solution to Poisson’s equation, and also by the application of equation 2.91.

2.4.2.8. Schottky Contacts

In the case of a metal-semiconductor contact, it is necessary to add the workfunction difference, Φ𝑚𝑠 , to the
potential in the semiconductor [47]. Φ𝑚 is a constant for a given metal, and Φ𝑠 is a function of the doping
in the semiconductor. The workfunction potential, Φ, when multiplied by q, is the difference between the
Fermi level and vacuum in the material. In essence, the workfunction difference represents the distance

762
 Figure 2-9. Schottky Contact, N-type.

between the Fermi level in the metal and the Fermi level in the semiconductor when considering the
individual band structures.

In the case of an n-type semiconductor, the semiconductor workfunction can be represented as

Φ𝑠 = 𝜒 + (𝐸𝐶 − 𝐸 𝐹𝑆 )/𝑞 (2.94)

where 𝜒 is the electron affinity in the semiconductor and q𝜒 is the distance between the conduction band
and vacuum in the semiconductor. 𝐸𝐶 is the conduction band energy and 𝐸 𝐹𝑆 is the Fermi level of the
semiconductor. Rewriting this expression in terms of the doping concentration, it becomes

𝑁𝑑
Φ𝑠 = 𝜒 + 𝐸 𝑔 /2 − 𝑉𝑡 𝑙𝑛( ) (2.95)
𝑛𝑖

In the case of a p-type semiconductor, the semiconductor workfunction can be represented as

Φ𝑠 = 𝜒 + 𝐸 𝑔 /2 + (𝐸 𝑖 − 𝐸 𝐹𝑆 )/𝑞 (2.96)

where 𝐸 𝑖 is the intrinsic value of the Fermi level, and can be approximated as the halfway point between the
conduction band (𝐸𝐶 ) and the valance band (𝐸 𝑉 ). Rewriting this expression in terms of the doping
concentration

𝑁𝑎
Φ𝑠 = 𝜒 + 𝐸 𝑔 /2 + 𝑉𝑡 𝑙𝑛( ) (2.97)
𝑛𝑖

763
 Figure 2-10. Schottky Contact, P-type.

For the TCAD devices in Xyce, for a node at a metal-semiconductor contact, the quantity Φ𝑚 − Φ𝑠 is added
to the potential at the node to account for the metal-semiconductor barrier. The current values of metal
workfunctions used in Xyce are given in Table 2-185. The values for electron affinity are given in
Table 2-186. The boundary condition for a metal electrode in Xyce is given by

𝑉𝑏𝑐 = 𝑉𝑐𝑘𝑡 + 𝑉𝑏𝑖 + Φ𝑚𝑠 (2.98)

where 𝑉𝑐𝑘𝑡 is the potential applied by the circuit to the electrode and 𝑉𝑏𝑖 is the "built-in" potential of the
semiconductor, a function of the semiconductor doping.

764
 Table 2-185. Material workfunction values
Metal Symbol Workfunction, Φ𝑚 (Volts)
aluminum al 4.10
p+-polysilicon ppoly 5.25
n+-polysilicon npoly 4.17
molybdenum mo 4.53
tungsten w 4.63
molybdenum disilicide modi 4.80
tungsten disilicide wdi 4.80
copper cu 4.25
platinum pt 5.30
gold au 4.80

Table 2-186. Electron affinities

Semiconductor Symbol Electron Affinity, 𝜒 (Volts)
Silicon si 4.17
Germanium ge 4.00
Galium Arsenide gaas 4.07
Silicon Dioxide sio2 0.97
Nitride nitride 0.97
Sapphire sapphire 0.97

765
2.4.2.9. Metal-Oxide-Semiconductor Contacts

To date in Xyce, only semiconductor material is included in the PDE solution domain. Metals and oxide
materials are only included through boundary conditions. This is an adequate approach for a lot of
problems. For some problems (such as modeling of low-dose radiation effects) modeling the oxide in more
detail, as a PDE, will become necessary. However, since oxides are usually very thin, compared with the
semiconductor domain, meshing both materials as part of the same simulation is difficult. Therefore,
incorporating the effects of a gate oxide as part of the gate boundary condition is a reasonable approach.
In the case of a contact to a metal-oxide-semiconductor structure, the separation of the Fermi energies in
the metal and the semiconductor at equilibrium is due to two effects: the workfunction difference between
the metal and the semiconductor, and the effective interface charge. These two effects cause the bands to
bend at the surface in equilibrium. The flatband voltage is the sum of these two terms [47]:

𝑄𝑖
𝑉𝐹 𝐵 = Φ𝑚𝑠 − (2.99)
𝐶𝑖

where Φ𝑚𝑠 is the metal-semiconductor workfunction difference, 𝑄 𝑖 is the value of interface charge (in
𝐶/𝑐𝑚 2 ), and 𝐶𝑖 is the oxide capacitance per unit area, which is given by

𝜖 𝑜𝑥 𝜖 0
𝐶𝑖 = (2.100)
𝑥𝑜

The voltage 𝑉𝐹 𝐵 is the amount of bias which, when applied to the gate, causes the electron energy bands to
be flat. This is the potential that is added to a boundary node in Xyce to account for a
metal-oxide-semiconductor barrier. The overall boundary condition for a contact to a
metal-oxide-semiconductor structure is given by

𝑉𝑏𝑐 = 𝑉𝑐𝑘𝑡 + 𝑉𝑏𝑖 + Φ𝑚𝑠 − 𝑄 𝑖 /𝐶𝑖 (2.101)

where 𝑉𝑐𝑘𝑡 is the potential applied by the circuit and 𝑉𝑏𝑖 is the "built-in" potential of the semiconductor.

2.4.2.10. NMOS Device

The default NMOS device used in Xyce has a substrate doping concentration of 1.0 × 1016 /𝑐𝑚 3 and an
oxide thickness of 1.0 × 10−6 𝑐𝑚. Since the ideal threshold voltage 𝑉𝑇 is given by

√︄
𝜖𝑠 2𝑞𝑁 𝐴 𝜙 𝐹
𝑉𝑇 = 2𝜙 𝐹 + 𝑥𝑜 (2.102)
𝜖 𝑜𝑥 𝜖 𝑠 𝜖0

𝑉𝑇 is equal to 0.892 V. for this device. Note that

1 𝑘𝑇 𝑁𝐴
𝜙𝐹 = [𝐸 𝑖 (𝑏𝑢𝑙 𝑘) − 𝐸 𝐹 ] = 𝑙𝑛( ) (2.103)
𝑞 𝑞 𝑛𝑖

for a p-type semiconductor substrate and

766
 𝑘𝑇 𝑁𝐷
𝜙𝐹 = − 𝑙𝑛( ) (2.104)
𝑞 𝑛𝑖

for an n-type substrate.

767
This page intentionally left blank.

768
3. COMMAND LINE ARGUMENTS

Xyce supports a handful of command line arguments which must be given before the netlist filename. While
most of these are intended for general use, others simply give access to new features that, while supported,
are not enabled by default. The general usage is as follows:
Xyce [arguments] <netlist filename>
Table 3-1 gives a list of supported command line options.1

Table 3-1. List of Xyce command line arguments.

Argument Description Usage Default
-h Help option. Prints usage and exits. -h -
-v Prints the version banner and exits. -v -
-license Prints the license text and exits. -license -
Prints a list of compiled-in options
-capabilities -capabilities -
and exits.
-delim
-delim Set the output file field delimiter. -
<TAB|COMMA|string>
Place the output result(s) into file(s)
with the specified basename and the output file
-o appropriate extension (e.g., -o <basename> name(s) based on
<basename>.prn or netlist file name.
<basename>.mt0).
Place the log output into specified Log output sent
-l -l <file>
file. to standard out.
No rawfile
-r Output a binary rawfile. -r <file>
written.
Use with -r to output a readable
(ASCII) rawfile. Without the
concurrent use of -r, -a will cause -r <file> -a Default rawfile is
-a
all .PRINT lines that use -a binary.
FORMAT=RAW to produce ASCII
rawfiles.
-nox Use the NOX nonlinear solver. -nox <ON|OFF> on
KLU for serial and
-linsolv <KLU| small circuits in
-linsolv Set the linear solver. KSPARSE|SUPERLU| parallel. AztecOO
AZTECOO|BELOS> for large circuits
in parallel.

1 Note that the “-h” option might list command line options not present in this table. These extra options are generally deprecated
and should not be used. Only the options listed in the table are considered supported features.

769
 Table 3-1. List of Xyce command line arguments.
Argument Description Usage Default
Print a terse summary of model
-param and/or device parameters, and -param -
default values.
Specify a file with simulation
-prf -prf <filename> -
parameters.
Specify a file to save simulation
-rsf -rsf <filename> -
responses functions.
Recompute .measure and/or .fft
-remeasure -remeasure -
results with existing data.
-syntax Check netlist syntax then exit. -syntax -
Check netlist syntax and topology,
-norun -norun -
then exit.
Suppress some of the
-quiet simulation-progress messages sent -quiet -
to stdout for transient simulations.
Output a list of all solution
-namesfile variables generated by the netlist -namesfile <filename> -
into <filename>
Output a list of all noise sources for
-noise_names_file
-noise_names_file all devices generated by the netlist -
<filename>
into <filename>
Set random number seed for If not provided,
expression library’s random number Xyce will
-randseed -randseed <number>
functions and also .SAMPLING generate a seed
analysis internally.
-maxord Maximum time integration order. -maxord <1..5> -
Maximum number of warning
-max-warnings -max-warnings <#> 100
messages.
-jacobian_test Jacobian matrix diagnostic. -jacobian_test -
ignore, which
Set precendence for multiply means use the last
defined parameters. Can be set to -redefined_params duplicate
-redefined_params
ignore (use last), usefirst, warn <option> parameter in the
or error. netlist without
warning or error.
The second
argument will
Hspice parser extensions. This will
-hspice-ext determine which
override several parser issues that
-hspice-ext <all|separator features are
are difficult to translate with scripts.
|units|math> enabled. For
details see
section 3.1.

A few other command line options are available that are typically only used in Xyce development. For
example the options -param, -info, -doc and -doc_cat are used to generate the device tables in this
guide. The options -jacobian_test and -namesfile can be useful in debugging new devices in Xyce.

770
The option -namesfile is also useful for determining the “fully qualified node names”, including the
subcircuit hierarchy, for nodes and internal variables for mutual inductors. The .PRINT section 2.1.31 has
more information on, and examples for, the -namesfile command line option.

3.1. Hspice extensions

The command line argument -hspice-ext is more complicated than most, so its explanation is given in
this section. The second argument to -hspice-ext can either be a single string or can be a comma
separated list.
• Using all will enable all the features.
• Using units sets the use of the “atto” prefix, designated by “a”, to mean a mutliplier of “1E-18”.
Without this setting, the “a” prefix means units of amperes.
• Using math sets the use of Hspice style math operators. Most importantly, it forces “log” to mean the
natural logarithm, rather than the base 10 logarithm.
• Using separator sets the use of the period “.” for subcircuit name separation. If this is not set, the
colon “:” is used instead.

3.2. Redefined parameters

Xyce, like most circuit simulators, permits multiple parameters of the same name in a netlist. However,
when multiple parameters have the same name, the default behavior of Xyce may different from other
codes. The -redefined_params command line option can be used to ensure compatibility. The available
options are the following:
• ignore is the default behavior, and will result in Xyce using the last parameter found in the netlist,
while emitting no warnings.
• uselast is synonymous with ignore.
• error will result in Xyce exiting with a fatal error if multiple parameters have the same name.
• warn will result in Xyce behaving the same as the default behavior (using the last parameter) but will
emit a warning for every duplicate parameter to the terminal.
• warning is synonymous with warn.
• uselastwarn is synonymous with warn.
• usefirst will result in Xyce using the first paramter of a given name in the netlist, without warnings
or errors. This is the behavior of some other simulators.
• usefirstwarn is the same as usefirst, except that it will emit a warning for every duplicate
parameter to the terminal.
The options that emit warnings or error messages to terminal output can result in very verbose output for
large circuits. For that reason warnings are not enabled as the default behavior. However, these options can
be useful for debugging netlists.

771
This page intentionally left blank.

772
4. RUNTIME ENVIRONMENT

4.1. Running Xyce in Serial

After ensuring that the directory into which Xyce was installed is in your PATH variable, one merely
executes the code by running the command, Xyce with the desired netlist name appended.

