Technische Universität Berlin

Faculty IV - Electrical Engineering and Computer Science

Department of Energy and Automation Technology
Electrical Drives

Master’s Thesis

DESIGN AND PROGRAMMING OF THE

SOFTWARE TO CONTROL THE SENSOR
AND SENSORLESS BLDCM OF THE
ELECTRICAL PROPULSION SYSTEM
FOR THE SOLAR AIRCRAFT

presented by

Pablo Armero Almazán

Matr.-Nr.: 362142

Date of submission: August 28, 2014

Supervisor: Prof. Dr.-Ing. Uwe Schäfer

Tutor: Rong Dong
Declaration

I hereby declare in lieu of an oath that I have produced this work by myself.
All used sources are listed in the bibliography and content taken directly or
indirectly from other sources is marked as such. This work has not been
submitted to any other board of examiners and has not yet been published.
I am fully aware of the legal consequences of making a false declaration.

Eidesstattliche Erklärung

Ich versichere an Eides statt, dass ich die vorliegende Arbeit selbstständig
angefertigt habe. Alle benutzten Quellen sind im Literaturverzeichnis aufge-
führt und die wörtlich oder inhaltlich entnommenen Stellen als solche ken-
ntlich gemacht. Die Arbeit wurde bisher keiner anderen Prüfbehörde vorgelegt
und noch nicht veröffentlicht. Ich bin mir bewusst, dass eine unwahre Erk-
lärung rechtliche Folgen hat.

Place/Date Signature
There is no lack of time, there is lack of interest.
Because when one really wants something,
dawn becomes day, Tuesday becomes Saturday
and a moment becomes an opportunity.
Acknowledgements

First and foremost, I have to thank my parents, without whose sacrifices

and efforts I would not be able to reach this goal. I only hope they are proud
of what their sons have achieved, their immeasurable work finish here, but
my journey has just started. I also need to thank my brother, who I took as
an example and encouraged me to improve day by day, and the rest of my
family, who have always been there to support me.
I would like to express my gratitude to the whole team of CITCEA-UPC,
but namely to Samuel Galceran, who have trusted me and helped me in many
ways, academically and professionally.
A very special thanks goes out to the staff of the Electrical Drives Institute
from the Technische Universität Berlin, for making my stay in Berlin easier
and help me with all I needed.
Last but not least, I must acknowledge my girlfriend, for being there to
support and encourage me at any moment and for bearing me when I am
insufferable.
In conclusion, I recognize that this whole path that now comes to an
end would not have been possible without all this surrounding, which is as
important as sometimes unrecognised.

I
Abstract

A Brushless Direct Current (BLDC) machine and converter are applied

for the electric propulsion system of the solar aircraft. BLDC motors are
electronically commutated instead of mechanically, as Direct Current (DC)
motors do, so the position of the rotor must be known to do the commutation
at the right moment.
The control of BLDC motors can be done in sensor or sensorless mode.
being the use of Hall-effect sensors the most widespread use inside the sen-
sored control. The use of the hall sensors divide a complete revolution of
the motor in 6 sectors, where in each sector only two phases are conducting
(block commutation, trapezoidal control).
However, considering the complexity of the high altitude, the electrical
propulsion system cannot work properly if the hardware of the Hall sensors
is broken. Therefore a sensorless control is necessary.
The BLDC motor provides an attractive candidate for sensorless opera-
tion because the nature of its excitation inherently offers a low-cost way to
extract rotor position information from motor-terminal voltages. The imple-
mented sensorless technique consists of sampling the non-fed phase voltage
and comparing it to half of the DC-link voltage to detect the zero crossing
point.
The most important part of the sensorless control is the sample of the
non-fed phase, so the software must synchronise the sample with the PWM
signal and apply the necessary delays in order to avoid commutation spikes
and execute a correct BEMF measure.
The software must be able to control the motor in both sensored or sen-
sorless mode and to change from one to the other if an error occurs, so the
code has to be compatible between the two modes.

III
Contents

Preface I
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . I
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III
Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XII
List of Code Snippets . . . . . . . . . . . . . . . . . . . . . . . . . . XIII

1. Introduction 1
1.1. Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2. Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3. BLDC Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3.1. History . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3.2. Operation of the BLDCM . . . . . . . . . . . . . . . . 3
Field Oriented Control . . . . . . . . . . . . . . . . . . 5
Hall effect sensors . . . . . . . . . . . . . . . . . . . . . 5
Sensorless control . . . . . . . . . . . . . . . . . . . . . 7

2. Hardware and software tools 13

2.1. XE167F Easy Kit . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.1. XE167F-96F66F . . . . . . . . . . . . . . . . . . . . . 14
2.1.2. The Easy Kit Board . . . . . . . . . . . . . . . . . . . 15
TM
2.2. DAvE version 2 . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3. Keil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3. Implementation 21
3.1. The Capture/Compare Unit 6 . . . . . . . . . . . . . . . . . . 21
3.1.1. PWM generation . . . . . . . . . . . . . . . . . . . . . 22
3.2. Speed and current controller . . . . . . . . . . . . . . . . . . . 23
3.2.1. MAC unit . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.2. Speed calculation . . . . . . . . . . . . . . . . . . . . . 26
Speed reference calculation . . . . . . . . . . . . . . . . 26

V
 Motor speed calculation . . . . . . . . . . . . . . . . . 26
3.3. Hall sensor control . . . . . . . . . . . . . . . . . . . . . . . . 27
3.4. Sensorless control . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.5. ADC measurements . . . . . . . . . . . . . . . . . . . . . . . . 30
3.6. Protections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.7. Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.8. Transition from Hall to Sensorless mode . . . . . . . . . . . . 35
3.8.1. Situation A . . . . . . . . . . . . . . . . . . . . . . . . 36
3.8.2. Situation B . . . . . . . . . . . . . . . . . . . . . . . . 36
3.8.3. Situation C . . . . . . . . . . . . . . . . . . . . . . . . 36
3.8.4. Situation D . . . . . . . . . . . . . . . . . . . . . . . . 37
3.8.5. Software implementation of the transition . . . . . . . 37
3.9. Software design . . . . . . . . . . . . . . . . . . . . . . . . . . 39

References 41

Appendices 47

A. Software structure and full code 47

B. Flowcharts 117
B.1. Flowcharts of the control process . . . . . . . . . . . . . . . . 117

VI
Acronyms

µC Microcontroller
ADC Analog-to-Digital Converter
BEMF Back Electro-Motive Force
BLDC Brushless Direct Current
CAN Controller Area Network
CCU6 Capture/Compare Unit 6
CPU Central Processing Unit
TM
DAVE Digital Application Virtual Engineer
DC Direct Current
DSP Digital Signal Processing
EMF Electro-Motive Force
FOC Field Oriented Control
GPIO General Purpose Input/Output
GPT General Purpose Timer
HW Hardware
IGBT Insulated Gate Bipolar Transistor
IGCT Integrated Gate Commutated Thyristor
ITC Interrupt Controller
JTAG Join Test Action Group
LED Light-Emitting Diode

VII
MAC Multiply-Accumulate unit
MOSFET Metal-Oxide-Semiconductor Field-Effect Transistor
PEC Peripheral Event Controller
PI Proportional-Integral
PMSM Permanent Magnet Synchronous Machine
PWM Pulse-Width Modulation
RPM Revolutions Per Minute
SW Software
UART Universal Asynchronous Receiver-Transmitter
USB Universal Serial Bus
ZCP Zero-Crossing Point

VIII
List of Tables
2.1. Pin selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.1. CCU6 register function in hall sensor mode . . . . . . . . . . . 27

3.2. Hall sensor pattern and corresponding output . . . . . . . . . 28
3.3. ADC channel measurement moment . . . . . . . . . . . . . . . 31
3.4. Error signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

IX
List of Figures

1.1. Full rotation of a BLDC motor . . . . . . . . . . . . . . . . . 4

1.2. Hall signal outputs depending on the magnet position . . . . . 7
1.3. Variation of the BEMF in the phases of a BLDC motor . . . . 8
1.4. Electrical model of a stator terminal . . . . . . . . . . . . . . 10

2.1. XE167 Easy Kit . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.2. Infineon Technologies’ XE166 family classification . . . . . . . 14
2.3. XE167 Easy Kit board . . . . . . . . . . . . . . . . . . . . . . 15
2.4. DAVE main screen for XE167F . . . . . . . . . . . . . . . . . 18
2.5. Keil programming interface . . . . . . . . . . . . . . . . . . . 20

3.1. Configuration of CCU6 channels . . . . . . . . . . . . . . . . . 22

3.2. Configuration of the CCU60_MODCTR register . . . . . . . . 22
3.3. Modulation of the output with T13 . . . . . . . . . . . . . . . 23
3.4. Cascade control block diagram . . . . . . . . . . . . . . . . . . 24
3.5. T12 operation in Hall sensor mode . . . . . . . . . . . . . . . 28
3.6. T13 operation in Sensorless mode . . . . . . . . . . . . . . . . 30
3.7. Measurements synchronised with T13 . . . . . . . . . . . . . . 31
3.8. Noise generated by the modulation of the conducting phases . 32
3.9. Interrupt priority . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.10. Hall sensors fail situation . . . . . . . . . . . . . . . . . . . . . 35

B.1. Flowchart of Main function . . . . . . . . . . . . . . . . . . . 118

B.2. Flowchart of Main_vInit . . . . . . . . . . . . . . . . . . . . . 119
B.3. Flowchart of TRAP_RESET . . . . . . . . . . . . . . . . . . 120
B.4. Flowchart of MOTOR_OPERATION . . . . . . . . . . . . . . 121
B.5. Flowchart of Motor_start . . . . . . . . . . . . . . . . . . . . 122
B.6. Flowchart of Motor_stop . . . . . . . . . . . . . . . . . . . . . 123
B.7. Flowchart of Init_control . . . . . . . . . . . . . . . . . . . . . 124
B.8. Flowchart of Commutation_init . . . . . . . . . . . . . . . . . 125
B.9. Flowchart of RampUp . . . . . . . . . . . . . . . . . . . . . . 126
B.10.Flowchart of Speed_ref_ramp . . . . . . . . . . . . . . . . . . 127
B.11.Flowchart of CCU60_viNodeI1 . . . . . . . . . . . . . . . . . 128

XI
B.12.Flowchart of Transition . . . . . . . . . . . . . . . . . . . . . . 129
B.13.Flowchart of CCU60_viNodeI2 . . . . . . . . . . . . . . . . . 130
B.14.Flowchart of Speed_calc . . . . . . . . . . . . . . . . . . . . . 131
B.15.Flowchart of CCU60_viNodeI3 . . . . . . . . . . . . . . . . . 132
B.16.Flowchart of Sensorless_calc . . . . . . . . . . . . . . . . . . . 133
B.17.Flowchart of ADC0_viSRN0 . . . . . . . . . . . . . . . . . . . 134
List of Code Snippets
2.1. User’s code gap example . . . . . . . . . . . . . . . . . . . . . 19

3.1. Unipolar PWM configuration . . . . . . . . . . . . . . . . . . 23

3.2. PI implementation using MAC unit . . . . . . . . . . . . . . . 25
3.3. Load of the next commutation pattern . . . . . . . . . . . . . 29
3.4. Dead-time for BEMF detection set by CC60_CC62R . . . . . 30
3.5. Forced delay for a correct BEMF measurement . . . . . . . . . 32
3.6. Transition from hall sensor control to sensorless control . . . . 37

XIII
 1

1. Introduction

1.1. Motivation

Electrical mobility consciousness has increased steadily over the past few
years. Seems that the use of electricity as power source is gaining credibility
and is settling on as the future substitute for fossil fuels for mobility, in favour
of a cleaner and better world.
Therefore, motion control plays a leading role in the future development
of this technology, due to the needs of an efficient drive with the best pos-
sible performance provided by a cost-effective solution, to make the most of
the available energy. But electrical mobility is not only related to ground
transportation, but all kinds of mobility, which have different issues and
challenges.

1.2. Objective

The aim of this project is to confront the problems that an electrical drive
can come across in an aircraft electrically propelled. In this case, the pos-
sibility of a failure of the position sensors is taken into account, and this
thesis is intended to find a suitable solution for the BLDC control, whether
the position sensors are functional or they break down. The implementation
of the project encompasses the pin selection for the hardware and the soft-
ware design for hall sensor mode or sensorless mode, as well as the detailed
explanation for further test and modification by future contributors to the
project.

CHAPTER 1. INTRODUCTION
2 1.3. BLDC MOTOR

1.3. BLDC Motor

1.3.1. History

The BLDC motor is developed on the basis of brushed DC motors. The

actual electrical machine theory was established when Faraday discovered the
electromagnetism induction phenomenon in 1831. The first DC motor was
made in the 1840s. Confined by the development of power electronic devices
and permanent magnet materials, BLDC motor was designed successfully
more than one century later. In 1915, Langmuir invented the mercury rec-
tifier to control grid electrode and made the DC/AC converter. Contrasting
the disadvantages of traditional motors, in the 1930s some scholar started de-
veloping brushless motors in which electronic commutation was implemented,
which made preparations for the BLDC motor. However, at that time, power
electronic devices were still in the early stage of development, scholars could
not find and appropriate commutation device.
In 1955, Harrison and Rye made the first patent claim for a thyristor
commutator circuit to take the place of mechanical commutation equipment.
This is exactly the rudiment of the BLDC motor. The principles of oper-
ation are as follows, when the rotor rotates, periodic Electro-Motive Force
(EMF) is induced in the signal winding, which leads to the conduction of
related thyristors. Hence, power windings feed by turns to achieve com-
mutation. However, the problems are, first, when the rotor stops rotating,
induced EMF cannot be produced in the signal windings and the thyristor
is not biased, so the power winding cannot feed the current and this type
of brushless motor has no starting torque. Furthermore, power consump-
tion is large because the gradient of the electric potential’s sloping part is
small. To overcome these problems, researchers introduced the commutators
with centrifugal plant or put an accessory steel magnet to ensure the motor
started reliably. But the former solution is more complex, while the latter
needs an additional starting pulse. After that, by numerous experiments
and practices, the electronic commutation brushless motor was developed
with the outcome of Hall elements in 1962, which inaugurated a new era in
productionization of BLDC motors. In the 1970s, a magnet sensing diode,
whose sensitivity is almost thousands of times greater than that of the Hall
element, was used successfully for the control of BLDC motors. Later, as
the electrical and electronics industry has been developing a large number of
high-performance power semiconductors and permanent magnet materials,
the solid ground for widespread use of BLDC motors is finally established.

CHAPTER 1. INTRODUCTION
1.3. BLDC MOTOR 3

In 1978, the Indramat branch of Mannesmann Corporation of the Federal

Republic of Germany officially launched the MAC brushless DC motor and
its drive system on Trade Shows in Hanover, which indicates that the BLDC
motor had entered into the practical stage. Since them, worldwide further
research has proceeded. Trapezoid-wave/square-wave and sine-wave BLDC
motors were developed successively. The sine-wave brushless DC motor is
the so-called Permanent Magnet Synchronous Machine (PMSM). With the
development of permanent magnet materials, microelectronics, power elec-
tronics, detection techniques, automation and control technology, especially
the power-switched devices like Insulated Gate Bipolar Transistor (IGBT),
Integrated Gate Commutated Thyristor (IGCT) and so on, the BLDC motors
in which electronic commutation is used are growing towars the intelligent,
high-frequency and integrated directions.
In the late 1990s, computer techniques and control theories developed
rapidly. Microprocessors such as microcontroller units (CPU), digital signal
processors (DSP), field programmable gate arrays (FPGA), complex pro-
grammable logic devices (CPLD) made unprecedented development while a
qualitative leap was taken in instruction speed and storage space, which fur-
ther promoted the evolution of BLDC motor. Moreover, a series of control
strategies and methods, such as sliding-mode variable structure control and
so on, are constantly used in BLDC motor drive systems. These methods
can improve the performance on torque-ripple minimization, dynamic and
steady-state speed response and system antidisturbance ability to some ex-
tent, as well as enlarge the application range and enrich the control theory [1].

1.3.2. Operation of the BLDCM

The BLDC motor unlike a DC motor, does not use brushes for commuta-
tion, instead, they are electronically commutated. They have many advan-
tages over DC motors, some of them are [2]:
• High dynamic response.
• High efficiency.
• Long operating life.
• Lower operation noise.
• Higher speed range.

CHAPTER 1. INTRODUCTION
 4 DC
2.1 Brushless 1.3. BLDC MOTOR
2 THEORIE

BLDC motors are synchronous motors, which means that the magnetic
gezeigt. Fürfield
einegenerated
kontinuierliche Drehung
by the stator des
and the Motorsfield
magnetic ist generated
es nötig die Leistungselektronik
by the rotor
rotate
sechs Mal mit at the einem
jeweils same frequency,
anderensoMOSFET-Paar
they do not experience slip as theDa
zu schalten. induction
mithilfe der Hall-
motors. To rotate the BLDC motor, the stator windings have to be energized
◦
sensoren sechs
in a Positionen mitonjeweils
sequence, based 60position.
the rotor Verschiebung dekodiert
A full rotation werden
period in können, werden
hall sensor
mode issoshown
die Hallsensoren in figure 1.1.
angebracht, dass mit einer Änderung einer Hallsensorkodierung eine
Änderung des MOSFET-Paars einhergeht.

(a) 0◦ (b) 60◦ (c) 120◦

(d) 180◦ (e) 240◦ (f) 300◦

AbbildungFigure 1.1.: Full rotation

3: Funktionsweise of a BLDC
eines BLDC motor [3]
in Y-Schaltung [2]

GMR-Sensor Mithilfe von Hallsensoren ist es möglich die Rotorposition bei einer Pol-
haarzahl von eins auf sechs Stellen einzugrenzen. Für einfache Anwendungen, bei de-
nen Position nicht genauer benötigt
CHAPTERwird, reichen diese Informationen, um das richtige
1. INTRODUCTION
MOSFET-Paar anzusteuern. Steigen die Anforderungen an Position und Drehzahldyna-
mik, muss eine genauere Methode für Position- und Drehzahlbestimmung verwendet wer-
den. Eine Möglichkeit ist die Verwendung eines giant magnetoresistance (GMR)-Sensors.
Der GMR-Sensor nutzt den GMR-Effekt, der bei Strukturen mit wiederholenden Schich-
1.3. BLDC MOTOR 5

There are many ways to control a BLDC motor, each one with its advan-
tages and disadvantages, the most used are:
• Field Oriented Control (FOC) (Sinusoidal)
• Hall sensor control (Trapezoidal)
• Sensorless control (Trapezoidal)

Field Oriented Control

FOC is a method which implies the measurement of motor currents and

transform them into a rotating coordinate system synchronised with the rotor
position, thus, to achieve that control the rotor position has to be directly
measured or estimated and the Hardware (HW) peripherals used for the task
must be suitable for an intensive computation compared to scalar controls,
which means more expensive and a more complicated control. On the other
side, it provides better dynamic responses and often more silent operations
[4].

Hall effect sensors

These kinds of devices are based on Hall-effect theory, which states that if
an electric current-carrying conductor is kept in a magnetic field, the mag-
netic field exerts a transverse force on the moving charge carriers that tends
to push them to one side of the conductor. A build-up of charge at the sides of
the conductors will balance this magnetic influence producing a measurable
voltage between the two sides of the conductor. The presence of this mea-
surable transverse voltage is called the Hall-effect because it was discovered
by Edwin Hall in 1879.
Most BLDC motors have three Hall sensors inside the stator on the non-
driving end of the motor. Whenever the rotor magnetic poles pass near the
Hall sensors they give a high or low signal indicating the N or S pole is
passing near the sensors as shown in figure 1.2 [5]. Based on the combination
of these three Hall sensor signals, the exact sequence of commutation can be
determined.
Embedding the Hall sensors into the stator is a complex process because
any misalignment in these Hall sensors with respect to the rotor magnets
will generate an error in determination of the rotor position. To simplify the

CHAPTER 1. INTRODUCTION
6 1.3. BLDC MOTOR

process of mounting the Hall sensors onto the stator some motors may have
the Hall sensor magnets on the rotor, in addition to the main rotor magnets.
Therefore, whenever the rotor rotates the Hall sensor magnets give the same
effect as the main magnets. The Hall sensors are normally mounted on a
printed circuit board and fixed to the enclosure cap on the non-driving end.
This enables users to adjust the complete assembly of Hall sensors to align
with the rotor magnets in order to achieve the best performance [6].
A BLDC motor is driven by voltage strokes coupled with the rotor position.
These strokes must be properly applied to the active phases of the three-phase
winding system so that the angle between the stator flux and the rotor flux
is kept close to 90o to get the maximum generated torque. Therefore, the
controller needs some means of determining the rotor’s orientation/position,
the Hall-effect sensors.
The process of switching the current to flow through only two phases for
every 60 electrical degree rotation of the rotor is called electronic commuta-
tion. The motor is supplied from a three-phase inverter, and the switching
actions can be simply triggered by the use of signals from position sensors
that are mounted at appropriate points around the stator.
Such a drive usually also has a current loop to regulate the stator current,
and an outer speed loop for speed control. The speed of the motor can be
controlled if the voltage across the motor is changed, which can be achieved
easily varying the duty cycle of the PWM signal used to control the six
switches of the three-phase bridge. Only two inverter switches, one in the
upper inverter bank and one in the lower inverter bank, are conducting at
any instant. These discrete switching events ensure that the sequence of
conducting pairs of stator terminals is maintained [7].
One of the hall sensors changes the state every 60 electrical degrees of ro-
tation. Given this, it takes six steps to complete an electrical cycle. However,
one electrical cycle may not correspond to a complete mechanical revolution
of the rotor. The number of electrical cycles to be repeated to complete a
mechanical rotation is determined by the rotor pole pairs. For each rotor pole
pair, one electrical cycle is completed. The number of electrical cycles/ro-
tations equals the rotor pole pairs [2]. This sequence of conducting pairs is
essential to the production of a constant output torque.
In summary, permanent magnet motor drives require a rotor position sen-
sor to properly perform phase commutation, but there are several drawbacks
when such types of position sensors are used. The main drawbacks are the
increased cost and size of the motor, and a special arrangement needs to be

CHAPTER 1. INTRODUCTION
1.3. BLDC MOTOR 7

made for mounting the sensors. Further, hall sensors are temperature sensi-
tive and hence the operation of the motor is limited, which could reduce the
system reliability because of the extra components and wiring [8]. To reduce
cost and improve reliability such position sensors may be eliminated. To this
end, many sensorless schemes have been reported for position (and speed)
control of BLDC motors [9].

Figure 1.2.: Hall signal outputs depending on the magnet position

One of the advantages of this control is that, with only two phases con-
ducting the motor can rotate and with the hall sensor signals the position
and speed can be calculated, so the Microcontroller (µC) does not need a
large computing capability as the ones used for FOC, reducing the final cost.
If sensors are used for FOC, normally they are more expensive than hall ef-
fect sensors, but also more precise, such as incremental encoders or resolvers.
The ease of control and the good performance/cost ratio makes the use of
hall effect sensors one of the most used control techniques for BLDC motors.

Sensorless control

Problems with the cost and reliability of rotor position sensors have moti-
vated research in the area of position sensorless BLDC machine drives. The
sensorless control does not use any position sensor, it is based on the mea-
surement of the non-conducting phase of the motor and the commutation
time is calculated based on the Back Electro-Motive Force (BEMF). The
BEMF in a BLDC motor is trapezoidal, as seen in figure 1.3, whereas in
PMSM is sinusoidal. A BLDC motor can be driven with sinusoidal currents
and a PMSM can be driven with direct currents, but for better performance,
PMSM should be excited by sinusoidal currents and BLDC motors by direct
currents.

CHAPTER 1. INTRODUCTION
8 1.3. BLDC MOTOR

Figure 1.3.: Variation of the BEMF in the phases of a BLDC motor [12]

The BLDC motor provides an attractive candidate for sensorless opera-

tion because the nature of its excitation inherently offers a low-cost way to
extract rotor position information from motor-terminal voltages. In the ex-
citation of a three-phase BLDC motor, except for the phase-commutation
periods, only two of the three phase windings are conducting at a time and
the no conducting phase carries the back-EMF. There are many categories of
sensorless control strategies [9]; however, the most popular category is based
on back electromotive forces or BEMFs [10]. Sensing back-EMF of unused
phase is the most cost efficient method to obtain the commutation sequence
in star wound motors. Since back-EMF is zero at standstill and proportional
to speed, the measured terminal voltage that has large signal-to-noise ratio
cannot detect zero crossing at low speeds. That is the reason why in all
BEMF-based sensorless methods the low-speed performance is limited, and
an open-loop starting strategy is required [11].
BEMF sensing methods for BLDC motors are split in two categories: direct
and indirect BEMF detection [13].

Direct BEMF detection methods The BEMF of floating phase is sensed

and its zero crossing is detected by comparing it with neutral point voltage.
This scheme suffers from high common node voltage and high frequency
noise due to the Pulse-Width Modulation (PWM) drive, so it requires low
pass filters and voltage dividers. This method includes:

CHAPTER 1. INTRODUCTION
1.3. BLDC MOTOR 9

• BEMF Zero-Crossing Point (ZCP) or Terminal Voltage Sensing

• PWM strategies

Indirect BEMF detection methods Because filtering introduces commuta-

tion delay at high speeds and attenuation causes reduction in signal sensitiv-
ity at low speeds, the range is narrowed in direct BEMF detection methods.
In order to reduce switching noise, the indirect BEMF detection methods:
• BEMF Integration
• Third harmonic Voltage Integration
• Free-wheeling Diode Conduction or Terminal Current Sensing
The method used in the scope of this project is sampling the non-fed phase
and compare this value to half of the DC-link voltage. The BEMF sampling is
done during PWM ON statm allowing up to a maximum duty cycle (100%).
The use of high frequency sampling without analog filtering and with the
possibility of digital filtering gives more flexibility to this method.
The BEMF can be calculated as in equation 1.1:

E = 2N lrBω (1.1)

where N is the number of winding turns per phase, l is the length of the
rotor, r is the internal radius of the rotor, B is the rotor magnet flux density
and ω is the motor’s angular velocity.
The main objective in this sensorless control is to detect the ZCP. The ZCP
is where the BEMF value is null, which occurs 30o after the commutation,
thus, at half time between two commutations.
In figure 1.4 is depicted the motor terminal model, where Lx is the phase
inductance, R is the phase resistance, E is the BEMF, Vn and Vx are, re-
spectively, the star connection and phase voltages referenced to ground [14].

Then, the equation of the electrical model is as in equation 1.2:

dIx
Vx = RIx + L + Ex + Vn (1.2)
dt

CHAPTER 1. INTRODUCTION
10 1.3. BLDC MOTOR

Ix Lx Ex

Vx Vn

Figure 1.4.: Electrical model of a stator terminal

Considering C as the non-fed phase, which means Ic = 0, the equation for

the three phases can be rewritten as:
dIa
Va = RIa + L + Ea + Vn
dt
dIb
Vb = RIb + L + Eb + Vn
dt
Vc = Ec + Vn

As phase C is not conducting, the current flows only through phase A and
B. Therefore:
Ia = −Ib

Including this into the above equations and adding the three terminal
voltage equations, the result is:

Va + Vb + Vc = Ea + Eb + Ec + 3Vn

As was depicted in figure 1.3, at the ZCP the sum of the three BEMF is
null, then the above equation can be rewritten as:

Va + Vb + Vc = +3Vn (1.3)

At ZCP the BEMF of phase C is also 0, so its phase equation is equal to:

CHAPTER 1. INTRODUCTION
1.3. BLDC MOTOR 11

Vc = Vn (1.4)

Inserting equation 1.4 in equation 1.3 and assuming that high-side Metal-
Oxide-Semiconductor Field-Effect Transistor (MOSFET) of phase A and low-
side MOSFET of phase B are closed, the final equation is:

VDC + 0 + Vc = +3Vc
where Vc at the ZCP is finally:

VDC
Vc = (1.5)
2

So, as seen in equation 1.5, the Software (SW) in the implemented sensor-
less control will detect the ZCP when the phase voltage is half of the DC link
and proceed to calculate the time for the next commutation period. This
process will be explained in detail in section 3.4.

CHAPTER 1. INTRODUCTION
2. Hardware and software tools

2.1. XE167F Easy Kit

The hardware used in this thesis is the Infineon Thechnologies’ Easy Kit
XE167F depicted in figure 2.1. This kit includes a µC from the XE166
family, the XE167F-96F66F and many peripherals such as an on board Uni-
versal Serial Bus (USB) to Join Test Action Group (JTAG)/Universal Asyn-
chronous Receiver-Transmitter (UART) interface, a potentiometer, a push-
button, Light-Emitting Diodes (LEDs) and the documentation and examples
needed for a quick start [15].

Figure 2.1.: XE167 Easy Kit

13
14 2.1. XE167F EASY KIT

2.1.1. XE167F-96F66F

The µC included in this Easy Kit is the XE167F-96F66F. This µC is a 144-

pin device of the 16-bit architecture family XE166, which uses the C166v2
core. In figure 2.2 a classification of the family products is shown.