4.2. Running Xyce in Parallel

Open MPI must be installed on the host machine. It may be download from
http://www.open-mpi.org/. Consult the documentation for help with installation.
After ensuring that the both the directory into which Xyce was installed and the directory in which mpirun
is found are in your PATH variable, one merely executes the code by running the command, mpirun
[mpirun options] Xyce [xyce options] with the desired netlist name appended.

4.3. Running Xyce on Sandia HPC and CEE Platforms

This version of Xyce has been installed centrally on Sandia HPC and CEE platforms, and requires
metagroup access. Contact the Xyce team for details on how to obtain this access.
Once you have registered for metagroup membership, the central installs of Xyce may be accessed by a
module load.
module load Xyce adds all required modules and sets all required environment variables to access the
normal version of Xyce. module load XyceRad does the same thing for the version Xyce containing
Sandia proprietary models.
module help Xyce provides some additional information about what the module does.
Consult the system documentation for help with submitting jobs on these platforms.
https://computing.sandia.gov/

773
This page intentionally left blank.

774
5. SETTING CONVERGENCE PARAMETERS FOR XYCE

Because the solution algorithms and methods within Xyce are different than those used by other circuit
simulation tools (e.g., SPICE), the overall convergence behavior is sometimes different, as are the
parameters which control this behavior.

5.1. Adjusting Transient Analysis Error Tolerances

Xyce uses a variable order trapezoid integration as its default scheme, and this method may also be
requested explicitly with the TIMEINT option METHOD=trap or METHOD=7. Trapezoid time-stepping is
second order accurate and does not have any numerical dissipation in its local truncation error. Variable
order trapezoid integration dynamically uses Backward Euler (BE) and trapezoid rule. When ERROPTION=1
is set with METHOD=7, trapezoid rule is used almost exclusively (BE only used at breakpoints). See table 2-5
for details.

Another time integration option is the second-order Gear method. It may be selected with the TIMEINT
option METHOD=gear or METHOD=8. See table 2-5 for details.

5.1.1. Setting RELTOL and ABSTOL

In Xyce, both the time integration package and the nonlinear solver package have RELTOL and ABSTOL
settings. Some general guidelines for settings parameters are [48]:

• Use the same RELTOL and ABSTOL values for both the TIMEINT and the
NONLIN-TRAN .OPTIONS statements.

• For a conservative approach (i.e., safe), set RELTOL=1.0E-(m+1) where m is the desired number of
significant digits of accuracy.

• Set ABSTOL to the smallest value at which the solution components (either voltage or current) are
essentially insignificant.

• Note that the above suggests that ABSTOL < RELTOL.

The current defaults for these parameters are ABSTOL=1.0E-6 and RELTOL = 1.0E-3. For a complete list
of the time integration parameters, see chapter 2.1.

775
5.2. Adjusting Nonlinear Solver Parameters (in transient mode)

In Xyce, the nonlinear solver options for transient analysis are set using the .OPTIONS NONLIN-TRAN line
in a netlist. This subsection gives some guidelines for setting this parameters.
• For guidelines on setting RELTOL and ABSTOL, see above.
• RHSTOL – This is the maximum residual error for each nonlinear solution. Xyce uses this as a
“safety” check on nonlinear convergence. Typically, 1.0E-2 (the default) works well.
• DELTAXTOL – This is the weighted update norm tolerance and is the primary check for nonlinear
convergence. Since it is weighted (i.e., normalized using RELTOL and ABSTOL), a value of 1.0
would give it the same accuracy as the time integrator. For robustness, the default is 0.33 but
sometimes a value of 0.1 may help prevent “time-step too small” errors. A value of 0.01 is
considered quite small.
• MAXSTEP – This is the maximum number of Newton (nonlinear) steps for each nonlinear solve. In
transient analysis, the default is 20 but can be increased to help prevent “time-step too small” errors.
This is roughly equivalent to ITL4 in SPICE.

776
6. QUICK REFERENCE FOR USERS OF OTHER SPICE CIRCUIT SIMULATORS

This chapter describes many of the differences between Xyce and other SPICE-like circuit simulaters. The
primary focus is on the difference between Orcad PSpice and Xyce, with an eye towards providing the
ability for those familiar with using PSpice to begin using Xyce quickly. PSpice compatibility was an early
focus for the Xyce project.
The Xyce team also supports a netlist translation tool, XDM (for Xyce Data Model) [49] . This tool
supports translation of netlists in Pspice, Hspice and Spectre format into Xyce formatted netlist files. There
is also a Xyce command line option (–hspice-ext all) which is designed to assist with a handful of
Hspice syntax issues that are particularly difficult to translate.
This chapter is likely not complete, and Xyce users might also consult specific sections of this Reference
Guide about particular Xyce commands. Those sections may have additional information on Xyce’s
incompatibilities with other circuit simulators, and how to work around them.

6.1. Differences Between Xyce and PSpice

This section is focused on the differences between Xyce and PSpice. However, some of this discussion also
applies to other SPICE-like circuit simulators.

6.1.1. Command Line Options

Command line arguments are supported in Xyce but they are different than those of PSpice. For a complete
reference, see Chapter 3.

6.1.2. Device Support

Most, but not all, devices commonly found in other circuit simulation tools are supported. Xyce also
contains enhanced versions of many semiconductor devices that simulate various environmental effects. For
the complete list, please see the Analog Device Summary in Table 2-35.

6.1.3. .OPTIONS Support

For the specific devices or models that are supported in Xyce, most of the standard netlist inputs are the
same as those in standard SPICE. However, the .OPTIONS command has several additional features used
to expose capabilities specific to Xyce. In particular, Xyce does not support the standard PSpice format
.OPTIONS line in netlists. Instead, options for each supported package are called according to the
following format.

General Form .OPTIONS <pkg> [<tag>=<value>]*

777
Arguments and
Options
DEVICE
Device Model
TIMEINT
Time Integration
NONLIN
Nonlinear Solver
NONLIN-TRAN
Transient Nonlinear Solver
NONLIN-HB
HB Nonlinear Solver
LOCA
Continuation/Bifurcation Tracking
LINSOL
Linear Solver
LINSOL-HB
HB Linear Solver
OUTPUT
Output
RESTART
Restart
HBINT
Harmonic Balance (HB)
SENSITIVITY
Direct and Adjoint sensitivity analysis

For a complete description of the supported options in Xyce, see section 2.1.25.

Known caveats are that the ABSTOL options have different definitions in PSpice and Xyce. Also, a PSpice
.OPTIONS VNTOL=<value> line can be mapped into these two Xyce lines:

.OPTIONS NONLIN ABSTOL=<value>

.OPTIONS NONLIN_TRAN ABSTOL=<value>

The PSpice ITL1 and ITL4 options are similar to the Xyce MAXSTEPS. In PSpice, ITL1 affects .DC
analyses, while ITL4 affects .TRAN analyses. In Xyce, .OPTIONS NONLIN refers to options for .DC
analyses, while .OPTIONS NONLIN-TRAN refers to options for .TRAN analyses. So, a feasible mapping is
PSpice .OPTIONS ITL1=20 becomes .OPTIONS NONLIN MAXSTEP=20 in Xyce. Similarly, PSpice
.OPTIONS ITL4=20 becomes .OPTIONS NONLIN-TRAN MAXSTEP=20 in Xyce. However, given that
PSpice and Xyce use different default values for ITL1 and ITL4 vs. MAXSTEPS, the best approach may be to
not translate the ITL1 and ITL4 lines into the corresponding Xyce netlist.

778
6.1.4. .PROBE vs. .PRINT

Xyce does not support the “.PROBE” statement. Output of Probe-format files, in .csd format, that are
readable by PSpice is done using the .PRINT netlist statement. See section 2.1.31 for the syntax for
FORMAT=PROBE. That section also describes wildcard support and access to subcircuit nodes in Xyce, both
of which are different than PSpice.

Xyce does not support PSpice style abbreviations in the .PRINT statement. For example, to print out the
value of the voltage at node A in a transient simulation you must request .PRINT TRAN V(A), not
.PRINT TRAN A. Xyce also does not support N() as a synonym for V() on .PRINT lines.

6.1.5. Converting PSpice ABM Models for Use in Xyce

Xyce is almost fully compatible with PSpice with respect to analog behavioral models. This includes the E,
F, G, and H device types. A notable exception to this compatibility is in the use of lead and device currents
in expressions used in controlled source definitions. That feature is not supported in Xyce. In addition, the
FREQ,LAPLACE and CHEBYSHEV forms for E and G sources or the ERROR qualifier are not supported in
Xyce..

6.1.6. Usage of .STEP Analysis

The implementation of .STEP in Xyce is not the same as that of PSpice. See section 2.1.36 for the syntax
and function of the .STEP function in Xyce.

6.1.6.1. Model Parameter Sweeps

PSpice requires extra keywords to apply a .STEP statement to a model parameter. Xyce handles model
parameters differently, and is actually somewhat more flexible than PSpice. Unfortunately, this means that
the two specifications are not compatible.

A model parameter in PSpice would be handled like this:

R1 1 2 RMOD 1
.model RMOD RES(R=30)
.step RES RMOD(R) 30 50 5

The equivalent way to specify this in Xyce would be:

R1 1 2 RMOD 1
.model RMOD RES(R=30)
.step RMOD:R 30 50 5

Note that Xyce does not require the RES keyword on the .STEP line. In PSpice, this keyword is needed to
specify what type of model is being used. Xyce actually has more flexibility than PSpice in this regard—any
model or instance variable can be set on the .STEP line using the same syntax.

Example: .step D101:IS 1.0e-3 5.0e-3 1.0e-3

779
In this example, D101 is the name of a model or instance, and IS is the name of the parameter within that
model or instance.

6.1.7. Behavioral Digital Devices

There are at least four significant differences. First, the instance line syntax for the Xyce digital behavioral
devices differs from PSpice. Second, Xyce uses one model card for the timing and Input/Output (I/O)
characteristics, while PSpice uses separate model cards for timing and I/O characteristics. The model cards
also have different parameters. Third, Xyce does support the DIGINITSTATE option. However, it has a
different default value than in PSpice. So, the DCOP calculations for flip-flops and latches may be different
in some cases between Xyce and PSpice. Finally, closely spaced input transitions to a gate (e.g., ones spaced
by less than the DELAY parameter of the Xyce model) may produce different behaviors in Xyce and PSpice.
Please consult Section 2.3.28 for more details.

6.1.8. Power Dissipation

PSpice supports printing the power dissipation of a device via syntax like W(<name>). At this time, not all
Xyce devices support power calculations. In addition, the Xyce results for the FET semiconductor devices
(J, M and Z devices) may differ from the PSpice results. Consult the Features Supported by Xyce Device
Models table in Section 2.3 and the individual sections on each device for more details. Additional
limitations on lead current and power calculations in Xyce are given in Section 2.3.3.
Example work-arounds are as follows, using either the node voltage at Node 2 or the lead current through
Resistor 2:

.DC V1 0 5 1
.param R2VAL=10
V1 1 0 5V
R1 1 2 10
R2 2 0 {R2VAL}
.PRINT DC V(2) {V(2)*V(2)/R2VAL} {I(R2)*I(R2)*R2VAL}

6.1.9. Dependent Sources with TABLE Syntax

The documented PSpice syntax for the TABLE form of the E and G sources is identical to the Xyce syntax
for those two devices. As an example, consider this E-source netlist line which conforms to the documented
PSpice and Xyce syntaxes:
E5 5 0 TABLE V(1,0) = (-2,-3) (2,3)
There is an equal sign between the expression {V(1,0)} and the list of value pairs (e.g., before (-2,-3)).
There is also a comma between the two values in each set of value pairs. However, it has been observed that
some PSpice versions will accept variants of the documented PSpice syntax. As examples, PSpice might
use this TABLE syntax, where the equal sign between the expression and the list of value pairs is missing
and there is an extra set of parentheses around the list of value pairs:
TABLE {EXPR} ((x1,y1) (x2,y2) ... (xn, yn))

780
PSpice might also specify the TABLE syntax without the commas between the two values in each set of
value-pairs. For example, this is a legal syntax in some PSpice versions:

TABLE {EXPR} = (x1 y1) (x2 y2) ... (xn yn)

So, the generic solution is to change these alternative PSpice syntaxes (and possibly others) to conform with
the Xyce E and G source TABLE syntax, which is (see also Sections 2.3.12 and 2.3.14):

TABLE {EXPR} = (x1,y1) (x2,y2) ... (xn, yn)

6.1.10. MODEL STATEMENTS

In PSpice, some .MODEL statements may have commas separating the list of parameters, which causes
problems in Xyce. A simple workaround is to replace those commas with spaces in the corresponding Xyce
.MODEL statements.