Figure 2.2.: Infineon Technologies’ XE166 family classification

Some of the main features of this µC are [16]:

• Maximum clock frequency of 66 MHz
• One-cycle multiply-and-accumulate instructions with the Multiply-Accumulate
unit (MAC)
• 64 kByte SRAM
• 768 kByte FLASH
• Two Synchronizable Analog-to-Digital Converter (ADC) with up to 24
channels, 10-bit resolution, conversion time below 1 µs, optional data
preprocessing (data reduction, range check)

CHAPTER 2. HARDWARE AND SOFTWARE TOOLS

2.1. XE167F EASY KIT 15

• Up to 4 Capture/Compare Unit 6 (CCU6)

• 5 General Purpose Timers (GPTs)
• C programmable
• 5 Controller Area Network (CAN) nodes
• 6 Serial channels
• Up to 118 General Purpose Input/Output (GPIO) lines
• Single power supply from 3.0 V to 5.5 V

2.1.2. The Easy Kit Board

The XE167F Easy Kit board includes all the necessary peripherals to per-
form a complete motion control chain. It includes two connectors BU101 and
BU102 specially designed to act as the interface to a power board for motion
control [15]. The Easy Kit board is shown in figure 2.3.

Figure 2.3.: XE167 Easy Kit board

CHAPTER 2. HARDWARE AND SOFTWARE TOOLS

16 2.1. XE167F EASY KIT

To develop the software, it is necessary to first select the needed pins. The
needs of the control are:
• 6 PWM signals
– 3 for high-side MOSFETs
– 3 for low-side MOSFETs
• 7 ADC channels
– 3 for phase voltages
– 1 for DC link voltage
– 1 for DC link current
– 2 for phase currents (for future FOC implementation)
• 3 pins for protection
– 1 for overvoltage
– 1 for overtemperature
– 1 for a driver error latch
• 3 inputs for hall sensors
To select the most suitable pins of the µC, the before mentioned connectors
BU101 and BU102 are used. The final pin selection, with their function and
connector of the board is detailed in the table 2.1.

CHAPTER 2. HARDWARE AND SOFTWARE TOOLS

2.1. XE167F EASY KIT 17

Purpose Pin µC function Connector

CC60 (10.0) Low-side 1
COUT60 (10.3) High-side 1
CC61 (10.1) Low-side 1
PWM outputs
COUT61 (10.4) High-side 2 BU101
CC62 (10.2) Low-side 3
COUT62 (10.5) High-side 3
Driver error output CTRAP (10.6) _ERR
CCU60_CCPOS0 (10.7) Hall 1
Hall sensors CCU60_CCPOS1 (10.8) Hall 2
CCU60_CCPOS2 (10.9) Hall 3
BU102
ADC0_CH3 (5.3) IDC
Currents ADC0_CH8 (5.8) Ia
ADC0_CH13 (5.13) Ib
ADC1_CH1 (15.1) Va
ADC1_CH4 (15.4) Vb X105A
Voltages
ADC1_CH7 (15.7) Vc
ADC1_CH2 (15.2) VDC X105B
Overvoltage GPIO (5.14) Port input
X106A
Overtemperature GPIO (5.11) Port input

Table 2.1.: Pin selection

Once the pins have been selected, the next step is to configure the periph-
erals of the µC to use this pins with their corresponding function.

CHAPTER 2. HARDWARE AND SOFTWARE TOOLS

 TM
18 2.2. DAVE VERSION 2

TM
2.2. DAvE version 2
TM
Digital Application Virtual Engineer (DAVE ) is a free application with
which the user can start developing the code for the µC. It provides a visual
interface that makes it easier to configure the different peripherals and ini-
tialise the desired application, generating the needed C-code with the main
driver functions and interrupts. The main screen of the XE167F-96F66F
TM
in the DAVE environment is depicted in figure 2.4. In this project, the
TM
second version of DAVE generation tools is used, because it is the one
compatible with the XE166 family [17].

Figure 2.4.: DAVE main screen for XE167F

This programm will generate the main function and the peripheral initial-
isation as well as the interrupt functions, so there is only the need to create
the application specific code inside the generated functions. For that pur-
pose, the generated files include gaps where the user can introduce its own
TM
code, so that, if the DAVE project is modified, this code will stay intact.
This gaps are implemented as in listing 2.1:

CHAPTER 2. HARDWARE AND SOFTWARE TOOLS

2.3. KEIL 19

while(1) // Enter the endless loop.

{
// USER CODE BEGIN (Main,4)
TRAP_RESET(); // Reset the software trap flag if requested by
the user.
MOTOR_OPERATION(); // Start or Stop the Motor as per the present
conditions.
RampUp(); // Ramp up(or down) the motor to the new speed.
if(P5_IN & 0x4800) // If overvoltage or overtempreature
detected...
{
status_overvoltemp = 1; // ... set the overtemperature and
overvoltage flag.
Motor_stop(); // ... stop the motor.
}
// USER CODE END
}
Listing 2.1: User’s code gap example

TM
Once the peripherals are configured with the DAVE environment, the
application code must be added and edited in the C compiler. Infineon
products can be programmed in many development environments, but the
most common used are Tasking and Keil.

2.3. Keil

To develop all the code and compile the software, the latest version of the
Keil compiler (Keil µVision3) is used. This software includes the necessary
TM
files to program the µC and can directly import a DAVE project.
This is the last step of the tool chain, where finally the code is devel-
oped and loaded into the µC. The programming environment is depicted in
figure 2.5.

CHAPTER 2. HARDWARE AND SOFTWARE TOOLS

20 2.3. KEIL

Figure 2.5.: Keil programming interface

CHAPTER 2. HARDWARE AND SOFTWARE TOOLS

3. Implementation

In this chapter are explained the main parts of the developed code, such
as the use of the CCU6, the speed and current controllers, the two control
modes, protection routines and the transition between control modes.
The most important peripherals used to develop the software are the ADCs
and the CCU6, the ones that receive the signals to decide the commutation
time and allow the protection interrupt routines and fast output responses.

3.1. The Capture/Compare Unit 6

The CCU6 is a high-resolution 16-bit capture and compare unit, specially

designed for AC drive control. It supports block commutation control and
control mechanisms for multi-phase machines. One of the main important
features of this unit is the possibility to operate in hall sensor mode, di-
TM
rectly configured with DAVE , which will configure the needed registers to
capture the hall sensor signals and synchronise the commutation [18, 19].
The main blocks inside the CCU6 are the two timers (the T12 and the
T13) and the Capture/Compare registers (CC6x).
The timer T12 is the one which executes the commutation of the outputs
and is also used to measure the speed of the motor. In both control methods it
operates as a timer, but the configuration of the compare registers is different
between the hall sensor and sensorless method.
The CCU6 module can also enable or disable the modulation of the out-
puts, as well as choosing which timer can do the modulation and select which
events generate an interrupt to have a full control of the process.
TM
See in figure 3.1 how the inputs are configured in the DAVE environ-
ment.

21
22 3.1. THE CAPTURE/COMPARE UNIT 6

Figure 3.1.: Configuration of CCU6 channels

3.1.1. PWM generation

The control method in both hall sensor mode and sensorless mode is with
unipolar PWM. This means that only the high-side MOSFET is modulated
when it is enabled, whereas the low-side MOSFET stays closed the whole
period. Unipolar switching mode reduces electromagnetic noise, DC bus
ripple and commutation losses because there is less switching [20]. As most
of the time the motor will run with hall sensors active, this method will be
implemented as the primary control, while the sensorless control will only
take part if the hall sensors fail.
The modulation is done with the T13, and the channels which are allowed
to be modulated by the T13 are configured in the register CCU60_MODCTR
shown in figure 3.2.

Figure 3.2.: Configuration of the CCU60_MODCTR register

CHAPTER 3. IMPLEMENTATION
3.2. SPEED AND CURRENT CONTROLLER 23

The even bits of the T13MODEN correspond to the outputs CC6x, while
the odd bits to the COUT6x. So the final configuration for the register is as
in listing 3.1.

CCU60_MODCTR = 0x2A80; // load CCU60 modulation control register

Listing 3.1: Unipolar PWM configuration

The high-side MOSFET will be closed when the T13 value is over the
compare register CCU60_CC63R, as shown in figure 3.3.

COUT60

CC60

COUT61

CC61

COUT62

CC62

CC63

Figure 3.3.: Modulation of the output with T13

3.2. Speed and current controller

The speed and current controller are both implemented using the common
used Proportional-Integral (PI) controller, but in a cascade control system.
A cascade control is characterized by a master control loop (the outer and
slower one) and a faster, slave, inner loop. The cascade control inner loop
compensates for specific disturbances at source and largely prevents them
from affecting the process being controlled, improving considerably the dy-
namic response. To implement a cascade control, the inner loop must have
a faster response than the outer loop, at least 3-5 times faster [21]. This is
permits that, in an electric motor, the speed can be the outer or primary loop
and the current can be the inner or secondary loop, due to the fact that the

CHAPTER 3. IMPLEMENTATION
24 3.2. SPEED AND CURRENT CONTROLLER

electrical time constant is normally much more smaller than the mechanical
time constant. A block diagram of a cascade loop is depicted in figure 3.4.

+ PI + PI PWM/ Va
V
ωref ωerr iref ierr α b BLDC
controller controller Driver Vc
- -
ω i

Figure 3.4.: Cascade control block diagram

The speed PI controller will execute every 1 ms, whereas the current PI
every 50 µs, synchronised with the T13 period match.

3.2.1. MAC unit

The cascade control implementation in software uses the MAC unit of the
µC. The MAC unit enhances strongly the DSP performance and 32 bit data
processing capability. It is a built-in unit in the CPU core, which means
that can be considered as an additional arithmetic unit dedicated for Digital
Signal Processing (DSP) and 32 bit calculation. In the execution stage of
the instruction processing pipeline, the instruction will be put into the ALU
or MAC unit to be executed according to which instruction set (normal or
MAC) the code comes from. If the instruction comes from normal instruction
set, the code execution will be done in the ALU unit. If the instruction is
from the MAC instruction set with CoXXX prefix, this instruction will be
processed in the MAC unit [22].
Infineon and Keil provides a library with many mathematical functions
which use this unit to reduce the calculation time in complex routines. In this
project some of these functions are used, i.e. the pi_controller64 function.
This function calculates the output of a PI controller whose inputs and output
are in 1Q15. The calculation is done in only 42 clock cycles, which means
only 636,4 ns.
Listing 3.2 shows the code where the PI is implemented. See that is totally
implemented in assembler and the commands used are referenced to the MAC
unit.

CHAPTER 3. IMPLEMENTATION
3.2. SPEED AND CURRENT CONTROLLER 25

int pi_controller64(long *pi_parameter,int reference,int actual)

{
long a;
#pragma asm

mov R12,MCW ;Save MCW register

mov MCW,#1536 ;Set saturation and shift left
mov R11,ZEROS ;Load zero in R11
CoLOAD R11,R9 ;Load Accumulator (High) with R9 (reference)
CoSUB R11,R10 ;error = reference - actual
CoSTORE R9 ,MAS ;Load error in R9
mov R3,[R8+] ;
CoLOAD R3,[R8+] ;Load yn (integral buffer) in accumulator
mov R1,R8 ;Save parameters addres in R1
mov R5,[R1+] ;Load Kp (proportional Constant) in R5
mov R6,[R1+] ;Load Ki = T0/Ti (integral Constant) in R6
CoMAC R6,R9 ;yn = Ki * error + yn
mov R6,[R1+] ;Load ymax (limit value max)
mov R7,[R1+] ;Load ymin (limit value min)
CoMIN R11,R6 ;Limit max yn
CoMAX R11,R7 ;Limit min yn
CoSTORE R4,MAH ;Store yn-high in R4
CoSTORE R3,MAL ;Store yn-low in R3
mov [-R8],R4 ;Store R4 in integral buffer(High)
mov [-R8],R3 ;Store R3 in integral buffer(Low)
CoMUL R5,R9 ;Kp * error
CoSHL #6 ;64 * Kp * error
CoADD R3,R4 ;y = yn + (64 * Kp * error)
CoMIN R11,R6 ;Limit max y
CoMAX R11,R7 ;Limit min y
CoSTORE R4,MAS ;Store y-high in R4 (return register)
mov MCW,R12 ;Restore MCW register
#pragma endasm
a = reference;
a = actual;
a = *pi_parameter;
}

Listing 3.2: PI implementation using MAC unit

CHAPTER 3. IMPLEMENTATION
26 3.2. SPEED AND CURRENT CONTROLLER

3.2.2. Speed calculation

Speed reference calculation

In both control methods, the speed reference is not set directly, but con-
stantly increased or decreased as a ramp. This makes that the controller
follows the new vale avoiding oscillations and great current spikes due to big
accelerations or decelerations. The RampUp() and Speed_ref_ramp() func-
tions are the ones that compute this constant approach to the end reference
value.
The function RampUp() will increment the counter every T13 period match,
i.e. every 50 µs, and when the counter value is equal to the predefined
time step, it will call the Speed_ref_ramp() and reset the counter. Inside
the Speed_ref_ramp(), the speed reference used in the PI is incremented or
decremented if it is not equal to the end reference (the final value to achieve).
The RampUp function also recalculates the time step necessary if the end
reference is modified, in order to reach the end value in the predefined time
stored in the variable rampup_time.

Motor speed calculation

The measured speed of the motor is calculated from different values of

the timer T12. The function Speed_calc receives as input parameters a T12
sample corresponding to the number of counts of the timer between commu-
tations (in hall sensor mode) or between ZCP (in sensorless mode), a counter
value and a limit value, which will define if there are enough samples to start
averaging the speed.
The main speed calculation is based on the use of a speed buffer which
can accumulate up to 6*Pole Pairs samples. That is because every 6 sector a
complete electrical revolution is done, the pole pairs indicate the number of
electrical revolutions per mechanical revolution. Therefore, the speed buffer
contains the amount of counts per mechanical revolution, which only needs to
be converted to Revolutions Per Minute (RPM). To do that, the equation 3.1
is used.

60
n= (3.1)
speed_buf f ∗ T 12resolution

CHAPTER 3. IMPLEMENTATION
3.3. HALL SENSOR CONTROL 27

However at the beginning there are not enough samples to implement the
average of the speed, so only the sector time is used, and the equation to
convert it to RPM is slightly different:

60
n=
sector_time ∗ 6 ∗ P p ∗ T 12resolution

3.3. Hall sensor control

The main control of the motor will be with the hall sensors active. The
hall sensor signals are connected to pins 10.7, 10.8 and 10.9 as shown in ta-
ble 2.1. In this mode, the timer T12 is configured as a timer, the registers
CCU60_CC61 and CCU60_CC62 as compare registers, and the CCU60_CC60
as a capture register, which will read the hall sensors input. In table 3.1, the
different registers and its function are listed.
The timer T12 clock frequency will be 1/256 the Central Processing Unit
(CPU) clock frequency, whereas the T13 will have the same clock as the
CPU. This makes the period of T13 to be short enough to modulate the
output signals between commutation (fpwm = 25 kHz), while the timer T12
can be used to measure the speed without overflowing too fast.

Register Configuration Function

Detects an edge in any hall
CCU60_CC60 Capture mode
sensor signal
Requests the new commuta-
tion if hall pattern is correct
CCU60_CC61 Compare mode
after a short time to avoid
noise
Will trigger the speed cal-
CCU60_CC62 Compare mode culation after the commuta-
tion

Table 3.1.: CCU6 register function in hall sensor mode

A drawing explaining how it works is shown in figure 3.5. When the

CCU60_CC60 detects an edge in the hall sensors inputs, it compares the
read pattern with the actual and the expected pattern, and, if it matches the
expected pattern, it will store the T12 value and reset the timer.

CHAPTER 3. IMPLEMENTATION
28 3.3. HALL SENSOR CONTROL

After a short time to avoid noise in the sensor signals, the commutation is
requested when CCU60_CC61 compare register is matched and is synchro-
nized at T13 one match. The speed calculation is done when T12 equals
CCU60_CC62 compare register [23].

Stores-value-in-
CC60_R Period-time-out-
and-reset-the-timer- error

T12PR
Launches-speed-
calculation

CC62_R
CC61_R
Commutation-
after-short-time
CCPOSx
1 0

COUT6x

Figure 3.5.: T12 operation in Hall sensor mode

In the interrupt serviced by a CCU60_CC62R compare match, the speed

is calculated based on the T12 counter between correct hall events and the
next commutation pattern is loaded. The commutation pattern of the motor
is listed in table 3.2.

Hall input 001 011 010 110 100 101

Clockwise C+B- A-C+ A-B+ C-B+ A+C- A+B-
Counterclockwise C-B+ A+C- A+B- C+B- A-C+ A-B+

Table 3.2.: Hall sensor pattern and corresponding output

This values are stored in the variable SP_Tab_r_l. Each element of this
variable consists of the current hall signal, the expected hall signal and the
corresponding output. The order of the table is selected so, for an index
equivalent at the actual hall sensor state, it returns the next commutation
pattern. So a load of the next pattern is done as in listing 3.3.

CHAPTER 3. IMPLEMENTATION
3.4. SENSORLESS CONTROL 29

CCU60_MCMOUTS = SP_Tab_r_l[((CCU60_MCMOUT & 0x3800) >> 11) +

direction]; // ... load the next commutation pattern.
Listing 3.3: Load of the next commutation pattern

The variable direction provides an offset to choose between the clockwise

or counterclokwise sequence.

3.4. Sensorless control

For the sensorless control we also use the T13 and T12 of the CCU6 with
the same configuration. But the registers are not configured as in hall sensor
mode, therefore, they are all in compare mode. Despite this change, the
CCU60_CC61 and CCU60_CC62 have the same function as in hall sensor
mode, in order to make the two control methods as much code-compatible
as possible [24].
The difference in this method is that the speed is not calculated between
commutations, but between ZCPs. Also the timer T12 is restarted after
detecting a ZCP. As was deduced in equation 1.5, the ZCP is reached when
the non-fed phase voltage equals half of the DC link voltage. When this it is
detected, the software will recalculate the next commutation time and store
it in CC60_CC61R, to request the commutation at the required moment.
The CC60_CC62R, which triggers the interrupt for the speed calculation,
has a value a little bit higher than CC60_CC61R, the same as in hall sensor
mode. The T12PR is still used to trigger a time-out error interrupt if needed.
To detect the ZCP, many measures must be done during each sector. These
measures can not be done randomly, because the measured value may not be
always appropriate. To evaluate the BEMF properly, the voltage measure-
ment must be done while the two phases are conducting, that is, after the
compare match of T13. The performance in sensorless mode is illustrated in
figure 3.6.
It is also important, that the BEMF measurement do not start just after
the commutation, since the commutation spikes could lead to a false ZCP,
this is achieved by the delay between CC60_CC61R and CC60_CC62R,
making CC60_CC62R set the start of a slope 0 (rising) or 1 (falling). The
code in 3.4 shows how inside the interrupt the corresponding slope, based on
the previous ZCP detection, is set. More detailed explanation in appendix B.

CHAPTER 3. IMPLEMENTATION
30 3.5. ADC MEASUREMENTS

StoresUvalueUinUCC60_R,U PeriodUtime-outU
resetUtheUtimerUT12UandUloadU error
CC61_R
T12PR
LaunchesUspeedU
calculation

CC62_R Dead-timeUwhereUBEMFU
CC61_R }U
valuesUareUignoredU

ZCP ZCP

COUT6x
CommutationUatU
CC61_RUmatch
Figure 3.6.: T13 operation in Sensorless mode

if (slope == 2) // If a falling edge has been detected...

{
slope = 0; // ... we need to look for rising edge.
}
if (slope == 3) // If a rising edge has been detected...
{
slope = 1; // ... we need to look for falling edge.
}

Listing 3.4: Dead-time for BEMF detection set by CC60_CC62R

3.5. ADC measurements

As explained in section 3.4, the exact time of the measurements is really

important for obtaining a reliable BEMF measure. Therefore, the measure-
ments are synchronised with the T13 period or compare match depending
on the signal to be measured. However, the measurement is not done at the
exact moment of the match, since the sample will include the commutation
spikes, but a short time after. Based on the ADC used by each measure and

CHAPTER 3. IMPLEMENTATION
3.5. ADC MEASUREMENTS 31

the time requirements, the table 3.3 shows when each channel is converted.
The figure 3.7 show the ADC conversions synchronised with the T13 during
the switch pattern A+ B-, that is, measuring BEMF of phase C.

ADC Channel Measurement moment

ADC0_CH3 T13 Compare match
ADC1_CH1 T13 Compare match (if non-fed)
ADC1_CH2 T13 Period match
ADC1_CH4 T13 Compare match (if non-fed)
ADC1_CH7 T13 Compare match (if non-fed)

Table 3.3.: ADC channel measurement moment

COUT60

CC61
Delay 4us

ADC0
IDC

ADC1
VC
VDC

T13
CC63

Figure 3.7.: Measurements synchronised with T13

As can be seen in figure 3.7, although the conversions of the phase voltages
samples are synchronised with the T13 compare match, there is a delay
between the compare event and the start of the conversion. This delay is
done in order to avoid the noise detection caused by the commutation of
the MOSFETs, as shown in figure 3.8, and to take into account the delay
introduced by the optocoupler used in the hardware to isolate the ADC from
the high voltage circuit.
This delay introduced by the optocoupler has been measured and is equal
to 4µs, so it is necessary to add a delay by software, to select the appropri-

CHAPTER 3. IMPLEMENTATION
32 3.5. ADC MEASUREMENTS

Figure 3.8.: Noise generated by the modulation of the conducting phases

ate time, the number of cycles used to service the interrupt and start the
conversion must be taken into account:

• Cycles to service the interrupt = 7-11

• Cycles of the arbitration round of the ADC = 4
• Cycles to the conversion request = 6 cycles

The 4µs delay is equivalent to 264 clock cycles, which means that the delay
function must last:

F orced delay = 264 − 11 − 6 − 4 = 243cycles

To implement this delay this the function _nop_() is used, which is equal to
1 cycle delay. Taking into account the optimization level and the loop check,
the final loop implemented is in the listing 3.5.
for(x=0; x<200; x++) {
_nop_(); // No
operation function delay
}
Listing 3.5: Forced delay for a correct BEMF measurement

CHAPTER 3. IMPLEMENTATION
3.6. PROTECTIONS 33

3.6. Protections

The protection routines are the most important, because they prevent the
motor from being damaged. There are many different errors that can cause
permanent damage to the motor, so the software must identify and prevent
the damage by stopping the motor if any error signal is detected and register
which fault caused the motor to stop. In table 3.4 are listed the errors and
how are they handled by the software.

Error Handled by
Overcurrent ADC0_SRN0INT (Over specified limit)
Overvoltage Main loop
Overtemperature Main loop
Driver error CCU60_IS_TRPF (TRAP)
Time-out CCU60_IS_T12PM (T12 Period match)
Wrong Hall Event CCU60_IS_WHE

Table 3.4.: Error signals

Most of the error signals will stop the motor in case they are latched, with
the exception of the overcurrent and wrong hall event erros. The overcurrent
interrupt will check the value of the measured current and, if its inside a
specified limit, it will not stop the motor until a predefined amount of time
is exceeded, avoiding a motor stop only if momentary current peaks occur.
The wrong hall event error occurs when the hall sensor signals fail, in this
case, before stopping the motor, the software will try to change to sensorless
control if possible. This error is the purpose of the thesis and is treated into
detail in section 3.8.

3.7. Interrupts

The architecture of the XE167F supports several mechanisms for fast and
flexible response to service requests from internal or external sources:
• Interrups generated by the Interrupt Controller (ITC)
• DMA transfers issued by the Peripheral Event Controller (PEC)
• Traps caused by the TRAP instruction or system specific states

CHAPTER 3. IMPLEMENTATION
34 3.7. INTERRUPTS

The normal interrupt processing has a total of 96 separate interrupt nodes

assignable to 16 priority levels subdivided in 8 groups, which allows the user
to specify in which order multiple interrupt requests must be handled. Dur-
ing a normal interrupt process, the CPU temporarily suspends the current
program execution and branches to an interrupt service routine to service
the requested flag [25].
In the application, there are many interrupts that can be active at the same
time. Therefore, the protection interrupts will be prioritised before the other
requests. In figure 3.9 is depicted the priority of the software interrupts. The
higher the interrupt level, the higher priority, and if two interrupts have the
same priority level, the CPU will service first that with a higher group level.

Figure 3.9.: Interrupt priority

The selection of the interrupts and their routines is one of the most im-
portant parts of the application, due to the fact that being two different
control methods, both need to work with the same interrupts and with as
much code-sharing as possible to allow a fast transition from one control to
the other and reduce the size of the final application.

CHAPTER 3. IMPLEMENTATION
3.8. TRANSITION FROM HALL TO SENSORLESS MODE 35

3.8. Transition from Hall to Sensorless mode

The main objective of the thesis is developing a software capable of switch-

ing from a hall sensor control to a sensorless control if the hall sensors break
stop working properly. To do this transition it is necessary to compare both
control methods and their code to unify as much as possible the functions,
interrupts and variables. Therefore, during the hall sensor control mode, the
BEMF of the non-fed phase is measured, so that if the sensors fail, the con-
trol method can be switched. This entails that the size of the application is
larger, as well as the CPU load, but on the other hand it can prevent the
motor from stopping in case of error event.
As the failure of the sensors cannot be predicted, the software has to do the
transition independently of when they occur. To achieve that, is necessary
to evaluate the different situations when the hall sensors can stop working
and proceed based on the moment of breaking.
The different situations when the hall sensors can fail are depicted in fig-
ure B.12.

A B C D
Expected0next0
commutation
Slope0=000or0Slope0=01
BEMF0not0detected0yet0in0 Slope0=020or0Slope0=03
this0sector BEMF0already0detected0
in0this0sector
ZCP

CC62R
CC61R

Last0correct0hall0
event

Figure 3.10.: Hall sensors fail situation

CHAPTER 3. IMPLEMENTATION
36 3.8. TRANSITION FROM HALL TO SENSORLESS MODE

3.8.1. Situation A

This situation can occur if, after a correct hall input, the sensors stop
working without having been done the commutation. If that happens, the
software has to force the commutation, load the next pattern and reconfigure
the timers and registers for the sensorless mode.
The reconfiguration of the registers is the same in all the possible situa-
tions, since it disables the hall sensor mode and enables all the registers as
in compare mode. It also disables the wrong hall event interrupt.
To restart the timer, as the BEMF has not been detected yet, the timer
T12 must be loaded with half of the sector time, which corresponds to the
actual value of zc, since instead of measuring time between ZCP, as the timer
is restarted every commutation, only stores half of the time between ZCP.
Therefore, the timer T12 is restarted with the variable zc plus the actual T12
value.

3.8.2. Situation B

In situation B, the only difference is that the commutation has already been
done, so there is no need to do it. The transition is done by reconfiguring
the registers, loading the next pattern and restarting the timer T12 as in
subsection 3.8.1. Neither in situation A nor in situation B the compare
registers CCU60_61R and CCU60_62R need to be modified, given that as
the BEMF has not been detected yet, the sensorless software control will
change them when it is detected.

3.8.3. Situation C

This stretch is the last stretch where the BEMF has not been detected.
Unlike the previous situations, the commutation and load of the next pattern
has already been done, so the routine only have to reconfigure the registers
and reload the timer T12 as in subsection 3.8.1.

CHAPTER 3. IMPLEMENTATION
3.8. TRANSITION FROM HALL TO SENSORLESS MODE 37

3.8.4. Situation D

The last stretch corresponds to the region where the BEMF has already
been detected, so in this case the compare registers must be reloaded to do
the next commutation and the timer is reset to a different value than in the
previous situation, due to the fact that now it will directly measure the time
between ZCP. In this case, the timer T12 is reset with the actual value of
T12 minus the actual zc value.

3.8.5. Software implementation of the transition

The implementation of the above situations is accomplished inside the

wrong hall event interrupt routine. The listing 3.6 shows the necessary code
to commplete a transition from hall sensor mode to sensorless mode.
if(CCU60_IS & 0x2000) // If CCU60_IS_WHE...
{ // ... wrong hall event detection.
// USER CODE BEGIN (NodeI1,18)
status_hs_error=1; // ... set the hall sensor error status flag.
if(status_bemf) // If bemf detection is possible...
{
bemf_aux = CCU60_T12; // ... save the actual T12 value.
if(slope == 0 || slope == 1) // If bemf not detected yet in
this sector...
{
CCU60_T12MSEL = 0x0111; // ... change hall sensor mode to
sensorless mode.
CCU60_IEN &= 0x97FF; // ... disable wrong hall event or
idle interrupts.
CCU60_TCTR4 = 0x0045; // ... T12 reset and stop.
CCU60_T12 = zc_time + bemf_aux; // ... set the
counter time to the half of the sector time (actual
zc_time).
CCU60_T12PR = (zc_time << 1) + zc_time; // ... set the
time out limit to 1.5 of the sector time.
CCU60_TCTR4 = 0x0040; // ... shadow transfer request for
T12.
CCU60_TCTR4 = 0x0002; // ... start T12.
if(bemf_aux < CCU60_CC61R) // If the commutation was not
done yet...
{
CCU60_MCMOUTS |= 0x0080; // ... force commutation.