In PSpice, some .MODEL statements may not have parentheses surrounding the list of parameters. While
Xyce also does not require parentheses in model cards, parentheses are accepted. The only Xyce requirement
is that if they are used then they must be paired with one left parenthesis before all of the parameters and
one right parentheses after all of the parameters. It is an error to have unmatched parentheses.

PSpice syntaxes where only a subset of the model parameters are enclosed within parentheses are also not
supported in Xyce. A PSpice example is:

.model somebjt NPN Is=1e-16 (Xti=3 Bf=100) Eg=1.11 NC=2

Nested parentheses, as is often seen when a DEV (deviation) is specified for a parameter in a PSpice model
statement, are also not allowed in Xyce. A PSpice example is:

.model someotherbjt NPN(Is=1e-16 Xti=3 (Bf=100 DEV=5%) Eg=1.11 NC=2)

The previous PSpice example also raises the issue of model parameters that are supported in PSpice but not
in Xyce. It that case, Xyce will issue a warning about the invalid parameter and the simulation will run.

Another common issue is a PSpice model parameter (e.g., BV=) without a value. That PSpice syntax error is
often silently ignored in PSpice, but flagged as a parsing error in a Xyce netlist.

Temperature coefficient (TC) specifications can be a problem also. The documented PSpice syntax is this,
with a comma between the two values.

TC=0.1,0.1

However, it has been observed that some PSpice versions allow the TC parameter to omit the comma
between those two values. That is not legal in Xyce.

6.1.11. .NODESET and .IC Statements

Xyce and PSpice differ in their capabilities to handle .NODESET and .IC statements in subcircuits. See
sections 2.1.22 and 2.1.14 for more details.

781
6.1.12. Piecewise Linear Sources

The preferred Xyce syntax for PWL sources does not use parentheses or commas within the time-voltage
pair listing. See Section 2.3.9 for more details.

The Xyce PWL source does not support the PSpice .IN format for file input. See Section 2.3.9 for the
ASCII text and .csv formats supported by Xyce for file input.

The Xyce repeat R=<value> syntax for PWL sources is not compatible with the PSpice REPEAT syntax for
PWL sources. Some work-arounds are as follows. This PSpice REPEAT FOREVER syntax:

VPWL1 1 0 PWL REPEAT FOREVER (0,0) (0.5,1) (1,0)

+ ENDREPEAT

is equivalent to this Xyce syntax:

VPWL1 1 0 PWL 0 0 0.5 1 1 0 R=0

Similarly, if the PSpice source has its time-voltage pairs in a .csv file, and the specified waveform starts at
time=0, then this PSpice syntax:

VPWL2 2 0 PWL
+ REPEAT FOREVER
+ FILE "data.csv"
+ ENDREPEAT

is equivalent to this Xyce syntax:

VPWL2 2 0 PWL file "data.csv" R=0

For more general PSpice REPEAT syntaxes, and especially for the PSpice REPEAT for N syntax, the user
might have to manually duplicate the PSpice waveform in a .csv file.

6.1.13. .AC Output

The Xyce .csd file format for a .AC analysis is different than the PSpice format, but is still viewable in the
PSpice A/D waveform viewer. This PSpice .PROBE statement:

.PROBE/CSDF V([1b]) VR([1b]) VI([1b])

will produce #N and #C lines in its netlistName.csd file like this, where the real and imaginary parts of
V(1b) are output for each data point on the #C line. The end-user can then use the PSpice A/D UI to choose
to plot the VR and VI quantities.

#N
‘V(1b)’ ‘V(1b)’ ‘V(1b)’
#C 1.0000000000E01 3
2.470E-02/-1.552E-01:1 2.470E-02/-1.552E-01:2 2.470E-02/-1.552E-01:3

This corresponding Xyce .PRINT AC statement:

.PRINT AC FORMAT=PROBE V(1b) VR(1b) VI(1b)

782
will produce #N and #C lines in its netlistName.csd file like this, where the real and imaginary parts of
V(1b) are still output on the #C line. However, in Xyce, the VR() and VI() operators return real-valued
quantities as shown below. This Xyce formatted file is still viewable in PSpice A/D.
#N
‘V(1b)’ ‘VR(1b)’ ‘VI(1b)’
#C 1.0000000000E01 3
2.470e-02/-1.552e-01:1 2.470e-02/0.000e+00:2 -1.552e-01/0.000e+00:3

6.1.14. Additional differences

Some other differences between Xyce and PSpice are described in Table 6-1. Users should also consult
Table 6-2, since that table lists more general incompatibilities that span multiple circuit simulators.

Table 6-1. Incompatibilities with PSpice.

Issue Comment
Xyce interprets log() in an expression as a base-10 log() while PSpice
𝑙𝑜𝑔() interprets log() as a natural log() function.
.VECTOR, .WATCH, and .PLOT
output control analysis are not Xyce does not support these commands.
supported.
.PZ analysis is not supported. Xyce does not support this command.
.DISTO analysis is not supported. Xyce does not support this command.
.TF analysis is not supported. Xyce does not support this command.
.AUTOCONVERGE is not supported. Xyce does not support this command.
The Xyce version of .SENS requires that the user specify exactly which
.SENS analysis is supported, but parameters are the subject of the sensitivity analysis. Additionally, Xyce
has a different syntax than PSpice. can compute sensitivities in transient and .AC as well as the .DC case
(unlike PSpice).
.NOISE analysis is supported, but The Xyce version of .NOISE is new enough that not all noise models
not all devices supported. have been implemented.
Xyce does not support these commands directly. However, Xyce does
.MC and .WCASE statistical analyses support a variety of UQ methods, including .SAMPLING,
are not supported. .EMBEDDEDSAMPLING and various polynomial chaos expansion
methods.
.DISTRIBUTION, which defines a Xyce does not support this command. This command goes along with
user distribution for tolerances, is .MC and .WCASE statistical analyses, which are also not directly
not supported. supported.
.LOADBIAS and .SAVEBIAS initial
condition commands are not Xyce does not support these commands.
supported.
.ALIASES, .ENDALIASES, are not
Xyce does not support these commands.
supported.
.STIMULUS is not supported. Xyce does not support this command.
.TEXT is not supported. Xyce does not support this command.
Xyce does not support this. Use the FORMAT=PROBE option of .PRINT
.PROBE does not work
instead. See section 2.1.31 for syntax.

783
 .OP is supported in Xyce, but will not produce the extra output normally
.OP only produces output in serial
associated with the .OP statement, if running a parallel build.
A requested pulsed source rise/fall time of zero really is zero in Xyce. In
Pulsed source rise time of zero other simulators, requesting a zero rise/fall time causes them to use the
printing interval found on the tran line.
Mutual Inductor Model Not the same as PSpice. This is a Sandia developed model.
Output variables have to be specified as a V(node) or I(source). Listing
.PRINT line shorthand
the node alone will not work.
BSIM3 level In Xyce the BSIM3 level=9. In PSpice the BSIM3 is level=7.
Interactive mode Xyce does not have an interactive mode.
Xyce has much tighter default solver tolerances than some other
simulators (e.g., PSpice), and thus often takes smaller time steps. As a
result, it will often take a greater number of total time steps for a given
Time integrator default tolerances
time interval. To have Xyce take time steps comparable to those of
PSpice, set the RELTOL and ABSTOL time integrator options to larger
values (e.g., RELTOL=1.0E-2, ABSTOL=1.0E-6).
Xyce does not support PSpice style .OPTION statements. In Xyce, the
.OPTIONS statements various packages all (potentially) have their own separate .OPTIONS line
in the netlist. For a complete description, see section 2.1.25.
Xyce does support a maximum time step-size control on the .tran line,
but we discourage its use. The time integration algorithms within Xyce
use adaptive time-stepping methods that adjust the time-step size
according to the activity in the analysis. If the simulator is not providing
enough accuracy, the RELTOL and ABSTOL parameters should be
DTMAX decreased for both the time integration package (.OPTIONS TIMEINT)
and the transient nonlinear solver package (.OPTIONS NONLIN-TRAN).
We have found that in most cases specifying the same maximum
timestep that PSpice requires for convergence actually slows Xyce down
by preventing it from taking larger timesteps when the behavior
warrants.
PSpice requires the use of a keyword UIC on the .TRAN line in order to
use initial conditions via IC keywords on instance lines. Doing so also
tells PSpice not to perform an operating point calculation. In Xyce, UIC
is ignored and produces a warning message. Xyce always uses initial
conditions specified with IC keywords, and the case of inductors and
.TRAN “UIC” keyword capacitors automatically inserts a fictitious voltage source around the
device that guarantees the correct potential drop across the device
during the operating point. If the user desires that Xyce not perform an
operating point calculation, but rather use an initial condition for a
transient run of all zero voltages, then the user should specify NOOP
instead.
Device temperatures in Xyce are specified through the .OPTIONS
Temperature specification DEVICE line. PSpice allows a .TEMP line that is not recognized (and is
ignored) by Xyce.
PSpice uses A and B to reference the two terminals of the lossless
Lead currents for lossless
tranmission line. So, Xyce uses I1() and I2(), while PSpice uses IA()
transmission lines
and IB() to access the lead currents for the device.
The use of those characters is fine in Xyce comment lines. It may be best
Extended ASCII characters in .LIB
to replace them with the printable equivalent on other Xyce netlist lines
files
though.

784
6.1.15. Translating Between PSpice and Xyce Netlists

Some internal Sandia users have found the following checklist to be helpful in getting their PSpice netlists
to run in Xyce. Additional changes may be needed in some cases.
For the .cir file:

• Change .LIB references to point to the modified libraries generated for use with Xyce.
• Change PROBE and PROBE64 statements to PRINT <Sim Type>
• Find cases where the PSpice netlist used N() rather than V().
• .DC has the keyword PARAM in PSpice. If it exists then remove it in the Xyce netlist.
• .OPTIONS TNOM=X is changed to .OPTIONS DEVICE TNOM=X in the Xyce netlist.
• .TEMP args does not exist in Xyce. The equivalent Xyce statement is .STEP TEMP LIST args
• The default time integrator tolerances can make Xyce take smaller timesteps on some circuits, and
therefore have slower simulation times. The Xyce timesteps can be increased at the expense of time
integration accuracy by loosening the integrator tolerances. Some users find that .OPTIONS
TIMEINT RELTOL=1e-2 ABSTOL=1e-4 leads to time steps more like PSpice’s.
• Move any .IC and .NODESET statements to the top-level, and use the fully qualified node names in
those statements.
• Adjust the syntax for any PWL sources, if needed, per Section 6.1.12.

For the .lib file:

• Add LEVEL=2 parameter to diode models.
• Fix the parentheses and comma differences between PSpice and Xyce.MODEL statements per Section
6.1.10.
• Find and modify any nested expression statements. This may entail replacing “{” with “(” in the
expression in the Xyce netlist.
• Fix the table syntax for dependent sources, as discussed in Section 6.1.9.

6.2. Differences Between Xyce and Other SPICE Simulators

This section covers some known differences between Xyce and other SPICE-like circuit simulators, besides
PSpice, as listed in Table 6-2. However, users of those other simulators (e.g., SPICE3F5, HSPICE, ngspice,
...) should also check the previous subsection on PSpice, since some of that discussion also applies here.

Table 6-2. Incompatibilities with Other Circuit Simulators.

Issue Comment
The .DC sweep calculation does not automatically output the
sweep variable. Only variables explicitly listed on the .PRINT
.DC sweep output.
line are output.

785
 In Xyce the MOSFET levels are not the same. In Xyce, a
MOSFET levels. BSIM3 is MOSFET level 9. Other simulators have different
levels for the BSIM3.
In Xyce the BSIM SOI (v3.2) is MOSFET level 10. Other
BSIM SOI v3.2 level.
simulators have different levels for the BSIM SOI.
In Xyce the BSIM4 is MOSFET levels 14 and 54. Other
BSIM4 level.
simulators have different levels for the BSIM4.
The manner of specifying a model parameter to be swept is
Syntax for .STEP is different. slightly different than in some other simulators. See the Xyce
Users’ and Reference Guides for details.
The Xyce switches are not compatible with the simple switch
implementation in SPICE3F5. The switch in Xyce smoothly
transitions between the ON and OFF resistances over a small
range between the ON and OFF values of the control signal
(voltage, current, or control expression). See the Xyce
Reference Guide for the precise equations that are used to
Switch is not the same as
compute the switch resistance from the control signal values.
SPICE3F5.
The SPICE3F5 switch has a single switching threshold voltage
or current, and RON is used above threshold while ROFF is
used below threshold. Xyce’s switch is considerably less likely
to cause transient simulation failures. Results similar to
SPICE3F5 can be obtained by setting VON and VOFF to the
same threshold value, but this is not a recommended practice.
Piecewise Linear (PWL) source not
See Sections 2.3.9 and 6.1.12 of the Xyce Reference Guide for
fully compatible with either
more details.
HSPICE or PSpice.
The “atto” prefix, which is designated by “a”, is acceptable in
Acceptable prefixes in the metric
HSPICE, but is not accepted in Xyce. The use of the “atto”
system.
prefix in Xyce must be replaced with “E-18”.
In Xyce hierarchical parameters, M (multiplier or multiplicity
factor) and S (scale), are not commonly supported. The M
Hierarchical parameters. parameter is only supported by the R, L, C and MOSFET
device models and some BJT device models (VBIC 1.3 and
MEXTRAM).
.MEASURE has some
See Section 2.1.18.14 of the Xyce Reference Guide for more
incompatibilities and differences
details.
with HSPICE.
The P() accessor for power may
give different results for
semiconductor devices (D, J, M, Q See Sections 2.3.8, 2.3.18, 2.3.20, 2.3.17 2.3.25 and 2.3.19 for
and Z devices) and the lossless more details.
transmission device (T device) than
with HSPICE.
The Xyce .OP statement provides
See Section 2.1.24 for more details.
less output than other simulators.
In SPICE3F5 and PSpice, initial conditions can be set on the
initial voltages and currents at each end of the lossless
Initial conditions for lossless and transmission line, but not for the lossy transmission line. In
lossy transmission lines Xyce, initial conditions can be set on the initial voltages and
currents at each end of the lossy transmission line, but not for
the lossless transmission line

786
 Some SPICE-style circuit simulators can use the .PRINT line
to (for example) print out the vds, vgb, vsd, etc. values for a
PMOS transistor (say, M1) using .PRINT TRAN vgs(M1)
vbs(M1) vds(M1). This is not directly supported in Xyce. See
Use of vgs(Mxxx) style syntax on
Section 2.3.17 for how this is supported with the N() syntax for
the .PRINT line
the BSIM3 and BSIM4 models. For other transistor devices,
use something like this on the Xyce .PRINT line, V(ng,ns)
where ng and ns are the names of the circuits nodes attached to
the gate and source terminals of the transistor.
Devices that may be expected to work in AC or HB analysis do
not at this time. For AC analysis this includes, but is not limited
to, the lossy transmission line (LTRA) and lossless
transmission line (TRA). The LTRA and TRA models will
need to be replaced with lumped transmission line models
(YTRANSLINE) to perform small-signal AC analysis. For
Some devices do not work in
harmonic balance, the two transmission line models do work
frequency-domain analysis
correctly in frequency domain.
Independent behavioral sources, such as a time-dependent B,
E, F, G, or H source, will not work correctly with either AC or
HB. However, such sources which are purely dependent (only
depending on solution variables and not time) will work in AC
and HB.

6.3. DC Operating Point Calculation Failures in Xyce

This section discusses various netlist problems that can cause Xyce to fail to get a DC Operating Point
(DCOP). Some of this discussion is “tutorial” in nature, but helps illustrate the issues.

6.3.1. Incompatible Voltage Constraints at Circuit Nodes

The Xyce DCOP calculation will fail if the netlist specifies incompatible voltage constraints at a given node
in the circuit. This netlist fragment will cause Xyce to fail to get a DCOP because the two voltage sources
obviously cannot both apply their assigned voltage at Node1.

VA Node1 0 1
VB Node1 0 2

This configuration is also not allowed because there is an infinite number of ways that the two voltage
sources can supply current to the rest of the circuit and still maintain the requested voltage.

VA Node1 0 1
VB Node1 0 1

With those two netlist fragments as background, the next two examples illustrate a “Xyce-unique” way that
DCOP failure can occur. This happens because initial conditions on capacitors in Xyce are enforced with
additional voltage sources during the DCOP. So, these two netlist fragments are identical to the two cases
given above, and will both cause a DCOP failure in Xyce. A similar problem can occur with other Xyce
devices that allow initial conditions, for voltage drops across the device, to be set.

787
 VA node1 0 1
CB node1 0 1.0pf IC=2

or

VA node1 0 1
CB node1 0 1.0pf IC=1

6.3.2. Multiple Voltage Constraints From Subcircuits or at Global Nodes

Similar incompatible voltage constraints can be caused by subcircuit definitions, if the subcircuits enforce
voltage constraints on one (or more) of their interface nodes. An example netlist fragment is given below. In
this example, subcircuits X1 and X2 are trying to enforce incompatible constraints at Node1 in the top-level
circuit. This is notionally identical to the first example in the previous subsection. However, these
incompatibilities can be harder to find if the subcircuit definitions are located in different library files.

X1 node1 0 MySubcircuitA
X2 node1 0 MySubcircuitB
.SUBCKT MYSUBCIRCUITA 1 2
VA 1 0 1
R1A 1 internalNodeA 0.5
R2A internalNodeA 2 0.5
.ENDS
.SUBCKT MYSUBCIRCUITB 3 4
VB 3 0 2
R1B 3 internalNodeB 0.5
R2B internalNodeB 4 0.5
.ENDS

Global nodes that have voltage sources applied to them from separate parts of the circuit (e.g, from within
subcircuit definitions) can cause yet another version of the DCOP failure modes given in the previous
subsection. If these two netlist statements are given in different subcircuit definitions then a Xyce DCOP
failure will occur.

Vpin1 $G_GlobalNode1 0 1
Vpin2 $G_GlobalNode1 0 2

Of course, the examples given above can occur in varied combinations.

6.3.3. NODESET and IC Statements in Subcircuits

As previously noted, Xyce does not support .NODESET and .IC statements in subcircuits. This is a common
cause of DCOP failure in Xyce when the same netlist converges in PSpice. See sections 2.1.22 and 2.1.14
for more details on how to move those .NODESET and .IC statements to the “top-level” in the Xyce netlist.

788
6.3.4. No DC Path to Ground for a Current Flow

A Xyce DCOP failure can occur if there is no DC path to ground at a node but a current flow must occur.
This can happen because of a typographic error during netlist entry. An simple example is as follows, where
the netlist line for R1 has 0 (“oh”) rather then 0 (“zero”). It can also happen when all of the current into a
subcircuit must flow through capacitors.
I1 1 0 1
R1 1 O 1
C1 1 0 2pF

6.3.5. Inductor Loops

An inductor loop with no DC path to ground will also typically cause a DCOP failure. A simple example
is:
V1 1 0 1
R1 1 2 1
L1 2 3 2uH
L2 2 3 2uH
R3 3 0 1

6.3.6. Infinite Slope Transistions

It is possible for a user to specify expressions that could have infinite-slope transitions with B-, E-, F-, G-
and H-sources. A common example is IF statements within those source definitions. This can often lead to
“timestep too small” errors when Xyce reaches the transition point. In some cases, it can also cause DCOP
failures. See Section 2.3.16 and the “Analog Behavioral Modeling” (ABM) chapter of the Xyce Users’
Guide [1] for guidance on using the B-source device and ABM expressions. Those recommendations also
apply to the E-, F-, G- and H-sources.

6.3.7. Simulation Settings

Automatic source stepping was added to Xyce in version 6.3. Xyce also automatically does Gmin stepping
when the DCOP calculation fails to converge. In addition, the time integration options normally do not
affect the DCOP calculation. So, adjusting the simulation settings for Xyce typically has no effect on the
DCOP calculation. However, if both of the automatic homotopy methods mentioned above do not work, and
none of the other netlist issues mentioned above exist, then Xyce does have other homotopy methods
available. See the Xyce Users’ Guide [1] for more details.

789
This page intentionally left blank.

790
7. QUICK REFERENCE FOR MICROSOFT WINDOWS USERS

Xyce is supported on Microsoft Windows. However, the primary targets for Xyce are high-performance
supercomputers and workstations, which are almost always running a variant of Unix. All of Xyce
developement is done on Unix platforms. Bearing this in mind, there are occasionally issues with using a
Unix application on a Windows platform. Some of these issues are described in the table below.

Table 7-1. Issues for Microsoft Windows.

Issue Comment
Xyce will expect library files, which are referenced in the netlist, to have
File names are case-sensitive exactly the same case as the actual filename. If not, Xyce will be unable
to find the library file.
The characters that mark the end of a line in Windows are a carriage
return followed by a Line Feed (CR+LF). In Unix-like systems (including
Windows endline characters are
Linux and OS X), the character is simply a Line Feed (LF). Moving a
different from other OS’s
file between the two systems does not usually cause issues, but users
should be aware of the difference in case problems arise.
Programs such as Microsoft Word by default use file formats that Xyce
cannot recognize. It is best not to use such programs to create netlists,
Xyce is unable to read proprietary
unless netlists are saved as *.txt files. If you must use a Microsoft editor,
file formats.
it is better to use Microsoft Notepad. In general, the best solution is to
use a Unix-style editor, such as Vi, Gvim, or Emacs.

791
This page intentionally left blank.

792
8. RAWFILE FORMAT

The rawfile format produced by Xyce closely follows SPICE3 conventions. Differences are noted in
section 8.3. Details on the both the ASCII and binary formats are provided here for reference.

8.1. ASCII Format

The ASCII file format can be created using the -a flag on the command line. See Chapter 3 for more
information.

The ASCII format standard dictates that the file consist of lines or sets of lines introduced by a keyword.
The Title and Date lines should be the first in the file, and should occur only once. They are followed by
the Plotname, Flags, No. Variables, and No. Points lines for each plot.

Listed next are sets of Variables, and Values lines. Let numvars be the number of variables (as specified
in the No. Variables line), and numpts be the number of points (as shown on the No. Points line).
After the Variables keyword there must be numvars declarations of outputs, and after the Values
keyword, there must be numpts lines, each consisting of numvars values.

Finally, Xyce also allows for a Version line to be placed after the No. Points line for compatibility with
various software programs.

See Table 8-1 for a summary of the above.

Table 8-1. Xyce ASCII rawfile format.

Issue Comment
An arbitrary string describing the circuit.
Title:
Date: A free-format date string.
Plotname: A string describing the analysis type.
Flags: A string describing the data type (real or complex).
No. Variables: The number of variables.
No. Points: The number of points.
The version of Xyce used to generate this output. By default the version is not
Version: (optional) output in the header. It can be output with the .options output
outputversioninrawfile=true option.
A newline followed by multiple lines, one for each variable, of the form [tab]
Variables:
<index> [tab] <name> [tab] <type>.
A newline followed by multiple lines, for each point and variable, of the form
Values: [tab] <value> with an integer index preceeding each set of points. Complex
values are output as [tab] <real component>, <imaginary component> .

793
8.2. Binary Format

The binary format is similar to the ASCII format, except that strings are null terminated rather than newline
terminated. In addition, all the values lines are stored in a binary format. The binary storage of real values
as double precision floats is architecture specific.
See Table 8-2 for a summary of the binary table format.

Table 8-2. Xyce binary rawfile format.

Issue Comment

An arbitrary string describing the circuit.

Title:
Date: A free-format date string.
Plotname: A string describing the analysis type.
Flags: A string describing the data type (real or complex).
No. Variables: The number of variables.
No. Points: The number of points.
The version of Xyce used to generate this output. By default the version is not
Version: (optional) output in the header. It can be output with the .options output
outputversioninrawfile=true option.
A newline followed by multiple lines, one for each variable, of the form [tab]
Variables:
<index> [tab] <name> [tab] <type>.
Each real data point is stored contiguously in sizeof(double) byte blocks.
Binary: Complex values are output as real and imaginary components in a block of size
2*sizeof(double) byte blocks.