CHAPTER 3. IMPLEMENTATION
38 3.8. TRANSITION FROM HALL TO SENSORLESS MODE

}
if(bemf_aux < CCU60_CC62R) // If the update of the pattern
was not done yet...
{
CCU60_MCMOUTS = CCU60_MCMOUTS = SP_Tab_r_l[((CCU60_MCMOUT
& 0x3800) >> 11) + direction]; // ... load next
commutation pattern.
}
}
else // If bemf has already been detected in this sector...
{
CCU60_T12MSEL = 0x0111; // ... change hall sensor mode
to sensorless mode.
CCU60_IEN &= 0x97FF; // ... disable wrong hall event
or idle interrupts.
CCU60_TCTR4 = 0x0045; // ... T12 reset and stop.
CCU60_T12 = bemf_aux - zc_time; // ... set the
counter time to the difference between the actual value
and half the sector time.
CCU60_T12PR = (zc_time << 1) + zc_time; // ... set the
time out limit to 1.5 the sector time.
CCU60_CC60SR = (zc_time << 1); // ... set the CC60R for
speed measurement with the time between zero crossing
points.
CCU60_CC61SR = zc_time - 5; // ... initialize the CCU61
register with half of the time between zero crossings,
// so that it commutates within the required time (with an
advance).
CCU60_CC62SR = CCU60_CC61SR + 10; // ... initialize CCU62
some time after to calculate the speed.
CCU60_TCTR4 = 0x0040; // ... shadow transfer request for
T12.
CCU60_TCTR4 = 0x0002; // ... start T12.
}
}
else // If bemf detection is not possible...
{
Motor_stop(); // ... stop the motor.
}
// USER CODE END
CCU60_ISR |= 0x2000; // ... clear flag CCU60_IS_WHE.
}

Listing 3.6: Transition from hall sensor control to sensorless control

CHAPTER 3. IMPLEMENTATION
3.9. SOFTWARE DESIGN 39

3.9. Software design

This section is intended to explain the basics of the software procedure,
but for a detailed explanation and understanding of the whole process it is
recommended to see appendix A and appendix B attached at the end of the
thesis.
The control process includes the following steps:
1. System clock initialisation and configuration
2. Peripheral initialisation
a) Input/Output ports
b) CCU60 Unit
c) ADCs
3. Register the offset error in the ADC measures
4. Initialise the start variables and the PIs variables
5. Enter the endless loop
6. Reset the trap flag to enable the motor start
7. Start the motor if there is no error or stop it if there is
a) Set the start and end speed reference
b) Initialise the control variables
c) Initialise and enable the CCU60 timers and main registers
8. Increase the speed reference if it is not the desired reference yet
9. Check the overvoltage and overtemperature pins in the endless loop
10. The process is handled by the interrupts
a) Protection interrupts
b) CCU60_CC62R interrupt calculates the speed and loads the next
pattern and applies a dead-time for the BEMF measurement after
commutation
c) T13 triggers measures, computes the duty cycle reference and
modulates the outputs

CHAPTER 3. IMPLEMENTATION
References

[1] Chang-Liang Xia. Permanent magnet brushless DC motor drives and

controls. 1 edition, June 2012.
[2] Padmaraja Yedamale. Brushless DC (BLDC) motor fundamentals. Mi-
crochip Technology Inc, page 20, 2003.
[3] Stepper and BLDC Motors Animation.
http://en.nanotec.com/support/tutorials/stepper-motor-and-
bldc-motors-animation.
[4] BLDC Vector Control (FOC).
http://www.st.com/web/catalog/apps/SE413/AS380/AC901.
[5] Artur Giedymin. Entwurf und Implementierung einer Regelung eines
hochdynamischen Elektroantriebs für eine adaptive Wankstabilisierung
nach ISO 26262. Master’s thesis, Technische Universität Berlin, October
2012.
[6] Ernesto Vázquez-Sánchez José Carlos Gamazo-Real and Jaime Gómez-
Gil. Position and Speed Control of Brushless DC Motors Using Sensor-
less Techniques and Application Trends. Sensors, July 2010.
[7] Satish Rajagopalan. Detection of Rotor and Load Faults in BLDC Mo-
tors Operating Under Stationary and Non-Stationary Conditions. PhD
thesis, Georgia Institute of Technology, Atlanta, GA, USA, August 2006.
[8] P Damodharan and Krishna Vasudevan. Indirect back-emf zero cross-
ing detection for sensorless bldc motor operation. In Power Electronics
and Drives Systems, 2005. PEDS 2005. International Conference on,
volume 2, pages 1107–1111. IEEE, 2005.
[9] Gui-Jia Su and John W McKeever. Low-cost sensorless control of brush-
less dc motors with improved speed range. Power Electronics, IEEE
Transactions on, 19(2):296–302, 2004.
[10] JX Shen and S Iwasaki. Sensorless control of ultrahigh-speed pm brush-

41
42 References

less motor using pll and third harmonic back emf. Industrial Electronics,
IEEE Transactions on, 53(2):421–428, 2006.
[11] JX Shen, ZQ Zhu, and David Howe. Sensorless flux-weakening control of
permanent-magnet brushless machines using third harmonic back emf.
IEEE Transactions on Industry Applications, 40(6):1629–1636, 2004.
[12] Daniel Torres. Sensorless BLDC Motor Control with Back-EMF Filter-
ing Using a Majority Function. Digi-Key Article Library, October 2011.
Contributed By Convergence Promotions LLC.
[13] U. Vinatha, S. Pola, and K.P. Vittal. Recent developments in con-
trol schemes of bldc motors. In Proceedings of the IEEE International
Conference on Industrial Technology (ICIT 2006), Mumbai, India, De-
cember 2006.
[14] Bilal Akin and Manish Bhardwaj. Sensorless Trapezoidal Control of
BLDC Motors. Texas Instruments Incorporated, 655303, Dallas, Texas
75265, July 2013. SPRABQ7.
[15] Infineon Technologies AG, 81726 Munich, Germany. XE166 family Easy
Kit Manual, 2007-06 edition, June 2007.
[16] Infineon Technologies AG, 81726 Munich, Germany. XE167 Datasheet,
2008-08 edition, August 2008.
[17] Infineon DAvE information website. http://www.infineon.com/cms/
en/product/microcontrollers/development-tools,-software-
and-kits/dave-tm-version-2-low-level-code-generation/
channel.html?channel=db3a3043134dde6001134ee4d3b30265.
[18] Infineon Technologies AG, 81726 Munich, Germany. Technical details
on CAPCOM6 module, January 2001. AP1607310.
[19] Infineon Technologies AG, 81726 Munich, Germany. XE166 Derivatives
Volume 2: Peripheral Units, 2008-08 edition, August 2008. User’s Man-
ual, V2.1.
[20] Eduardo Viramontes. BLDC Motor Control with Hall Effect Sensors
Using the 9S08MP. Systems and Applications Engineering Freescale
Technical Support, April 2010. AN4058.
[21] Peter Thomas. How to tune cascade loops. Control Specialists Ltd.,
expertune user conference edition, April 2007.

References
[22] Infineon Technologies AG, 81726 Munich, Germany. DSP Optimization
Guide for XC2000, XE1 6 6 and XC166 Microcontroller Families with
MAC Unit, 2007-10 edition, October 2007. APP1611311.
[23] Infineon Technologies AG, 81726 Munich, Germany. Speed control
of Brushless DC motor with Hall sensor using DAvE Drive for In-
fineon XC164CM/CS microcontrollers, 2007-07-04 edition, July 2007.
AP1611710.
[24] Infineon Technologies AG, 81726 Munich, Germany. Sensorless Control
of BLDC Motor using XE164F Microcontroller, 2010-03 edition, March
20110. AP16173.
[25] Infineon Technologies AG, 81726 Munich, Germany. XE166 Derivatives
Volume 1: System Units, 2008-08 edition, August 2008. User’s Manual,
V2.1.

43
Appendices

45
Appendix A.

Software structure and full code

TM
The software is structured based on the generated files by DAVE project.
The created files are stored in the folder Dave files, whereas the own created
files are in the folder User files.
TM
The header files of the generated DAVE files are not modified, on the
other hand, the source files have been modified in the gaps provided for that
purpose to develop the control application.
The user files are 3:
FUNC.H Includes all the INLINE functions included in the main file.
FUNC.C Contains the global variables and functions created by the user.
CONFIG.H Defines the motor and application parameters needed.
There are other files used, which are provided by Infineon and Keil, such as
the DSP library files (DspLib_Keil.h, FloatTo1Q15.c and the PI function).
The main code files are attached in the following pages:

• MAIN.c in page 48 • CCU60.h in page 90

• MAIN.h in page 54 • FUNC.c in page 92
• ADC0.c in page 57 • FUNC.h in page 103
• ADC0.h in page 65 • CONFIG.h in page 106
• ADC1.c in page 68 • DSP_LibKeil.h in page 107
• ADC1.h in page 75 • FloatTo1Q15.c in page 113
• CCU60.c in page 78 • PI_CTRL.c in page 115

47
48

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.C

//****************************************************************************
// @Module Project Settings
// @Filename MAIN.C
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains the project initialization function.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:06
//
//****************************************************************************

// USER CODE BEGIN (MAIN_General,1)

// USER CODE END

//****************************************************************************
// @Project Includes
//****************************************************************************

#include "MAIN.H"

// USER CODE BEGIN (MAIN_General,2)

#include "FUNC.H"
// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (MAIN_General,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************

// USER CODE BEGIN (MAIN_General,4)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (MAIN_General,5)

// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (MAIN_General,6)

extern BIT status_overvoltemp;

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 49

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.C

// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

// USER CODE BEGIN (MAIN_General,7)

// USER CODE END

//****************************************************************************
// @External Prototypes
//****************************************************************************

// USER CODE BEGIN (MAIN_General,8)

extern void ADC_OFFSET(int CH_OFFSET);
extern void RampUp(void);
// USER CODE END

//****************************************************************************
// @Prototypes Of Local Functions
//****************************************************************************

// USER CODE BEGIN (MAIN_General,9)

// USER CODE END

//****************************************************************************
// @Function void MAIN_vInit(void)
//
//----------------------------------------------------------------------------
// @Description This function initializes the microcontroller.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (Init,1)

// USER CODE END

void MAIN_vInit(void)
{
// USER CODE BEGIN (Init,2)

// USER CODE END

// globally disable interrupts

PSW_IEN = 0;

/// -----------------------------------------------------------------------
/// Configuration of the System Clock:
/// -----------------------------------------------------------------------
/// - VCO clock used, input clock is connected
/// - input frequency is 8,00 MHz
/// - system clock is 66.00 MHz

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

50

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.C

MAIN_vUnlockProtecReg(); // unlock write security

MAIN_vChangeFreq(); // load PLL control register

// -----------------------------------------------------------------------
// Initialization of the Peripherals:
// -----------------------------------------------------------------------

// initializes the Parallel Ports

IO_vInit();

// initializes the Capture / Compare Unit 60 (CCU60)

CCU60_vInit();

// initializes the Analog / Digital Converter (ADC0)

ADC0_vInit();

// initializes the Analog / Digital Converter (ADC1)

ADC1_vInit();

// -----------------------------------------------------------------------
// Initialization of the Bank Select registers:
// -----------------------------------------------------------------------

// USER CODE BEGIN (Init,3)

// USER CODE END

MAIN_vLockProtecReg(); // lock write security

// globally enable interrupts

PSW_IEN = 1;

} // End of function MAIN_vInit

//****************************************************************************
// @Function void MAIN_vUnlockProtecReg(void)
//
//----------------------------------------------------------------------------
// @Description This function makes it possible to write one protected
// register.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (UnlockProtecReg,1)

// USER CODE END

void MAIN_vUnlockProtecReg(void)
{
uword uwPASSWORD;

SCU_SLC = 0xAAAA; // command 0

SCU_SLC = 0x5554; // command 1

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 51

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.C

uwPASSWORD = SCU_SLS & 0x00FF;

uwPASSWORD = (~uwPASSWORD) & 0x00FF;

SCU_SLC = 0x9600 | uwPASSWORD; // command 2

SCU_SLC = 0x0000; // command 3

} // End of function MAIN_vUnlockProtecReg

//****************************************************************************
// @Function void MAIN_vLockProtecReg(void)
//
//----------------------------------------------------------------------------
// @Description This function makes it possible to lock one protected
// register.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (LockProtecReg,1)

// USER CODE END

void MAIN_vLockProtecReg(void)
{
uword uwPASSWORD;

SCU_SLC = 0xAAAA; // command 0

SCU_SLC = 0x5554; // command 1

uwPASSWORD = SCU_SLS & 0x00FF;

uwPASSWORD = (~uwPASSWORD) & 0x00FF;

SCU_SLC = 0x9600 | uwPASSWORD; // command 2

SCU_SLC = 0x1800; // command 3; new PASSWOR is 0x00

uwPASSWORD = SCU_SLS & 0x00FF;

uwPASSWORD = (~uwPASSWORD) & 0x00FF;
SCU_SLC = 0x8E00 | uwPASSWORD; // command 4

} // End of function MAIN_vLockProtecReg

//****************************************************************************
// @Function void MAIN_vChangeFreq(void)
//
//----------------------------------------------------------------------------
// @Description This function is used to select the external crystal and
// configure the system frequency to 80Mhz/66Mhz.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

Page: 4

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

52

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.C

// USER CODE BEGIN (ChangeFreq,1)

// USER CODE END

void MAIN_vChangeFreq(void)
{
SCS_SwitchToHighPrecBandgap();

//For application and internal application resets, the complete PLL configuration cou
ld be avoided
//The entry from application resets and internal application reset is covered in the
following differentiation
//in int/ext clock in lock/unlocked state.

if ((SCU_PLLSTAT & 0x0004) == 0x0004) // fR derived from Internal clock

{
//Normal startup state during boot and the clock
//has to be in the next step configured on the external crystal
//use XTAL/VCO, count XTAL clock

SCS_StartXtalOsc(1); // Starts the crystal oscillator

SCS_SwitchSystemClock(1); // System clock is increased to ta
rget speed (80/66 MHz)

else // fR derived from external crysta

l clock
{
if ((SCU_PLLSTAT & 0x1009) == 0x1009) // fR derived from external crysta
l clock + VCO is locked
{
//usually after an application reset where clock need not be configured
again.
//check K2/P/N values and decide whether these values have to be adapted
based on application needs.
NOP();
//usually the PLL losss of Lock TRAP schould be enabled here.
}
else //fR derived from external crystal
clock + VCO is not locked

{
//estimate the K1 value and the current frequency
//reduce K2/P/N values in steps so that the frequency
//jumps is limited to 20MHz or factor of 5 whichever is minimum
NOP();
}

} // End of function MAIN_vChangeFreq

//****************************************************************************
// @Function void main(void)
//
//----------------------------------------------------------------------------
// @Description This is the main function.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//

Page: 5

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 53

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.C

//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (Main,1)

// USER CODE END

void main(void)
{
// USER CODE BEGIN (Main,2)

// USER CODE END

MAIN_vInit();

// USER CODE BEGIN (Main,3)

ADC_OFFSET(1); // Calculate the offset for phase A voltage.
ADC_OFFSET(4); // Calculate the offset for phase B voltage.
ADC_OFFSET(7); // Calculate the offset for phase C voltage.
ADC_OFFSET(3); // Calculate the offset for DC link current.
ADC_OFFSET(2); // Calculate the offset for DC link voltage.
INIT_CTRL_VAR(); // Initialise the main variables for start-up.
// USER CODE END

while(1)
{

// USER CODE BEGIN (Main,4)

TRAP_RESET(); // Reset the software trap flag if requested by
the user.
MOTOR_OPERATION(); // Start or Stop the Motor as per the present c
onditions.
RampUp(); // Ramp up(or down) the motor to the new speed.
if(P5_IN & 0x4800) // If overvoltage or overtempreature detected..
.
{
status_overvoltemp = 1; // ... set the overtemperature and overvolt
age flag.
Motor_stop(); // ... stop the motor.
}
// USER CODE END