8.3. Special Notes

• Complex data points are only output under AC analysis.

• Commands and Options lines are not used.
• Binary header is formatted ASCII.
• Xyce can output an optional Version line in the header.

794
BIBLIOGRAPHY

[1] Eric R. Keiter, Richard L. Schiek, Heidi K. Thornquist, Ting Mei, Jason C. Verley, Karthik V.
Aadithya, Joshua D. Schickling, and Gary L. Hennigan. Xyce Parallel Electronic Simulator: Users’
Guide, Version 7.8. Technical Report SAND2023-13274, Sandia National Laboratories,
Albuquerque, NM, 2023.
[2] Orcad PSpice User’s Guide. Technical report, Orcad, Inc., 1998.
[3] Armin W. Doerry. Catalog of window taper functions for sidelobe control. Technical Report
SAND2017-4042, Sandia National Laboratories, Albuquerque, NM, 2017.
[4] Alan V. Oppenheim and Ronald W. Schafer. Digital Signal Processing. Prentice Hall, 1975.
[5] HSPICE User’s Guide. Technical report, Synopsys, Mountain View, California, 2008.
[6] PSPICE A/D Reference Guide. Technical report, Cadence, San Jose, California, 2008.
[7] Touchstone File Format Specification, Version 2.0. Technical report, Open IBIS Forum, 2009.
[8] Hiroshi Akima. A new method of interpolation and smooth curve fitting based on local procedures. J.
ACM, 17(4):589–602, October 1970.
[9] Gisela Engeln-Müllges and Frank Uhlig. Akima and Renner Subsplines, pages 341–352. Springer
Berlin Heidelberg, Berlin, Heidelberg, 1996.
[10] Jean-Paul Berrut and Lloyd N. Trefethen. Barycentric lagrange interpolation. SIAM Rev, 46:501–517.
[11] Arthur P. Mattuck. Introduction to Analysis. Prentice Hall, Inc., 1999.
[12] H. K. Gummel and H. C. Poon. An integral charge control model of bipolar transistors. Bell Sys.
Techn. J., May-June:827–852, 1970.
[13] Matthias Rudolph. Documentation of the fbh hbt model, 2005.
[14] The SiGe group at Auburn University. Mextram web site, 2015.
[15] T. A. Fjeldly, T. Ytterdal, and M. Shur. Introduction to Device Modeling and Circuit Simulation.
Wiley InterScience, 1998.
[16] A. S. Grove. Physics and Technology of Semiconductor Devices. John Wiley and Sons, Inc., 1967.
[17] P. Antognetti and G. Massobrio. Semiconductor Device Modeling with SPICE. McGraw-Hill, 1988.
[18] J.H. Huang, Z.H. Liu, M.C. Jeng, K. Hui, M. Chan, P.K. KO, and C. Hu. BSIM3 Manual. Technical
report, Department of Electrical Engineering and Computer Science, University of California,
Berkeley, CA 94720.
[19] M. Bucher, C. Lallement, C. Enz, F. Theodoloz, and F. Krummenacher. The EPFL/EKV MOSFET
Model Equations for Simulation Technical Report: Model Version 2.6. Technical report, Electonics
Laboratories, Swiss Federal Institute of Technology (EPFL), Lausanne, Switzerland, September 1997.

795
[20] H. Shichman and D. A. Hodges. Modeling and simulation of insulated-gate field-effect transistor
switching circuits. IEEE Journal of Solid-State Circuits, SC-3:285, September 1968.
[21] A. Vladimirescu and S. Lui. The Simulation of MOS Integrated Circuits using SPICE2. Technical
Report Memorandum No. M80/7, February 1980.
[22] B. J. Sheu, D. L. Scharfetter, P.-K. Ko, and M.-C. Jeng. BSIM: Berkeley Short-channel IGFET Model
for MOS Transistors. IEEE Journal of Solid-State Circuits, SC-22:558–566, August 1987.
[23] J. R. Pierret. A MOS Parameter Extraction Program for the BSIM Model. Technical Report
Memorandum No. M84/99 and M84/100, November 1984.
[24] Ping Yang, Berton Epler, and Pallab K. Chatterjee. An investigation of the charge conservation
problem for MOSFET circuit simulation. IEEE Journal of Solid-State Circuits, SC-18(1), February
1983.
[25] BSIM3v3.1 Manual. Technical report, Department of Electrical Engineering and Computer Science,
University of California, Berkeley, CA 94720.
[26] J. C. Bowers and H. A. Neinhaus. SPICE2 Computer Models for HEXFETs. Technical Report
Application Note 954A, International Rectifier Corporation, Berkeley, CA 94720.
[27] P. Su, H. Wan, S. Fung, M. Chan, A. Niknejad, and C. Hu. BSIM3SOI3.1 MOSFET MODEL Users’
Manual. Technical report, Department of Electrical Engineering and Computer Science, University of
California, Berkeley, CA 94720.
[28] P. Su, H. Wan, S. Fung, M. Chan, A. Niknejad, and C. Hu. BSIM3SOI3.2 documentation. Technical
report, Department of Electrical Engineering and Computer Science, University of California,
Berkeley, CA 94720.
[29] G. Gildenblat, X. Li, W.Wu, H. Wang, A. Jha, R. van Langevelde, G.D.J. Smit, A.J. Scholten, and
D.B.M. Klaassen. PSP: An advanced surface-potential-based MOSFET model for circuit simulation.
IEEE Transactions on Electron Devices, 53(9):1979–1993, 2006.
[30] V. Sriramkumar, N. Paydavosi, J. Duarte, D. Lu, C. Lin, M. Dunga, S. Yao, T. Morshed, A. Niknejad,
and C. Hu. BSIM-CMG 107.0.0 Multi-Gate MOSFET Compact Model Technical Manual. Technical
report, Department of Electrical Engineering and Computer Science, University of California,
Berkeley, CA 94720.
[31] M. Bucher, A. Bazigos, E. Kitonaki, and F. Krummenacher. Recent advances in the EKV3 MOS
transistor model. In NANOTECH 2006, Workshop on Compact Models, May 2006.
[32] M. Bucher, D. Diamantakos, A. Bazigos, and F. Krummenacher. Design-oriented characterization and
parameter extraction methodologies for the EKV3 MOSFET model. In NANOTECH 2007 Workshop
on Compact Modeling, May 2007.
[33] J.S. Roychowdhury, A.R. Newton, and D.O. Pederson. Algorithms for the transient simulation of lossy
interconnect. IEEE Trans. Computer-Aided Design of Integrated Circuits and Systems, 13(1), 1994.
[34] Tom Quarles. Spice3f5 Users’ Guide. Technical report, University of California-Berkeley, Berkeley,
California, 1994.
[35] Abnita Munankarmay and Michael A. Heroux. A comparison of two equivalent real formulations for
complex-valued linear systems part 1: Introduction and method. American Journal of Undergraduate
Research, 1(3), 2002.

796
[36] Federico Milano. Power System Modelling and Scripting. Springer, 2010.
[37] Prabha Kundur. Power System Stability and Control. McGraw Hill, 1994.
[38] Shahar Kvatinsky, Keren Talisveyberg, Dmitry Fliterand, Avinoam Kolodny, Uri C. Weiser, and
Eby G. Friedman. Models of memristors for spice simulations. IEEE 27th Convention of Electrical
and Electronics Engineers in Israel, pages 1–5, 2012.
[39] Shahar Kvatinsky, Eby G. Friedman, Avinoam Kolodny, and Uri C. Weiser. Team: Threshold adaptive
memristor model. IEEE Transactions On Circuits And Systems—I, 60:211–221, January 2013.
[40] Shahar Kvatinsky, Eby G. Friedman, Avinoam Kolodny, and Uri C. Weiser. Generalized memristive
device spice model and its application in circuit design. IEEE Transactions on Computer-Aided
Design of Integrated Circuits and Systems, 32, August 2013.
[41] Niroula J., Agarwal S., Jacobs-Gedrim R., Schiek R., Hughart D., Hsia A., James C., and Marinella
M. Compact model of resistive memory for pulsed analog and neuromorphic applications.
Subbmitted to Journal of Computational Electronics, 2017.
[42] C. D. Thurmond. The standard thermodynamic function of the formation of electrons and holes in Ge,
Si, GaAs, and GaP. J. Electrochem. Soc., 122:1133, 1975.
[43] D.M. Caughey and R.E. Thomas. Carrier mobilities in silicon empirically related to doping and field.
Proc. IEEE, 55:2192–2193, 1967.
[44] N.D. Arora, J.R. Hauser, and D.J. Roulston. Electron and hole mobilities in silicon as a function of
concentration and temperature. IEEE Transactions on Electron Devices, ED-29:292–295, 1982.
[45] J.M. Dorkel and Ph. Leturq. Carrier mobilities in silicon semi-empirically related to temperature,
doping, and injection level. Solid-State Electronics, 24(9):821–825, 1981.
[46] C. Lombardi, S. Manzini, A. Saporito, and M. Vanzi. A physically based mobility model for
numerical simulation of nonplanar devices. IEEE Transactions on Computer-Aided Design,
7(11):1164–1170, November 1988.
[47] Ben G. Streetman and Sanjay Banerjee. Solid State Electronic Devices. Prentice-Hall, Upper Saddle
River, New Jersey, 2000.
[48] K.E. Brenan, S.L Campbell, and L.R. Petzold. Numerical Solution of Initial-Value Problems in
Differential-Algebraic Equations. Society for Industrial and Applied Mathmatics, Philadelphia, 1996.
[49] Gary J. Templet, Richard L. Schiek, and Jason C. Verley. Xyce xdm netlist translator user guide,
version 2.7. Technical Report SAND2023-07308, Sandia National Laboratories, 2022.

797
APPENDIX A. Third Party Licenses

Xyce makes use of code developed by various third parties. The following text is provided to comply with
the licenses of the codes that require it.
The ksparse solver in Xyce contains code derived from SPICE 3f5 source code:

Copyright (c) 1985-1991 The Regents of the University of California.

All rights reserved.

Permission is hereby granted, without written agreement and without

license or royalty fees, to use, copy, modify, and distribute this
software and its documentation for any purpose, provided that the
above copyright notice and the following two paragraphs appear in all
copies of this software.

IN NO EVENT SHALL THE UNIVERSITY OF CALIFORNIA BE LIABLE TO ANY PARTY

FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN IF
THE UNIVERSITY OF CALIFORNIA HAS BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGE.

THE UNIVERSITY OF CALIFORNIA SPECIFICALLY DISCLAIMS ANY WARRANTIES,

INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE
PROVIDED HEREUNDER IS ON AN "AS IS" BASIS, AND THE UNIVERSITY OF
CALIFORNIA HAS NO OBLIGATION TO PROVIDE MAINTENANCE, SUPPORT, UPDATES,
ENHANCEMENTS, OR MODIFICATIONS.

Xyce’s linear solver makes use of the AMD library:

AMD, Copyright (c), 1996-2015, Timothy A. Davis,

Patrick R. Amestoy, and Iain S. Duff. All Rights Reserved.
Used in Xyce under the LGPL v2.1 license.

Parallel builds of Xyce use the Open MPI library:

Most files in this release are marked with the copyrights of the
organizations who have edited them. The copyrights below are in no
particular order and generally reflect members of the Open MPI core
team who have contributed code to this release. The copyrights for
code used under license from other parties are included in the
corresponding files.

798
Copyright (c) 2004-2010 The Trustees of Indiana University and Indiana
University Research and Technology
Corporation. All rights reserved.
Copyright (c) 2004-2010 The University of Tennessee and The University
of Tennessee Research Foundation. All rights
reserved.
Copyright (c) 2004-2010 High Performance Computing Center Stuttgart,
University of Stuttgart. All rights reserved.
Copyright (c) 2004-2008 The Regents of the University of California.
All rights reserved.
Copyright (c) 2006-2010 Los Alamos National Security, LLC. All rights
reserved.
Copyright (c) 2006-2010 Cisco Systems, Inc. All rights reserved.
Copyright (c) 2006-2010 Voltaire, Inc. All rights reserved.
Copyright (c) 2006-2011 Sandia National Laboratories. All rights
reserved.
Copyright (c) 2006-2010 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
Copyright (c) 2006-2010 The University of Houston. All rights
reserved.
Copyright (c) 2006-2009 Myricom, Inc. All rights reserved.
Copyright (c) 2007-2008 UT-Battelle, LLC. All rights reserved.
Copyright (c) 2007-2010 IBM Corporation. All rights reserved.
Copyright (c) 1998-2005 Forschungszentrum Juelich, Juelich
Supercomputing Centre, Federal Republic of
Germany
Copyright (c) 2005-2008 ZIH, TU Dresden, Federal Republic of Germany
Copyright (c) 2007 Evergrid, Inc. All rights reserved.
Copyright (c) 2008 Chelsio, Inc. All rights reserved.
Copyright (c) 2008-2009 Institut National de Recherche en
Informatique. All rights reserved.
Copyright (c) 2007 Lawrence Livermore National Security, LLC.
All rights reserved.
Copyright (c) 2007-2009 Mellanox Technologies. All rights reserved.
Copyright (c) 2006-2010 QLogic Corporation. All rights reserved.
Copyright (c) 2008-2010 Oak Ridge National Labs. All rights reserved.
Copyright (c) 2006-2010 Oracle and/or its affiliates. All rights
reserved.
Copyright (c) 2009 Bull SAS. All rights reserved.
Copyright (c) 2010 ARM ltd. All rights reserved.
Copyright (c) 2010-2011 Alex Brick <[email protected]>. All rights
reserved.
Copyright (c) 2013-2014 Intel, Inc. All rights reserved.
Copyright (c) 2011-2014 NVIDIA Corporation. All rights reserved.

Additional copyrights may follow

799
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:

- Redistributions of source code must retain the above copyright

notice, this list of conditions and the following disclaimer.

- Redistributions in binary form must reproduce the above copyright

notice, this list of conditions and the following disclaimer listed
in this license in the documentation and/or other materials
provided with the distribution.

- Neither the name of the copyright holders nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.

The copyright holders provide no reassurances that the source code

provided does not infringe any patent, copyright, or any other
intellectual property rights of third parties. The copyright holders
disclaim any liability to any recipient for claims brought against
recipient by any third party for infringement of that parties
intellectual property rights.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS

"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Xyce uses the Trilinos Solver Framework:

Trilinos: An Object-Oriented Solver Framework

Copyright (2001) Sandia Corporation

Copyright (2001) Sandia Corporation. Under the terms of Contract

DE-AC04-94AL85000, there is a non-exclusive license for use of this
work by or on behalf of the U.S. Government. Export of this program
may require a license from the United States Government.

1. Redistributions of source code must retain the above copyright

notice, this list of conditions and the following disclaimer.

800
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.

3. Neither the name of the Corporation nor the names of the

contributors may be used to endorse or promote products derived from
this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY SANDIA CORPORATION "AS IS" AND ANY

EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL SANDIA CORPORATION OR THE
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

NOTICE: The United States Government is granted for itself and others
acting on its behalf a paid-up, nonexclusive, irrevocable worldwide
license in this data to reproduce, prepare derivative works, and
perform publicly and display publicly. Beginning five (5) years from
July 25, 2001, the United States Government is granted for itself and
others acting on its behalf a paid-up, nonexclusive, irrevocable
worldwide license in this data to reproduce, prepare derivative works,
distribute copies to the public, perform publicly and display
publicly, and to permit others to do so.

NEITHER THE UNITED STATES GOVERNMENT, NOR THE UNITED STATES DEPARTMENT
OF ENERGY, NOR SANDIA CORPORATION, NOR ANY OF THEIR EMPLOYEES, MAKES
ANY WARRANTY, EXPRESS OR IMPLIED, OR ASSUMES ANY LEGAL LIABILITY OR
RESPONSIBILITY FOR THE ACCURACY, COMPLETENESS, OR USEFULNESS OF ANY
INFORMATION, APPARATUS, PRODUCT, OR PROCESS DISCLOSED, OR REPRESENTS
THAT ITS USE WOULD NOT INFRINGE PRIVATELY OWNED RIGHTS.

Some versions of Xyce use the Intel Math Kernel Library:

Intel Simplified Software License (Version April 2018)

Copyright (c) 2018 Intel Corporation.

Use and Redistribution. You may use and redistribute the software (the
"Software"), without modification, provided the following conditions
are met:

801
* Redistributions must reproduce the above copyright notice and the
following terms of use in the Software and in the documentation
and/or other materials provided with the distribution.

* Neither the name of Intel nor the names of its suppliers may be used
to endorse or promote products derived from this Software without
specific prior written permission.

* No reverse engineering, decompilation, or disassembly of this

Software is permitted.

Limited patent license. Intel grants you a world-wide, royalty-free,

non-exclusive license under patents it now or hereafter owns or
controls to make, have made, use, import, offer to sell and sell
("Utilize") this Software, but solely to the extent that any such
patent is necessary to Utilize the Software alone. The patent license
shall not apply to any combinations which include this software. No
hardware per se is licensed hereunder.

Third party and other Intel programs. "Third Party Programs" are the
files listed in the "third-party-programs.txt" text file that is
included with the Software and may include Intel programs under
separate license terms. Third Party Programs, even if included with
the distribution of the Materials, are governed by separate license
terms and those license terms solely govern your use of those
programs.

DISCLAIMER. THIS SOFTWARE IS PROVIDED "AS IS" AND ANY EXPRESS OR

IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, AND
NON-INFRINGEMENT ARE DISCLAIMED. THIS SOFTWARE IS NOT INTENDED FOR USE
IN SYSTEMS OR APPLICATIONS WHERE FAILURE OF THE SOFTWARE MAY CAUSE
PERSONAL INJURY OR DEATH AND YOU AGREE THAT YOU ARE FULLY RESPONSIBLE
FOR ANY CLAIMS, COSTS, DAMAGES, EXPENSES, AND ATTORNEYS’ FEES ARISING
OUT OF ANY SUCH USE, EVEN IF ANY CLAIM ALLEGES THAT INTEL WAS
NEGLIGENT REGARDING THE DESIGN OR MANUFACTURE OF THE MATERIALS.

LIMITATION OF LIABILITY. IN NO EVENT WILL INTEL BE LIABLE FOR ANY

DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. YOU AGREE TO INDEMNIFY AND
HOLD INTEL HARMLESS AGAINST ANY CLAIMS AND EXPENSES RESULTING FROM
YOUR USE OR UNAUTHORIZED USE OF THE SOFTWARE.

802
No support. Intel may make changes to the Software, at any time
without notice, and is not obligated to support, update or provide
training for the Software.

Termination. Intel may terminate your right to use the Software in the
event of your breach of this Agreement and you fail to cure the breach
within a reasonable period of time.

Feedback. Should you provide Intel with comments, modifications,

corrections, enhancements or other input ("Feedback") related to the
Software Intel will be free to use, disclose, reproduce, license or
otherwise distribute or exploit the Feedback in its sole discretion
without any obligations or restrictions of any kind, including without
limitation, intellectual property rights or licensing obligations.

Compliance with laws. You agree to comply with all relevant laws and
regulations governing your use, transfer, import or export (or
prohibition thereof) of the Software.

Governing law. All disputes will be governed by the laws of the United
States of America and the State of Delaware without reference to
conflict of law principles and subject to the exclusive jurisdiction
of the state or federal courts sitting in the State of Delaware, and
each party agrees that it submits to the personal jurisdiction and
venue of those courts and waives any objections. The United Nations
Convention on Contracts for the International Sale of Goods (1980) is
specifically excluded and will not apply to the Software.

Xyce’s implementation of the Diode-CMC model, version 2.0.0, is derived from Verilog-A sources
provided under the following license:

Silicon Integration Initiative (Si2)

Compact Model Coalition In-Code Statement

Software is distributed as is, completely without warranty or service

support. NXP Semiconductors, Hiroshima University, and Silicon
Integration Initiative, Inc. (Si2), along with their employees are not
liable for the condition or performance of the software.

NXP Semiconductors, Hiroshima University, and Si2 own the copyright

and grant users a perpetual, irrevocable, worldwide, non-exclusive,
royalty-free license with respect to the software as set forth below.

NXP Semiconductors, Hiroshima University, and Si2 hereby disclaim all

implied warranties.

NXP Semiconductors, Hiroshima University, and Si2 grant the users the
right to modify, copy, and redistribute the software and

803
documentation, both within the user’s organization and externally,
subject to the following restrictions

1. The users agree not to charge for the NXP Semiconductors,

Hiroshima University, and Si2 -developed code itself but may charge
for additions, extensions, or support.

2. In any product based on the software, the users agree to

acknowledge NXP Semiconductors, Hiroshima University, and Si2 that
developed the software. This acknowledgment shall appear in the
product documentation.

3. Redistributions to others of source code and documentation must

retain the copyright notice, disclaimer, and list of conditions.

4. Redistributions to others in binary form must reproduce the

copyright notice, disclaimer, and list of conditions in the
documentation and/or other materials provided with the distribution.

Xyce’s implementations of the L-UTSOI and EKV 2.6 models are derived from Verilog-A sources provided
under the ECL-2.0 license. The L-UTSOI is Copyright 2020 CEA-Leti.
The ECL-2.0 license text is as follows:

Licensed under Educational Community License, Version 2.0 (the

"License"); you may not use this file except in compliance with the
License. You may obtain a copy of the license at

http://opensource.org/licenses/ECL-2.0

Unless required by applicable law or agreed to in writing, software

distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied. See the License for the specific language governing
permissions and limitations under the License.

Xyce’s implementation of the MEXTRAM model, version 504.12.1, is derived from Verilog-A sources
provided under the following license:

Copyright (c) 2000-2007, NXP Semiconductors

Copyright (c) 2007-2014, Delft University of Technology
Copyright (c) 2015, Auburn University

INTELLECTUAL PROPERTY NOTICE, DISCLAIMER AND LICENSE

The Mextram model and documentation presented at this website, denoted

as the Model, has been developed by NXP Semiconductors until 2007,
Delft University of Technology from 2007 to 2014, and Auburn

804
University since April 2015.

The Model is distributed as is, completely without any expressed or

implied warranty or service support. NXP Semiconductors, Delft
University of Technology, Auburn University and their employees are
not liable for the condition or performance of the Model.

NXP Semiconductors, Delft University of Technology, Auburn University

own the copyright and grant users a perpetual, irrevocable, worldwide,
non-exclusive, royalty-free license with respect to the Model as set
forth below.

NXP Semiconductors, Delft University of Technology, Auburn University

hereby disclaim all implied warranties.

NXP Semiconductors, Delft University of Technology, Auburn University

grant the users the right to modify, copy, and redistribute the Model
and documentation, both within the user’s organization and externally,
subject to the following restrictions

1. The users agree not to charge for the code itself but may charge
for additions, extensions, or support.

2. In any product based on the Model, the users agree to acknowledge

NXP Semiconductors, Delft University of Technology, Auburn
University that developed the Model. This acknowledgment shall
appear in the product documentation.

3. The users agree to obey all restrictions governing redistribution

or export of the Model.

4. The users agree to reproduce any copyright notice which appears on

the Model on any copy or modification of such made available to
others.

Xyce’s implementation of the HICUM/L0 model, version 1.32, is derived from Verilog-A sources provided
under the following license:

Software is distributed as is, completely without warranty or service

support. Michael Schroter and his team members are not liable for the
condition or performance of the software.

Michael Schroter owns the copyright and grants users a perpetual,

irrevocable, worldwide, non-exclusive, royalty-free license with
respect to the software as set forth below.

Michael Schroter hereby disclaims all implied warranties.

805
Michael Schroter grants the users the right to modify, copy, and
redistribute the software and documentation, both within the user’s
organization and externally, subject to the following restrictions.

1. The users agree not to charge for the model owner’s code itself but
may charge for additions, extensions, or support.

2. In any product based on the software, the users agree to

acknowledge Michael Schroter who developed the model and software.
This acknowledgment shall appear in the product documentation.

3. Redistributions to others of source code and documentation must

retain the copyright notice, disclaimer, and list of conditions.

4. Redistributions to others in binary form must reproduce the

copyright notice, disclaimer, and list of conditions in the
documentation and/or other materials provided with the distribution

Xyce’s implementation of the HICUM/L2 model, version 2.34, is derived from Verilog-A sources provided
under the following license:

The Software is distributed as is, completely without expressed or

implied warranty or service support. Michael Schroter and his team
members are not liable for the condition or performance of the
software.