} // End of function main

// USER CODE BEGIN (MAIN_General,10)

// USER CODE END

Page: 6

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

54

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.H

//****************************************************************************
// @Module Project Settings
// @Filename MAIN.H
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains all function prototypes and macros for
// the MAIN module.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:06
//
//****************************************************************************

// USER CODE BEGIN (MAIN_Header,1)

// USER CODE END

#ifndef _MAIN_H_
#define _MAIN_H_

//****************************************************************************
// @Project Includes
//****************************************************************************

// USER CODE BEGIN (MAIN_Header,2)

#include <mac.h>
// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (MAIN_Header,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************
#define KEIL

#define SEG(func) (unsigned int)(((unsigned long)((void (far*)(void))func) >> 16))

#define SOF(func) (unsigned int)(((void (far *) (void))func))

// USER CODE BEGIN (MAIN_Header,4)

#define C_BDATA int bdata

#define C_SBIT sbit
#define BIT volatile bit
#define C_SDATA sdata

#define INLINE static __inline

#define C_ATBIT(var, bit_pos) =(var)^(bit_pos)

#define C_LSHIFT(var, pos) (var)

#define C_RSHIFT(var, pos) (var)

#define C_MAC(icode, var1, var2) _mac_((icode), (var1), (var2))

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 55

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.H

#define C_MACL(icode, var) _macl_((icode), (var))

#define C_CoLOAD(var) C_MACL(CoLOAD, (var))

#define C_CoLOAD_0_1(var1, var2) C_MAC(CoLOAD, 0, (var2))
#define C_CoLOAD_1_0(var1, var2) C_MAC(CoLOAD, (var1), 0)

#define C_CoMAX(var) C_MACL(CoMAX, (var))

#define C_CoMAX_0_1(var1, var2) C_MAC(CoMAX, 0, (var2))
#define C_CoMAX_1_0(var1, var2) C_MAC(CoMAX, (var1), 0)

#define C_CoMIN(var) C_MACL(CoMIN, (var))

#define C_CoMIN_0_1(var1, var2) C_MAC(CoMIN, 0, (var2))
#define C_CoMIN_1_0(var1, var2) C_MAC(CoMIN, (var1), 0)

#define C_CoSUB(var) C_MACL(CoSUB, (var))

#define C_CoSUB_0_1(var1, var2) C_MAC(CoSUB, 0, (var2))
#define C_CoSUB_1_0(var1, var2) C_MAC(CoSUB, (var1), 0)

#define C_CoMULsu(var1, var2) _mac_(CoMULsu, (var1), (var2))

#define C_CoMULu(var1, var2) _mac_(CoMULu, (var1), (var2))
#define C_CoMUL(var1, var2) _mac_(CoMUL, (var1), (var2))
#define C_CoSTORE() _lmac_()
#define C_CoABS() _macv_(CoABSacc)
#define C_CoSTOREMAS() _imac_sat_()
#define C_CoSTOREMAS_MAL() _lmac_sat_()

#define C_CoMACsu(var1, var2) _mac_(CoMACsu, var1, var2)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (MAIN_Header,5)

struct PI_cont // PI-controller parameter

{
long yn; // PI integral buffer
int kp; // Proportional constant

int ki; // Integral constant

int ymax; // Maximal ouput limit
int ymin; // Minimal ouput limit
};
// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (MAIN_Header,6)

// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

// USER CODE BEGIN (MAIN_Header,7)

// USER CODE END

//****************************************************************************
// @Prototypes Of Global Functions

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

56

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\MAIN.H

//****************************************************************************

void MAIN_vUnlockProtecReg(void);

void MAIN_vLockProtecReg(void);

void MAIN_vChangeFreq(void);

// USER CODE BEGIN (MAIN_Header,8)

// USER CODE END

//****************************************************************************
// @Interrupt Vectors
//****************************************************************************

// USER CODE BEGIN (MAIN_Header,9)

// USER CODE END

//****************************************************************************
// @Project Includes
//****************************************************************************

#include <Intrins.h>

#include "XE16xREGS.H"
#include "SCS.H"

#include "IO.H"
#include "CCU60.H"
#include "ADC0.H"
#include "ADC1.H"

// USER CODE BEGIN (MAIN_Header,10)

#include "DspLib_Keil.h"
#include "CONFIG.H"
// USER CODE END

#endif // ifndef _MAIN_H_

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 57

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

//****************************************************************************
// @Module Analog / Digital Converter (ADC0)
// @Filename ADC0.C
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains functions that use the ADC0 module.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:07
//
//****************************************************************************

// USER CODE BEGIN (ADC0_General,1)

// USER CODE END

//****************************************************************************
// @Project Includes
//****************************************************************************

#include "MAIN.H"

// USER CODE BEGIN (ADC0_General,2)

// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (ADC0_General,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************

// USER CODE BEGIN (ADC0_General,4)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (ADC0_General,5)

// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (ADC0_General,6)

extern BIT status_overcurrent; // Overcurrent detection

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

58

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

extern int C_SDATA oc_counter; // Counter to specify the time limi

t for motor Overcurrent
// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

// USER CODE BEGIN (ADC0_General,7)

// USER CODE END

//****************************************************************************
// @External Prototypes
//****************************************************************************

// USER CODE BEGIN (ADC0_General,8)

extern void Motor_stop(void);
// USER CODE END

//****************************************************************************
// @Prototypes Of Local Functions
//****************************************************************************

// USER CODE BEGIN (ADC0_General,9)

// USER CODE END

//****************************************************************************
// @Function void ADC0_vInit(void)
//
//----------------------------------------------------------------------------
// @Description This is the initialization function of the ADC function
// library. It is assumed that the SFRs used by this library
// are in reset state.
//
// Following SFR fields will be initialized:
// GLOBCTR - Global Control
// RSPR0 - Priority and Arbitration Register
// ASENR - Arbitration slot enable register
// CHCTRx - Channel Control Register x
// RCRx - Result Control Register x
// KSCFG - Module configuration Register
// INPCR - Input class Registers
// CHINPRx - Channel Interrupt register
// EVINPRx - Event Interrupt register
// SYNCTR - Synchronisation control register
// LCBRx - Limit check boundary register
// PISEL - Port input selection
// QMR0 - Sequential 0 mode register
// CRMR1 - Parallel mode register
// QMR2 - Sequential 2 mode register
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 59

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

// USER CODE BEGIN (ADC0_Init,1)

// USER CODE END

void ADC0_vInit(void)
{
// USER CODE BEGIN (ADC0_Init,2)

// USER CODE END

/// -----------------------------------------------------------------------
/// Configuration of ADC0 kernel configuration register:
/// -----------------------------------------------------------------------
ADC0_KSCFG = 0x0003; // load ADC0 kernel configuration register

/// - the ADC module clock is enabled

/// - the ADC module clock = 66,00 MHz
///

_nop_(); // one cycle delay

_nop_(); // one cycle delay

/// -----------------------------------------------------------------------
/// Configure global control register:
/// -----------------------------------------------------------------------
/// --- Conversion Timing -----------------
/// - conversion time (CTC) = 01,29 us

/// _Analog clock is 1/5th of module clock and digital clock is 1/1 times
/// of module clock

/// - the permanent arbitration mode is selected

ADC0_GLOBCTR = 0x0004; // load global control register

/// -----------------------------------------------------------------------
/// Configuration of Arbitration Slot enable register and also the Source
/// Priority register:
/// -----------------------------------------------------------------------
/// - Arbitration Slot 0 is disabled

/// - Arbitration Slot 1 is enabled

/// - Arbitration Slot 2 is disabled

/// - the priority of request source 0 is low

/// - the wait-for-start mode is selected for source 0
/// - the priority of request source 1 is high
/// - the wait-for-start mode is selected for source 1
/// - the priority of request source 2 is low
/// - the wait-for-start mode is selected for source 2
ADC0_ASENR = 0x0002; // load Arbitration Slot enable register

ADC0_RSPR0 = 0x0032; // load Priority and Arbitration register

/// -----------------------------------------------------------------------
/// Configuration of Channel Control Registers:
/// -----------------------------------------------------------------------
/// Configuration of Channel 3
/// - the result register0 is selected
/// - the limit check 7 is selected

/// - the reference voltage selected is Standard Voltage (Varef)

/// - the input class selected is Input Class 0

/// - LCBR0 is selected as upper boundary

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

60

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

/// - LCBR3 is selected as lower boundary

ADC0_CHCTR3 = 0x007C; // load channel control register

/// -----------------------------------------------------------------------
/// Configuration of Sample Time and Resolution:
/// -----------------------------------------------------------------------

/// Sample phase is extended by 48 Analog clock cycles

/// 10 bit resolution selected

ADC0_INPCR0 = 0x0030; // load input class0 register

/// 10 bit resolution selected

ADC0_INPCR1 = 0x0000; // load input class1 register

/// -----------------------------------------------------------------------
/// Configuration of Result Control Registers:
/// -----------------------------------------------------------------------
/// Configuration of Result Control Register 0
/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR0 = 0x0000; // load result control register 0

/// Configuration of Result Control Register 1

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR1 = 0x0000; // load result control register 1

/// Configuration of Result Control Register 2

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR2 = 0x0000; // load result control register 2

/// Configuration of Result Control Register 3

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR3 = 0x0000; // load result control register 3

/// Configuration of Result Control Register 4

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR4 = 0x0000; // load result control register 4

/// Configuration of Result Control Register 5

/// - the data reduction filter is disabled

Page: 4

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 61

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

/// - the event interrupt is disabled

/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR5 = 0x0000; // load result control register 5

/// Configuration of Result Control Register 6

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR6 = 0x0000; // load result control register 6

/// Configuration of Result Control Register 7

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC0_RCR7 = 0x0000; // load result control register 7

/// -----------------------------------------------------------------------
/// Configuration of Channel Interrupt Node Pointer Register:
/// -----------------------------------------------------------------------
/// - the SR0 line become activated if channel 3 interrupt is generated

ADC0_CHINPR0 = 0x0000; // load channel interrupt node pointer

// register

ADC0_CHINPR4 = 0x0000; // load channel interrupt node pointer

// register

ADC0_CHINPR8 = 0x0000; // load channel interrupt node pointer

// register

ADC0_CHINPR12 = 0x0000; // load channel interrupt node pointer

// register

/// -----------------------------------------------------------------------
/// Configuration of Event Interrupt Node Pointer Register for Source
/// Interrupts:
/// -----------------------------------------------------------------------
/// - the SR 0 line become activated if the event 1 interrupt is generated

ADC0_EVINPR0 = 0x0000; // load event interrupt set flag register

/// -----------------------------------------------------------------------
/// Configuration of Event Interrupt Node Pointer Register for Result
/// Interrupts:
/// -----------------------------------------------------------------------

ADC0_EVINPR8 = 0x0000; // load event interrupt set flag register

ADC0_EVINPR12 = 0x0000; // load event interrupt set flag register

/// -----------------------------------------------------------------------
/// Configuration of Service Request Nodes 0 - 3 :
/// -----------------------------------------------------------------------
/// SRN0 service request node configuration:
/// - SRN0 interrupt priority level (ILVL) = 7
/// - SRN0 interrupt group level (GLVL) = 0
/// - SRN0 group priority extension (GPX) = 0

Page: 5

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

62

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

ADC_0IC = 0x005C;

/// -----------------------------------------------------------------------
/// Configuration of Limit Check Boundary:
/// -----------------------------------------------------------------------

ADC0_LCBR0 = 0x0198; // load limit check boundary register 0

ADC0_LCBR1 = 0x0E64; // load limit check boundary register 1

ADC0_LCBR2 = 0x0540; // load limit check boundary register 2

ADC0_LCBR3 = 0x07C8; // load limit check boundary register 3

/// -----------------------------------------------------------------------
/// Configuration of Gating source and External Trigger Control:
/// -----------------------------------------------------------------------
/// - No Gating source selected for Arbitration Source 0

/// - the trigger input ETR00 is selected for Source 0

/// - No Gating source selected for Arbitration Source 1

/// - the trigger input ETR00 is selected for Source 1

/// - No Gating source selected for Arbitration Source 2

/// - the trigger input ETR00 is selected for Source 1

ADC0_PISEL = 0x0444; // load external trigger control register

/// -----------------------------------------------------------------------
/// Configuration of Conversion Queue Mode Register:Sequential Source 0
/// -----------------------------------------------------------------------
/// - the gating line is permanently Disabled
/// - the external trigger is disabled
/// - the trigger mode 0 is selected

ADC0_QMR0 = 0x0000; // load queue mode register

/// -----------------------------------------------------------------------
/// Configuration of Conversion Queue Mode Register:Sequential Source 2
/// -----------------------------------------------------------------------
/// - the gating line is permanently Disabled
/// - the external trigger is disabled
/// - the trigger mode 0 is selected

ADC0_QMR2 = 0x0000; // load queue mode register

/// -----------------------------------------------------------------------
/// Configuration of Conversion Request Mode Registers:Parallel Source
/// -----------------------------------------------------------------------
/// - the gating line is permanently Enabled
/// - the external trigger is disabled
/// - the source interrupt is disabled
/// - the autoscan functionality is disabled

ADC0_CRMR1 = 0x0001; // load conversion request mode register 1

/// -----------------------------------------------------------------------
/// Configuration of Synchronisation Registers:
/// -----------------------------------------------------------------------
/// - ADC0 is master
ADC0_SYNCTR |= 0x0010; // Synchronisation register

P5_DIDIS = 0x0008; // Port 5 Digital input disable register

Page: 6

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 63

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

ADC0_GLOBCTR |= 0x0300; // turn on Analog part

// USER CODE BEGIN (ADC0_Init,3)

// USER CODE END

} // End of function ADC0_vInit

//****************************************************************************
// @Function void ADC0_viSRN0(void)
//
//----------------------------------------------------------------------------
// @Description This is the interrupt service routine for the Service
// Request Node 0 of the ADC0 module.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (ADC0_viSRN0,0)

// USER CODE END

void ADC0_viSRN0(void) interrupt ADC0_SRN0INT

{

if((ADC0_CHINFR & 0x0008) == 0x0008) //Channel3 interrupt

{
ADC0_CHINCR = 0x0008; // Clear Channel3 interrupt

// USER CODE BEGIN (ADC0_viSRN0,7)

if (((ADC1_RESR0>>2)&0x03FF) < 682) // motor instantane
ous current below shutdown
{ // current limi
t ?
oc_counter++; // if no, incre
ment s/w counter.
if (oc_counter == TIME_CURRENT_MAX) // s/w counter
reached the specified limit ?
{
status_overcurrent = 1; // if yes, stop
the motor
Motor_stop();
}
}
else // if motor 'cu
rrent' sample is higher
{ // than the spe
cified limit (10 A), stop the motor
status_overcurrent = 1;
Motor_stop();
}
// USER CODE END