Michael Schroter owns the copyright and grants users a perpetual,

irrevocable, worldwide, non-exclusive, royalty-free license with
respect to the software as set forth below.

Michael Schroter hereby disclaims all implied warranties.

Michael Schroter grants the users the right to modify, copy, and
redistribute the software and documentation, both within the user’s
organization and externally, subject to the following restrictions.

1. The users agree not to charge for the model owner code itself but
may charge for additions, extensions, or support.

2. In any product based on the software, the users agree to

acknowledge Michael Schroter that developed the model and software.
This acknowledgment shall appear in the product documentation.

3. Users agree to obey all government restrictions governing

redistribution or export of the software.

4. The Users agree to reproduce any copyright notice which appears on

the software and documentation on any copy or modification of such

806
 made available to others.

Xyce’s implementations of the BSIM3 v.3.2.2, the BSIM4 v. 4.6.1, the BSIM4 v. 4.7.0, and the BSIM-SOI
v. 3.2, are based on the original code of those devices provided by University of California, Berkeley. They
all have the following license:

Software is distributed as is, completely without warranty or service

support. The University of California and its employees are not liable
for the condition or performance of the software.

The University owns the copyright but shall not be liable for any
infringement of copyright or other proprietary rights brought by third
parties against the users of the software.

The University of California hereby disclaims all implied warranties.

The University of California grants the users the right to modify,

copy, and redistribute the software and documentation, both within the
user’s organization and externally, subject to the following
restrictions

1. The users agree not to charge for the University of California code
itself but may charge for additions, extensions, or support.

2. In any product based on the software, the users agree to

acknowledge the UC Berkeley BSIM Research Group that developed the
software. This acknowledgment shall appear in the product
documentation.

3. The users agree to obey all U.S. Government restrictions governing

redistribution or export of the software.

4. The users agree to reproduce any copyright notice which appears on

the software on any copy or modification of such made available
to others.

Version 4.8.2 of the BSIM4 is licensed under the Educational Community License, Version 2.0:

Copyright 2020 University of California

Licensed under Educational Community License, Version 2.0 (the

"License"); you may not use this file except in compliance with the
License. You may obtain a copy of the license at

http://opensource.org/licenses/ECL-2.0

Unless required by applicable law or agreed to in writing, software

807
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied. See the License for the specific language governing
permissions and limitations under the License.

Xyce’s implementation of the BSIM6 model, version 6.1.1, is derived from Verilog-A sources provided
under the following license:

Software is distributed as is, completely without warranty or service

support. The University of California and its employees are not liable
for the condition or performance of the software.

The University owns the copyright but shall not be liable for any
infringement of copyright or other proprietary rights brought by third
parties against the users of the software.

The University of California hereby disclaims all implied warranties.

The University of California grants the users the right to modify,

copy, and redistribute the software and documentation, both within the
user’s organization and externally, subject to the following
restrictions

1. The users agree not to charge for the University of California code
itself but may charge for additions, extensions, or support.

2. In any product based on the software, the users agree to

acknowledge the UC Berkeley BSIM Research Group that developed the
software. This acknowledgment shall appear in the product
documentation.

3. The users agree to obey all U.S. Government restrictions governing

redistribution or export of the software.

4. The users agree to reproduce any copyright notice which appears on

the software on any copy or modification of such made available
to others.

Xyce’s implementations of the BSIM-SOI model, versions 4.5 and 4.6.1, are derived from Verilog-A
sources provided under the following license:

The Regents of the University of California and Analog Devices, Inc.

("Authors") own the copyright but shall not be liable for any
infringement of copyright or other proprietary rights brought by third
parties against the users of the software.

The Authors jointly grant the users the right to modify, copy, and

808
redistribute the software and documentation, both within the user’s
organization and externally, subject to the following conditions:

1. The users agree not to charge for the original source code itself
but may charge for additions, extensions, or support.

2. In any product based on the software, the users agree to acknowledge

the UC Berkeley BSIM Research Group that developed the software. This
acknowledgment shall appear in the product documentation.

3. The users agree to obey all U.S. Government restrictions governing

redistribution or export of the software.

4. Redistributions of source code must retain the above copyright

notices, this list of conditions and the following disclaimer.

5. Redistributions in binary form must reproduce the above copyright

notices, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.

6. Users may not claim authorship of the original software. Modified

versions of the software must be plainly marked as such.

7. Users may not suggest any sponsorship or endorsement by the Authors

of users’ product.

THIS SOFTWARE IS PROVIDED BY AUTHORS AS IS AND ANY EXPRESS OR IMPLIED

WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, AND
NONINFRINGEMENT OF THIRD PARTY RIGHTS ARE DISCLAIMED. IN NO EVENT SHALL
THE AUTHORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Xyce’s implementations of the BSIM-CMG model, versions 107.0.0 and 110.0.0, are derived from
Verilog-A sources provided under the following license:

Software is distributed as is, completely without warranty or service

support. The University of California and its employees are not liable
for the condition or performance of the software.

The University owns the copyright but shall not be liable for any
infringement of copyright or other proprietary rights brought by third

809
parties against the users of the software.

The University of California hereby disclaims all implied warranties.

The University of California grants the users the right to modify,

copy, and redistribute the software and documentation, both within the
user’s organization and externally, subject to the following
restrictions

1. The users agree not to charge for the University of California code
itself but may charge for additions, extensions, or support.

2. In any product based on the software, the users agree to

acknowledge the UC Berkeley BSIM Research Group that developed the
software. This acknowledgment shall appear in the product
documentation.

3. The users agree to obey all U.S. Government restrictions governing

redistribution or export of the software.

4. The users agree to reproduce any copyright notice which appears on

the software on any copy or modification of such made available
to others.

Xyce’s implementation of the MVS model, version 2.0.0, is derived from Verilog-A sources provided under
the following license:

Copyright @ 2013 Massachusetts Institute of Technology (MIT)

The terms under which the software and associated documentation (the
Software) is provided are as the following:

The Software is provided "as is", without warranty of any kind,

express or implied, including but not limited to the warranties of
merchantability, fitness for a particular purpose and noninfringement.
In no event shall the authors or copyright holders be liable for any
claim, damages or other liability, whether in an action of contract,
tort or otherwise, arising from, out of or in connection with the
Software or the use or other dealings in the Software.

MIT grants, free of charge, to any users the right to modify, copy,
and redistribute the Software, both within the user’s organization and
externally, subject to the following restrictions:

1. The users agree not to charge for the MIT code itself but may
charge for additions, extensions, or support.

2. In any product based on the Software, the users agree to

810
 acknowledge the MIT VS Model Research Group that developed the
software. This acknowledgment shall appear in the product
documentation.

3. The users agree to obey all U.S. Government restrictions governing

redistribution or export of the software.

4. The users agree to reproduce any copyright notice which appears on

the software on any copy or modification of such made available to
others.

Xyce’s implementation of the PSP model, version 102.5.0, is derived from Verilog-A sources provided
under the following license:

INTELLECTUAL PROPERTY NOTICE, DISCLAIMER AND LICENSE

The compact model software and documentation presented at this website

form a whole that will henceforth be denoted as the "Model".

The Model presented at this website has been co-developed by NXP

Semiconductors and Arizona State University until and including 2011.
For this part of the Model, NXP Semiconductors claims undivided
ownership and copyrights.

Since 2012 until today the Model has been co-developed by

NXP Semiconductors and Delft University of Technology and
for this part each claim undivided ownership and copyrights.

DISCLAIMER

The owners are fully free to further develop, adapt and extend the
Model as they judge necessary or desirable.

The Model is distributed as is, completely without any express or

implied warranty, or service support. The owners and their employees
are not liable in any way for the condition or performance of the
Model. The owners hereby disclaim all implied warranties.

LICENSE

NXP Semiconductors and Delft University of Technology hereby grant

users a perpetual, irrevocable, worldwide, non-exclusive, royalty-free
license with respect to Versions of the Model which have been released
through this website http://psp.ewi.tudelft.nl.

811
NXP Semiconductors and Delft University of Technology grant the users
the right to modify, copy and redistribute the Model, both within the
user’s organization and externally, subject to the following
restrictions.

RESTRICTIONS

1. The users agree not to charge for the Model itself but may
charge for additions, extensions, or support.

2. In any product based on the Model, the users agree to acknowledge

the owners as developers of the Model. This acknowledgment shall
appear in the product documentation.

3. The users agree to only use the name of CMC standard models to
identify implementations of the CMC standard models which produce
the same outputs as Standard code for the same inputs passing all
CMC QA tests.

4. The users agree to obey all government restrictions governing

redistribution or export of the software.

5. The users agree to reproduce any copyright notice which appears on

the software and documentation on any copy or modification of such
made available to others.

Xyce’s implementations of the PSP model, version 103.4.0, and the JUNCAP diode, are derived from
Verilog-A sources provided under the following license:

Software is distributed as is, completely without warranty or service

support. The Commissariat a l’energie atomique et aux energies
alternatives (CEA), NXP Semiconductors, and Delft University of
Technology, along with their employees are not liable for the
condition or performance of the software.

NXP Semiconductors, Delft University of Technology, and CEA own the

copyright and grant users a perpetual, irrevocable, worldwide,
non-exclusive, royalty-free license with respect to the software as
set forth below.

NXP Semiconductors, Delft University of Technology, and CEA hereby

disclaim all implied warranties.

NXP Semiconductors, Delft University of Technology, and CEA grant the

users the right to modify, copy, and redistribute the software and
documentation, both within the user’s organization and externally,
subject to the following restrictions:

812
 1. The users agree not to charge for the NXP Semiconductors, Delft
University of Technology, and CEA-developed code itself but may
charge for additions, extensions, or support.

2. In any product based on the software, the users agree to

acknowledge NXP Semiconductors, Delft University of Technology,
and CEA that developed the software. This acknowledgment shall
appear in the product documentation.

3. Redistributions to others of source code and documentation must

retain the copyright notice, disclaimer, and list of conditions.

4. Redistributions to others in binary form must reproduce the

copyright notice, disclaimer, and list of conditions in the
documentation and/or other materials provided with the
distribution.

Xyce’s implementation of the EKV 3.0 model is derived from Verilog-A sources developed by the EKV
Team of the Electronics Laboratory-TUC (Technical University of Crete). They are included in Xyce under
license from Technical University of Crete. The official web site of the EKV model is
http://ekv.epfl.ch/.
Due to licensing restrictions, the EKV MOSFETs are not available in open-source versions of Xyce.
The license for EKV3 authorizes Sandia National Laboratories to distribute EKV3 only in binary
versions of the code.

813
This page intentionally left blank.

814
Index

Xyce OUTPUT, 108

ABSTOL, 775 PARSER, 117
RELTOL, 775 PCES, 112
convergence, 775 RESTART, 109
.AC, 22 SAMPLES, 111
.DCVOLT, 27, 46 SENSITIVITY, 113
.EMBEDDEDSAMPLING, 28 TIMEINT, 97
.FFT, 33 .OP, 93
.FOUR, 38 .PARAM, 118
.FUNC, 40 subcircuit scoping, 119, 158
allowed names, 40 .PCE, 120
subcircuit scoping, 40, 158 .PREPROCESS
.GLOBAL_PARAM, 43 ADDRESISTORS, 125
.GLOBAL, 42 REMOVEUNUSED, 124
.HB, 45 REPLACEGROUND, 123
.IC, 46 .PRINT, 23, 25, 45, 92, 126, 160
.INC, 47 AC, 23, 133
.LIB, 48 DC, 25, 135
.LIN, 50 ES, 141
.MEASURE FFT, 82 HB, 45, 135
.MEASURE continuous, 76 NOISE, 92, 137
.MEASURE, 52, 56 PCE, 141
.MODEL, 175 TRAN, 138, 160
subcircuit scoping, 87, 158 AC Analysis, 133
.NODESET, 90 DC Analysis, 135
.NOISE, 91 Device Parameters and Internal Variables,
.OPTIONS, 94, 778, 784 145
DEVICE, 95 EMBEDDEDSAMPLING Analysis, 141
DIAGNOSTIC, 97 Harmonic Balance Analysis, 135
DIST, 114 Homotopy Analysis, 138, 139
EMBEDDEDSAMPLES, 111 Noise Analysis, 137
FFT, 115 Parameter Stepping, 142
HBINT, 114 PCE Analysis, 141
LINSOL-AC, 107 Transient Analysis, 138
LINSOL-HB, 107 Wildcards, 144
LINSOL, 104 .RESULT, 147
LOCA, 103 .SAMPLING, 148
MEASURE, 116 .SAVE, 150
NONLIN-TRAN, 101 .SENS, 151
NONLIN, 101 .STEP, 153

815
.SUBCKT, 157, 175 Decade sweeps, 155
.TRAN, 108, 159 Linear sweeps, 153
ABSTOL, 98, 102 List sweeps, 155
BPENABLE, 100 Octave sweeps, 155
DATA, 24 step, 153
DC, 25 transient, 101, 159, 160
EXITSTEP, 101 Aztec, 104, 107
EXITTIME, 101
MAXSTEP, 102 behavioral digital
NLNEARCONV, 98 device instance parameters, 712, 717
NLSMALLUPDATE, 98 device model parameters, 712, 717
RELTOL, 98, 102 bias point, 160
RESTARTSTEPSCALE, 98 bipolar junction transistor
1d pde (level 1) device instance parameters, 245
device instance parameters, 741 device model parameters, 245
2d pde (level 2) BJT, 239
device instance parameters, 745 BJT level 230
device output variables, 263
AC analysis, 22 BJT level 234
accelerated mass device, 721 device output variables, 269
Accelerated Mass Devices, 176 BJT level 504
algorithm device output variables, 274
time integration, 784 BJT level 505
analog behavioral modeling device output variables, 281
POLY, 172 bsim-cmg finfet v107.0.0
polynomial expression, 172 device instance parameters, 600
analysis device model parameters, 600
AC, 22 bsim-cmg finfet v108.0.0
control parameters, 94 device instance parameters, 629
DC, 25, 101 device model parameters, 630
Data Sweeps, 26, 155 bsim-cmg finfet v110.0.0
Decade sweeps, 25 device instance parameters, 563
Linear sweeps, 25 device model parameters, 564
List Sweeps, 26 bsim-soi 4.5.0
Octave sweeps, 26 device instance parameters, 412
EMBEDDEDSAMPLING, 28 device model parameters, 413
HB, 45 bsim-soi 4.6.1
LHS, 28, 148 device instance parameters, 381
MC, 28, 148 device model parameters, 382
Monte Carlo, 28, 148 bsim3
Noise, 91 device instance parameters, 312
op, 93 device model parameters, 312
options, 94 bsim3 soi
PCE, 28, 120, 148 device instance parameters, 326
pce, 120 device model parameters, 327
SAMPLING, 28, 148 bsim4
sampling, 28, 148 device instance parameters, 347
STEP device model parameters, 349

816
bsim6 digital devices, 176, 711
device instance parameters, 442 digital devices, Y type, 716
device model parameters, 443 diode, 175, 206
bsource, 237 equations, 175
generic switch, 176, 703
capacitor, 185 ideal delay device, 709
device instance parameters, 186 independent current source, 175
device model parameters, 187 independent voltage source, 176, 220, 225
checkpoint, 109 inductor, 175, 190
command line, 769 JFET, 176, 285
arguments, 769 linear device, 176
constants (EXP,PI), 40 lossless transmission line, 705
controlled switch, 701 LTRA, 176
device model parameters, 702 ltra, 699
convergence, 775 lumped transmission line, 707
current controlled current source, 233 memristor, 176
current controlled voltage source, 236 MESFET, 176, 289
current source MOSFET, 176, 291
current controlled, 233 mutual inductor, 175
independent, 220 mutualinductor, 193
nonlinear dependent, 237 nodes, 181
voltage controlled, 234 nonlinear dependent source, 175
package options, 95
DC analysis, 25, 101 PDE Devices, 176, 741
Data sweeps, 26, 155 port device, 176, 230
Decade sweeps, 25 resistor, 176, 200
Linear sweeps, 25 subcircuit, 176
List sweeps, 26 transmission line, 176
Octave sweeps, 26 lossless, 705
delay element lossy, 699
device instance parameters, 710 lumped, 707
device voltage controlled current source, 175, 234
ABM device voltage controlled switch, 176
PSpice equivalent, 779 voltage controlled voltage source, 175, 231
ACC Devices, 176 diagnostic
accelerated mass devices, 176 options, 97
analog, 175 Digital Devices, 176
analog device summary, 175 digital devices, 711
B source, 175 diode, 206
bipolar junction transistor (BJT, 176 device instance parameters, 207
BJT, 239 device model parameters, 207
bsource, 237 operating temperature, 206
capacitor, 175, 185 DIODE level 200
controlled switch, 701 device output variables, 211
current controlled current source, 175, 233 DIODE level 2002
current controlled switch, 176 device output variables, 215
current controlled voltage source, 175, 236 diodecmc 2.0.0
delay device, 709 device instance parameters, 211

817
 device model parameters, 212 jfet
dist device instance parameters, 286, 287
options, 114, 115 device model parameters, 286, 287
dopingprofiles juncap200 diode
composite parameters, 748 device instance parameters, 208
device model parameters, 208
ekv mosfet version 2.6
device instance parameters, 673 Latin Hypercube Sampling analysis, 28, 148
device model parameters, 673 layer
ekv3 mosfet composite parameters, 747
device instance parameters, 676 lead currents, 129, 184
device model parameters, 676 legal characters, 183
embedded sampling analysis, 28 level parameter, 88
EMBEDDEDSAMPLING analysis, 28 LHS analysis, 28, 148
expressions lin
operators, 162 device model parameters, 704
SPICE functions, 169 linear device, 704
lossless transmission line, 705
fbh hbtx v2.1 lossy transmission line, 699
device instance parameters, 257 device model parameters, 699
device model parameters, 257 lumped transmission line, 707
fft, 33 device instance parameters, 707
additional metrics, 36 device model parameters, 707
options, 35, 37, 115, 116 lutsoi mosfet
remeasure, 37 device instance parameters, 684
window functions, 35 device model parameters, 684
generic switch, 703 MC analysis, 28, 148
measure
harmonic balance analysis, 45
err1 measure, 70
hicum l0 v1.32
err2 measure, 70
device instance parameters, 260
error measure, 70
device model parameters, 260
additional examples, 68
hicum v2.4.0
expression support, 64
device model parameters, 265
HSPICE compatibility, 73
ideal transmission line measurement output, 63
device instance parameters, 705 measurement windows, 64
illegal characters, 183 options, 116, 117
independent voltage source, 220, 225 re-measure, 65
inductor, 190 supported operators for AC measures, 71
device instance parameters, 192 supported operators for Noise measures, 71
device model parameters, 192 suppressing measure output, 69
initial condition, 27, 46, 90 unsupported types, 72
DCVOLT, 27, 46 measure continuous
IC, 46 AT and TD qualifiers for TRIG-TARG, 81
NODESET, 90 HSPICE compatibility, 81
measurement output, 79
JFET, 285 rise, fall and cross qualifiers, 80

818
measure fft device output variables, 596
additional examples, 85 mosfet level 2
re-measure, 85 device instance parameters, 306
memristor, 733 device model parameters, 306
memristorpem MOSFET level 2002
device instance parameters, 737 device output variables, 671
device model parameters, 738 mosfet level 3
memristorteam device instance parameters, 308
device instance parameters, 735 device model parameters, 308
device model parameters, 736 mosfet level 6
memristoryakopcic device instance parameters, 310
device instance parameters, 736 device model parameters, 310
device model parameters, 736 MOSFET level 70
MESFET, 289 device output variables, 411
mesfet MOSFET level 70450
device instance parameters, 290 device output variables, 440
device model parameters, 290 MOSFET level 77
mextram 504.12.1 device output variables, 469
device instance parameters, 271 mutualinductor, 193
device model parameters, 271 B-H loop calculations, 198
mextram 504.12.1 with self heating
Converting nonlinear to linear mutual
device instance parameters, 278
indcutors, 198
device model parameters, 278
mutual inductor equations, 196
Microsoft Windows, 791
mvs etsoi 2.0.0
model
device model parameters, 664
binning, 88
mvs hemt 2.0.0
definition, 87
device model parameters, 664
level parameter, 88
mvsg-hv hemt model
scale, 88
device instance parameters, 666
model binning, 88, 117
device model parameters, 666
Monte Carlo analysis, 28, 148
MOSFET, 291
mosfet level 1 netlist
device instance parameters, 304 commands, 22
device model parameters, 304 comment, 161
MOSFET level 102 expression operators, 162
device output variables, 484 functions, 169
MOSFET level 10240 in-line comment, 161
device output variables, 696 line continuation, 161
MOSFET level 103 model definition, 175
device output variables, 519 nodes, 181
MOSFET level 1031 reference, 21
device output variables, 555 subcircuit, 175
MOSFET level 107 node
device output variables, 627 composite parameters, 747, 748
MOSFET level 108 Noise analysis, 91
device output variables, 661 nonlinear mutual inductor
MOSFET level 110 device model parameters, 194

819
Operating Point, 25 rawfile, 793
operating point analysis, 93 ASCII, 793
output binary, 794
control, 52, 76, 82 region
save, 150 composite parameters, 749
resistor, 200
parallel device instance parameters, 202, 203
computing, 19 device model parameters, 202, 204
parameters restart, 109
convergence, 775 file, 110
parser two-level, 110
options, 117 results
PCE analysis, 28, 120, 148 fft, 33
PDE Devices, 176, 741 measure, 56
1D instance parameters, 741, 745 measure continuous, 77
model parameters, 741 measure fft, 82
Physical Models, 751 output control, 52, 76, 82
time integration parameters, 100 output options, 108
POLY, 172 print, 23, 25, 45, 92, 126, 160
port device, 230 sens, 151
power calculations, 184
power grid, 722 SAMPLING analysis, 148
power mosfet sampling analysis, 148
device instance parameters, 377 Sandia National Laboratories, 19
device model parameters, 377 save operating point conditions, 150
powergridbranch scale, 88, 117
device instance parameters, 723 sensitivity, 151
powergridbusshunt solvers
device instance parameters, 725 continuation
powergridgenbus options, 103
device instance parameters, 728 control parameters, 94
powergridtransformer embeddedsampling
device instance parameters, 727 options, 111
psp102va legacy mosfet 102.5 hb
device instance parameters, 472 options, 114
device model parameters, 473 homotopy
psp103va mosfet options, 103
device instance parameters, 490 linear
device model parameters, 491 Aztec, 104, 107
psp103va mosfet with self-heating iterative (preconditioned Krylov methods),
device instance parameters, 526 104
device model parameters, 527 options, 104, 107
PSpice, 777, 783, 785 sparse-direct, 104
E device, 779 Trilinos, 104, 107
F device, 779 nonlinear
G device, 779 options, 101, 776
H device, 779 nonlinear-transient
Probe, 127 options, 102

820
 PCE Voltage Nodes, 181
options, 112 voltage source
pce current controlled, 236
options, 112 independent, 225
sampling nonlinear dependent, 237
options, 111 voltage controlled, 231
sensitivity
options, 113 Y type digital devices, 716
sensitivty
options, 113
time integration
options, 775
time integration, 160, 784
options, 97, 101
STEP analysis
Decade sweeps, 155
Linear sweeps, 153
List sweeps, 155
Octave sweeps, 155
step parametric analysis, 153
subcircuit, 157, 739
designation, 158
name, 157
nesting, 158
node zero, 158
scoping, 158

TCAD Devices, 741

Physical Models, 751
time integration parameters, 100
time step
size, 160, 784
transient analysis, 101, 159
error tolerances, 775
Trilinos, 104, 107

Unix, 20
Users of PSpice, 777
Users of Xyce on Microsoft Windows, 791

VBIC (temperature considerations), 243

vbic 1.3 3t
device instance parameters, 248
device model parameters, 248
vbic 1.3 4t
device instance parameters, 253
device model parameters, 253
voltage controlled current source, 234
voltage controlled voltage source, 231

821
DISTRIBUTION

Email—Internal

Name Org. Sandia Email Address

Technical Library 1911 [email protected]

822
823
Sandia National Laboratories is
a multimission laboratory
managed and operated by
National Technology &
Engineering Solutions of Sandia
LLC, a wholly owned subsidiary
of Honeywell International Inc.,
for the U.S. Department of
Energy’s National Nuclear
Security Administration under
contract DE-NA0003525.