} // End of function ADC0_viSRN0

Page: 7

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

64

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.C

// USER CODE BEGIN (ADC0_General,10)

// USER CODE END

Page: 8

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 65

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.H

//****************************************************************************
// @Module Analog / Digital Converter (ADC0)
// @Filename ADC0.H
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains all function prototypes and macros for
// the ADC0 module.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:07
//
//****************************************************************************

// USER CODE BEGIN (ADC0_Header,1)

// USER CODE END

#ifndef _ADC0_H_

#define _ADC0_H_

//****************************************************************************
// @Project Includes
//****************************************************************************

// USER CODE BEGIN (ADC0_Header,2)

// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (ADC0_Header,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************

// This parameter identifies ADC0 analog channel 0

#define ADC0_ANA_0 0

// This parameter identifies ADC0 analog channel 1

#define ADC0_ANA_1 1

// This parameter identifies ADC0 analog channel 2

#define ADC0_ANA_2 2

// This parameter identifies ADC0 analog channel 3

#define ADC0_ANA_3 3

// This parameter identifies ADC0 analog channel 4

#define ADC0_ANA_4 4

// This parameter identifies ADC0 analog channel 5

#define ADC0_ANA_5 5

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

66

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.H

// This parameter identifies ADC0 analog channel 6

#define ADC0_ANA_6 6

// This parameter identifies ADC0 analog channel 7

#define ADC0_ANA_7 7

// This parameter identifies ADC0 analog channel 8

#define ADC0_ANA_8 8

// This parameter identifies ADC0 analog channel 9

#define ADC0_ANA_9 9

// This parameter identifies ADC0 analog channel 10

#define ADC0_ANA_10 10

// This parameter identifies ADC0 analog channel 11

#define ADC0_ANA_11 11

// This parameter identifies ADC0 analog channel 12

#define ADC0_ANA_12 12

// This parameter identifies ADC0 analog channel 13

#define ADC0_ANA_13 13

// This parameter identifies ADC0 analog channel 14

#define ADC0_ANA_14 14

// This parameter identifies ADC0 analog channel 15

#define ADC0_ANA_15 15

// This parameter identifies ADC0 -Source 0

#define ADC0_SOURCE_0 0

// This parameter identifies ADC0 -Source 1

#define ADC0_SOURCE_1 1

// This parameter identifies ADC0 -Source 2

#define ADC0_SOURCE_2 2

// USER CODE BEGIN (ADC0_Header,4)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (ADC0_Header,5)

// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (ADC0_Header,6)

// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 67

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC0.H

// USER CODE BEGIN (ADC0_Header,7)

// USER CODE END

//****************************************************************************
// @Prototypes Of Global Functions
//****************************************************************************

void ADC0_vInit(void);

// USER CODE BEGIN (ADC0_Header,8)

// USER CODE END

//****************************************************************************
// @Interrupt Vectors
//****************************************************************************

#define ADC0_SRN0INT 0x28

// USER CODE BEGIN (ADC0_Header,9)

// USER CODE END

#endif // ifndef _ADC0_H_

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

68

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.C

//****************************************************************************
// @Module Analog / Digital Converter (ADC1)
// @Filename ADC1.C
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains functions that use the ADC1 module.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:07
//
//****************************************************************************

// USER CODE BEGIN (ADC1_General,1)

// USER CODE END

//****************************************************************************
// @Project Includes
//****************************************************************************

#include "MAIN.H"

// USER CODE BEGIN (ADC1_General,2)

// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (ADC1_General,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************

// USER CODE BEGIN (ADC1_General,4)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (ADC1_General,5)

// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (ADC1_General,6)

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 69

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.C

// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

// USER CODE BEGIN (ADC1_General,7)

// USER CODE END

//****************************************************************************
// @External Prototypes
//****************************************************************************

// USER CODE BEGIN (ADC1_General,8)

// USER CODE END

//****************************************************************************
// @Prototypes Of Local Functions
//****************************************************************************

// USER CODE BEGIN (ADC1_General,9)

// USER CODE END

//****************************************************************************
// @Function void ADC1_vInit(void)
//
//----------------------------------------------------------------------------
// @Description This is the initialization function of the ADC function
// library. It is assumed that the SFRs used by this library
// are in reset state.
//
// Following SFR fields will be initialized:
// GLOBCTR - Global Control
// RSPR0 - Priority and Arbitration Register
// ASENR - Arbitration slot enable register
// CHCTRx - Channel Control Register x
// RCRx - Result Control Register x
// KSCFG - Module configuration Register
// INPCR - Input class Registers
// CHINPRx - Channel Interrupt register
// EVINPRx - Event Interrupt register
// SYNCTR - Synchronisation control register
// LCBRx - Limit check boundary register
// PISEL - Port input selection
// QMR0 - Sequential 0 mode register
// CRMR1 - Parallel mode register
// QMR2 - Sequential 2 mode register
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (ADC1_Init,1)

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

70

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.C

// USER CODE END

void ADC1_vInit(void)
{
// USER CODE BEGIN (ADC1_Init,2)

// USER CODE END

/// -----------------------------------------------------------------------
/// Configuration of ADC0 kernel configuration register:
/// -----------------------------------------------------------------------
ADC0_KSCFG = 0x0003; // load ADC0 kernel configuration register

/// - the ADC module clock is enabled

/// - the ADC module clock = 66,00 MHz
///

_nop_(); // one cycle delay

_nop_(); // one cycle delay

/// -----------------------------------------------------------------------
/// Configure global control register:
/// -----------------------------------------------------------------------
/// --- Conversion Timing -----------------
/// - conversion time (CTC) = 01,29 us

/// _Analog clock is 1/5th of module clock and digital clock is 1/1 times
/// of module clock

/// - the permanent arbitration mode is selected

ADC1_GLOBCTR = 0x0004; // load global control register

/// -----------------------------------------------------------------------
/// Configuration of Arbitration Slot enable register and also the Source
/// Priority register:
/// -----------------------------------------------------------------------
/// - Arbitration Slot 0 is enabled

/// - Arbitration Slot 1 is enabled

/// - Arbitration Slot 2 is disabled

/// - the priority of request source 0 is 2

/// - the wait-for-start mode is selected for source 0
/// - the priority of request source 1 is high
/// - the wait-for-start mode is selected for source 1
/// - the priority of request source 2 is low
/// - the wait-for-start mode is selected for source 2
ADC1_ASENR = 0x0003; // load Arbitration Slot enable register

ADC1_RSPR0 = 0x0032; // load Priority and Arbitration register

/// -----------------------------------------------------------------------
/// Configuration of Channel Control Registers:
/// -----------------------------------------------------------------------
/// Configuration of Channel 1
/// - the result register0 is selected
/// - the limit check 0 is selected

/// - the reference voltage selected is Standard Voltage (Varef)

/// - the input class selected is Input Class 0

/// - LCBR0 is selected as upper boundary

/// - LCBR1 is selected as lower boundary

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 71

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.C

ADC1_CHCTR1 = 0x0004; // load channel control register

/// Configuration of Channel 2

/// - the result register1 is selected
/// - the limit check 0 is selected

/// - the reference voltage selected is Standard Voltage (Varef)

/// - the input class selected is Input Class 0

/// - LCBR0 is selected as upper boundary

/// - LCBR1 is selected as lower boundary

ADC1_CHCTR2 = 0x1004; // load channel control register

/// Configuration of Channel 4

/// - the result register0 is selected
/// - the limit check 0 is selected

/// - the reference voltage selected is Standard Voltage (Varef)

/// - the input class selected is Input Class 0

/// - LCBR0 is selected as upper boundary

/// - LCBR1 is selected as lower boundary

ADC1_CHCTR4 = 0x0004; // load channel control register

/// Configuration of Channel 7

/// - the result register0 is selected
/// - the limit check 0 is selected

/// - the reference voltage selected is Standard Voltage (Varef)

/// - the input class selected is Input Class 0

/// - LCBR0 is selected as upper boundary

/// - LCBR1 is selected as lower boundary

ADC1_CHCTR7 = 0x0004; // load channel control register

/// -----------------------------------------------------------------------
/// Configuration of Sample Time and Resolution:
/// -----------------------------------------------------------------------

/// 10 bit resolution selected

ADC1_INPCR0 = 0x0000; // load input class0 register

/// 10 bit resolution selected

ADC1_INPCR1 = 0x0000; // load input class1 register

/// -----------------------------------------------------------------------
/// Configuration of Result Control Registers:
/// -----------------------------------------------------------------------
/// Configuration of Result Control Register 0
/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR0 = 0x0000; // load result control register 0

Page: 4

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

72

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.C

/// Configuration of Result Control Register 1

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR1 = 0x0000; // load result control register 1

/// Configuration of Result Control Register 2

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR2 = 0x0000; // load result control register 2

/// Configuration of Result Control Register 3

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR3 = 0x0000; // load result control register 3

/// Configuration of Result Control Register 4

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR4 = 0x0000; // load result control register 4

/// Configuration of Result Control Register 5

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR5 = 0x0000; // load result control register 5

/// Configuration of Result Control Register 6

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR6 = 0x0000; // load result control register 6

/// Configuration of Result Control Register 7

/// - the data reduction filter is disabled
/// - the event interrupt is disabled
/// - the wait-for-read mode is disabled

/// - the FIFO functionality is disabled

ADC1_RCR7 = 0x0000; // load result control register 7

/// -----------------------------------------------------------------------
/// Configuration of Channel Interrupt Node Pointer Register:
/// -----------------------------------------------------------------------
/// - the SR0 line become activated if channel 1 interrupt is generated

Page: 5

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 73

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.C

/// - the SR0 line become activated if channel 2 interrupt is generated

ADC1_CHINPR0 = 0x0000; // load channel interrupt node pointer

// register

/// - the SR0 line become activated if channel 4 interrupt is generated

/// - the SR0 line become activated if channel 7 interrupt is generated

ADC1_CHINPR4 = 0x0000; // load channel interrupt node pointer

// register

/// -----------------------------------------------------------------------
/// Configuration of Event Interrupt Node Pointer Register for Source
/// Interrupts:
/// -----------------------------------------------------------------------
/// - the SR 0 line become activated if the event 0 interrupt is generated
/// - the SR 0 line become activated if the event 1 interrupt is generated

ADC1_EVINPR0 = 0x0000; // load event interrupt set flag register

/// -----------------------------------------------------------------------
/// Configuration of Event Interrupt Node Pointer Register for Result
/// Interrupts:
/// -----------------------------------------------------------------------

ADC1_EVINPR8 = 0x0000; // load event interrupt set flag register

ADC1_EVINPR12 = 0x0000; // load event interrupt set flag register

/// -----------------------------------------------------------------------
/// Configuration of Service Request Nodes 0 - 3 :
/// -----------------------------------------------------------------------

/// -----------------------------------------------------------------------
/// Configuration of Limit Check Boundary:
/// -----------------------------------------------------------------------

ADC1_LCBR0 = 0x0198; // load limit check boundary register 0

ADC1_LCBR1 = 0x0E64; // load limit check boundary register 1

ADC1_LCBR2 = 0x0554; // load limit check boundary register 2

ADC1_LCBR3 = 0x0AA8; // load limit check boundary register 3

/// -----------------------------------------------------------------------
/// Configuration of Gating source and External Trigger Control:
/// -----------------------------------------------------------------------
/// - No Gating source selected for Arbitration Source 0

/// - the trigger input ETR01 is selected for Source 0

/// - No Gating source selected for Arbitration Source 1

/// - the trigger input ETR00 is selected for Source 1

/// - No Gating source selected for Arbitration Source 2

/// - the trigger input ETR00 is selected for Source 1

ADC1_PISEL = 0x0444; // load external trigger control register

/// -----------------------------------------------------------------------
/// Configuration of Conversion Queue Mode Register:Sequential Source 0
/// -----------------------------------------------------------------------
/// - the gating line is permanently Enabled

Page: 6

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

74

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.C

/// - the external trigger is enabled

ADC1_QMR0 = 0x0005; // load queue mode register

/// -----------------------------------------------------------------------
/// Configuration of Conversion Queue Mode Register:Sequential Source 2
/// -----------------------------------------------------------------------
/// - the gating line is permanently Disabled
/// - the external trigger is disabled
/// - the trigger mode 0 is selected

ADC1_QMR2 = 0x0000; // load queue mode register

/// -----------------------------------------------------------------------
/// Configuration of Conversion Request Mode Registers:Parallel Source
/// -----------------------------------------------------------------------
/// - the gating line is permanently Enabled
/// - the external trigger is disabled
/// - the source interrupt is disabled
/// - the autoscan functionality is disabled

ADC1_CRMR1 = 0x0001; // load conversion request mode register 1

/// -----------------------------------------------------------------------
/// Configuration of Synchronisation Registers:
/// -----------------------------------------------------------------------
/// - ADC1 is master
ADC1_SYNCTR |= 0x0010; // Synchronisation register

P15_DIDIS = 0x0096; // Port 15 Digital input disable register

ADC1_GLOBCTR |= 0x0300; // turn on Analog part

// USER CODE BEGIN (ADC1_Init,3)

// USER CODE END

} // End of function ADC1_vInit

// USER CODE BEGIN (ADC1_General,10)

// USER CODE END

Page: 7

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 75

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.H

//****************************************************************************
// @Module Analog / Digital Converter (ADC1)
// @Filename ADC1.H
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains all function prototypes and macros for
// the ADC1 module.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:07
//
//****************************************************************************

// USER CODE BEGIN (ADC1_Header,1)

// USER CODE END

#ifndef _ADC1_H_

#define _ADC1_H_

//****************************************************************************
// @Project Includes
//****************************************************************************

// USER CODE BEGIN (ADC1_Header,2)

// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (ADC1_Header,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************

// This parameter identifies ADC1 analog channel 0

#define ADC1_ANA_0 0

// This parameter identifies ADC1 analog channel 1

#define ADC1_ANA_1 1

// This parameter identifies ADC1 analog channel 2

#define ADC1_ANA_2 2

// This parameter identifies ADC1 analog channel 3

#define ADC1_ANA_3 3

// This parameter identifies ADC1 analog channel 4

#define ADC1_ANA_4 4

// This parameter identifies ADC1 analog channel 5

#define ADC1_ANA_5 5

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

76

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.H

// This parameter identifies ADC1 analog channel 6

#define ADC1_ANA_6 6

// This parameter identifies ADC1 analog channel 7

#define ADC1_ANA_7 7

// This parameter identifies ADC1 -Source 0

#define ADC1_SOURCE_0 0

// This parameter identifies ADC1 -Source 1

#define ADC1_SOURCE_1 1

// This parameter identifies ADC1 -Source 2

#define ADC1_SOURCE_2 2

// USER CODE BEGIN (ADC1_Header,4)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (ADC1_Header,5)

// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (ADC1_Header,6)

// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

// USER CODE BEGIN (ADC1_Header,7)

// USER CODE END

//****************************************************************************
// @Prototypes Of Global Functions
//****************************************************************************

void ADC1_vInit(void);

// USER CODE BEGIN (ADC1_Header,8)

// USER CODE END

//****************************************************************************
// @Interrupt Vectors
//****************************************************************************

// USER CODE BEGIN (ADC1_Header,9)

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 77

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\ADC1.H

// USER CODE END

#endif // ifndef _ADC1_H_

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

78

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

//****************************************************************************
// @Module Capture / Compare Unit 60 (CCU60)
// @Filename CCU60.C
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains functions that use the CCU60 module.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:07
//
//****************************************************************************

// USER CODE BEGIN (CCU60_General,1)

// USER CODE END

//****************************************************************************
// @Project Includes
//****************************************************************************

#include "MAIN.H"

// USER CODE BEGIN (CCU60_General,2)

// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (CCU60_General,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************

// USER CODE BEGIN (CCU60_General,4)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (CCU60_General,5)

// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (CCU60_General,6)

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 79

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

// ADC offset calculation

extern long C_SDATA adc_offset_acumm; // Accumulator for averagin
g of ADC offset.
extern int C_SDATA adc_offset_resA; // Computed ADC offset for
bemf A.
extern int C_SDATA adc_offset_resB; // Computed ADC offset for
bemf B.
extern int C_SDATA adc_offset_resC; // Computed ADC offset for
bemf C.
extern int C_SDATA adc_offset_resVdc; // Computed ADC offset for
DC link voltage.
extern int C_SDATA adc_offset_resIdc; // Computed ADC offset for
DC link current.

extern BIT status_overcurrent; // Overcurrent detection.

extern BIT status_hs_error; // Wrong hall event.
extern BIT status_time_out; // Time_out detection.
extern BIT status_ctrap; // Ctrap detection.
extern BIT status_bemf; // Bemf detection.
extern unsigned int C_SDATA speed_meas; // Measured speed.
extern int C_SDATA hall_counter; // Counter for a valid speed measur
ement.
extern int const SP_Tab_r_l[]; // Hall pattern.
extern int C_SDATA direction; // Index-offset for reading the swi
tching pattern, depending on the specified motor direction.
extern int C_SDATA PI_spd_out; // PI speed output.
extern int C_SDATA PI_curr_out; // PI current output.
extern int spd_loop; // Speed closed loop prescaler.
extern struct PI_cont C_SDATA PI_spd; // PI speed variable.
extern struct PI_cont C_SDATA PI_curr; // PI current variable.
extern unsigned int C_SDATA speed_meas; // Measured speed.
extern unsigned int C_SDATA speed_ref; // Desired motor speed reference.
extern int C_SDATA Idc_meas; // DC link current measurement.
extern unsigned int C_SDATA speed_meas; // Variable to store the actual mot
or speed (sector).
extern int C_SDATA increment_free; // Flag to specify that T13 interru
pt has occured.
extern unsigned int C_SDATA bemf_aux; // Stores the current value of T12
to change to sensorless mode.
extern int C_SDATA slope; // Slope of the bemf: Rising = 0, F
alling = 1.
extern unsigned int C_SDATA zc_time; // Time between zero crossing point
s in sensorless mode
extern unsigned int C_SDATA compare_value; // DC link voltage/2

// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

// USER CODE BEGIN (CCU60_General,7)

int C_SDATA speed_ref_pu; // Reference speed in pu.
int C_SDATA speed_meas_pu; // Actual speed in pu.
// USER CODE END

//****************************************************************************
// @External Prototypes
//****************************************************************************

// USER CODE BEGIN (CCU60_General,8)

extern void Motor_stop(void);
extern int Speed_calc(unsigned int time_meas, int* speed_counter, int limit);
extern int pi_controller64(long *pi_parameter,int reference,int actual);
extern void Sensorless_calc(void);

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

80

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

// USER CODE END

//****************************************************************************
// @Prototypes Of Local Functions
//****************************************************************************

// USER CODE BEGIN (CCU60_General,9)

// USER CODE END

//****************************************************************************
// @Function void CCU60_vInit(void)
//
//----------------------------------------------------------------------------
// @Description This is the initialization function of the CCU60 function
// library. It is assumed that the SFRs used by this library
// are in reset state.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (Init,1)

// USER CODE END

void CCU60_vInit(void)
{
// USER CODE BEGIN (Init,2)

// USER CODE END

/// -----------------------------------------------------------------------
/// Configuration of KERNEL REGISTERS :
/// -----------------------------------------------------------------------
/// - CCU60 Module is enabled.
/// - The CCU60 module clock = 66,000 MHz.
/// - T12 is enabled.
/// - T13 is enabled.
/// - MCM is enabled.

CCU60_KSCFG = 0x0003; // Kernel State Configuration Register

_nop_(); //No operation function as Delay

_nop_(); //No operation function as Delay
_nop_(); //No operation function as Delay
_nop_(); //No operation function as Delay

/// -----------------------------------------------------------------------
/// Configuration of CCU60 Timer 12:
/// -----------------------------------------------------------------------
/// - Timer 12 Input clock factor (T12CLK) is 0
/// - prescaler factor is 1
/// - Timer 12 run bit is reset
/// - Single shot mode is disabled
/// - Timer 12 works in edge aligned mode
/// - Interrupt on period match is enabled
/// - Interrupt on one match is disabled
/// - No External run selection is selected.

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 81

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

/// - Timer mode is selected.

/// -

CCU60_T12PR = 0xFFFF; // load CCU60 T12 period register

/// -----------------------------------------------------------------------
/// Configuration of T13HR Signal:
/// -----------------------------------------------------------------------
/// - Signal T13HRA is selected as Input

/// -----------------------------------------------------------------------
/// Configuration of CCU60 Timer 13:
/// -----------------------------------------------------------------------
/// - Timer 13 Input Clock factor (T13CLK) is 0
/// - prescaler factor is 0
/// - Timer 13 run bit is reset
/// - Trigger control is disabled
/// - No External run selection is selected.
/// - Timer mode is selected.
/// -
/// - Single shot mode is disabled
/// - Interrupt on period match is enabled
/// - Interrupt on compare match is enabled

CCU60_T13PR = 0x0A4F; // load CCU60 T13 period register

CCU60_TCTR0 = 0x0008; // load CCU60 timer control register 0

CCU60_TCTR2 = 0x0020; // load CCU60 timer control register 2

/// -----------------------------------------------------------------------
/// Configuration of Multi Channel Mode:
/// -----------------------------------------------------------------------
/// - Multi channel mode is enabled

/// - switching selelction:

/// - Transfer on T12 channel 1 compare match

/// - Switching synchronization:

/// - Synchronisation on T13 period match

/// -----------------------------------------------------------------------
/// Configuration of CCU60 Channel 0:
/// -----------------------------------------------------------------------
/// - Hall sensor mode is selected
/// - The output CC60 is set to the passive state
/// - The trap functionality of the pin CC60 is enabled
/// - The passive level of the output CC60 is '0'
/// - The output COUT60 is set to the passive state
/// - The trap functionality of the pin COUT60 is enabled
/// - The passive level of the output COUT60 is '0'
/// - Dead time generation is disabled

/// - Generation interrupt on flag ICC60R is disabled

/// - Generation interrupt on flag ICC60F is disabled

CCU60_CC60SR = 0x0000; // Load CCU60 capture/compare shadow

// register for channel 0

/// -----------------------------------------------------------------------
/// Configuration of CCU60 Channel 1:
/// -----------------------------------------------------------------------
/// - Hall sensor mode is selected
/// - The output CC61 is set to the passive state
/// - The trap functionality of the pin CC61 is enabled
/// - The passive level of the output CC61 is '0'
/// - The output COUT61 is set to the passive state

Page: 4

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

82

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

/// - The trap functionality of the pin COUT61 is enabled

/// - The passive level of the output COUT61 is '0'
/// - Dead time generation is disabled

/// - Generation interrupt on flag ICC61R is disabled

/// - Generation interrupt on flag ICC61F is disabled

CCU60_CC61SR = 0x0000; // Load CCU60 capture/compare shadow

// register for channel 1

/// -----------------------------------------------------------------------
/// Configuration of CCU60 Channel 2:
/// -----------------------------------------------------------------------
/// - Hall sensor mode is selected
/// - The output CC62 is set to the passive state
/// - The trap functionality of the pin CC62 is enabled
/// - The passive level of the output CC62 is '0'
/// - The output COUT62 is set to the passive state
/// - The trap functionality of the pin COUT62 is enabled
/// - The passive level of the output COUT62 is '0'
/// - Dead time generation is disabled

/// - Generation interrupt on flag ICC62R is enabled

/// - Generation interrupt on flag ICC62F is disabled

CCU60_CC62SR = 0x0000; // Load CCU60 capture/compare shadow

// register for channel 2

/// -----------------------------------------------------------------------
/// Configuration of CCU60 Channel 3:
/// -----------------------------------------------------------------------
/// - T13 output is not inverted

CCU60_CC63SR = 0x0000; // load CCU60 capture/compare shadow

// register for channel 3

CCU60_T12DTC = 0x0003; // load CCU60 dead time control register for

// timer T12

CCU60_T12MSEL = 0x0888; // load CCU60 T12 campture/compare mode

// select register

CCU60_MODCTR = 0x2A80; // load CCU60 modulation control register

CCU60_MCMCTR = 0x0014; // load CCU60 multi channel mode control

// register

/// -----------------------------------------------------------------------
/// Configuration of CCU60 trap control:
/// -----------------------------------------------------------------------
/// - Trap can only be generated by SW by setting the bit TRPF
/// - The trap state is left immediately without any synchronization to
/// T12 or T13
/// - The trap state can be left as soon as bit TRPF is reset by SW
/// (according to TRPPEN, TRPM0 and TRPM1)
/// - Trap interrupt is enabled

CCU60_TRPCTR = 0x3F07; // Load CCU60 trap control register

/// -----------------------------------------------------------------------
/// Configuration of CCU60 interrupt control:
/// -----------------------------------------------------------------------
/// - For channel 0 interrupts is node I1 selected
/// - For channel 1 interrupts is node I1 selected
/// - For channel 2 interrupts is node I2 selected

Page: 5

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 83

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

/// - For correct hall event interrupt is node I2 selected

/// - For error interrupts is node I1 selected
/// - For T12 interrupts is node I1 selected
/// - For T13 interrupts is node I3 selected

CCU60_INP = 0x35A5; // Load CCU60 capture/compare interrupt node

// pointer register

CCU60_IEN = 0x2790; // Load CCU60 capture/compare interrupt

// enable register

/// -----------------------------------------------------------------------
/// Configuration of the used CCU60 Channel Port Pins:
/// -----------------------------------------------------------------------
/// - P10.0 is used for CCU60 output(CC60)
/// - P10.1 is used for CCU60 output(CC61)
/// - P10.2 is used for CCU60 output(CC62)
/// - P10.3 is used for CCU60 output(COUT60)
/// - P10.4 is used for CCU60 output(COUT61)
/// - P10.5 is used for CCU60 output(COUT62)
/// - P10.7 is used for CCU60 input(CCPOS0A)
/// - P10.8 is used for CCU60 input(CCPOS1A)
/// - P10.9 is used for CCU60 input(CCPOS2A)
/// - P10.6 is used for CCU60 input(CTRAPA)

P10_IOCR00 = 0x00A0; //set direction register

P10_IOCR01 = 0x00A0; //set direction register
P10_IOCR02 = 0x00A0; //set direction register
P10_IOCR03 = 0x00A0; //set direction register
P10_IOCR04 = 0x00A0; //set direction register
P10_IOCR05 = 0x00A0; //set direction register

CCU60_PISELL = 0xC000; // Load CCU60 Port Input Selection register

/// -----------------------------------------------------------------------
/// Configuration of the used CCU60 Channels Interrupts:
/// -----------------------------------------------------------------------
/// NodeI1 service request node configuration:
/// - NodeI1 interrupt priority level (ILVL) = 8
/// - NodeI1 interrupt group level (GLVL) = 0
/// - NodeI1 group priority extension (GPX) = 0

CCU60_1IC = 0x0060;

/// NodeI2 service request node configuration:

/// - NodeI2 interrupt priority level (ILVL) = 5
/// - NodeI2 interrupt group level (GLVL) = 1
/// - NodeI2 group priority extension (GPX) = 0

CCU60_2IC = 0x0055;

/// NodeI3 service request node configuration:

/// - NodeI3 interrupt priority level (ILVL) = 6
/// - NodeI3 interrupt group level (GLVL) = 2
/// - NodeI3 group priority extension (GPX) = 0

CCU60_3IC = 0x005A;

/// -----------------------------------------------------------------------
/// Configuration of T12, T13 ---- CCU60_TCTR4 Register:
/// -----------------------------------------------------------------------
/// - Enable shadow transfer of T12 and T13
/// - Timer 12 run bit is reset
/// - Timer 13 run bit is reset
CCU60_TCTR4 = 0x4040; // load CCU60 timer control register 4

Page: 6

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

84

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

// USER CODE BEGIN (NodeI3,3)

// USER CODE END

} // End of function CCU60_vInit

//****************************************************************************
// @Function void CCU60_viNodeI1(void)
//
//----------------------------------------------------------------------------
// @Description This is the interrupt service routine for the CCU60 node
// I1. If the content of the corresponding compare timer
// (configurable) equals the content of the capture/compare
// register or if a capture event occurs at the associated
// port pin, the interrupt request flag is set and an
// interrupt is triggered (only if enabled).
// Please note that you have to add application specific code
// to this function.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (NodeI1,1)

// USER CODE END

void CCU60_viNodeI1(void) interrupt CCU60_NodeI1_INT

{
// USER CODE BEGIN (NodeI1,2)

// USER CODE END

if(CCU60_IS & 0x0080) // if CCU60_IS_T12PM

{
// Timer T12 period match detection

// USER CODE BEGIN (NodeI1,19)

status_time_out=1; // ... set the time-out
status flag.
Motor_stop(); // ... stop the motor.
// USER CODE END

CCU60_ISR |= 0x0080; // clear flag CCU60_IS_T12PM

}

if(CCU60_IS & 0x0400) // if CCU60_IS_TRPF

{
// Trap detection

// USER CODE BEGIN (NodeI1,17)

status_ctrap=1; // ... set the trap sta
tus flag.
Motor_stop(); // ... stop the motor.
// USER CODE END

CCU60_ISR |= 0x0400; // clear flag CCU60_IS_TRPF

}

Page: 7

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 85

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

if(CCU60_IS & 0x2000) // if CCU60_IS_WHE

{
// Wrong hall event detection

// USER CODE BEGIN (NodeI1,18)

status_hs_error=1; // ... set the hall sen
sor error status flag.

if(status_bemf) // If bemf detection is

possible...
{
bemf_aux = CCU60_T12; // ... save the act
ual T12 value.
if(slope == 0 || slope == 1) // If bemf not dete
cted yet in this sector...
{
CCU60_T12MSEL = 0x0111; // ... change h
all sensor mode to sensorless mode.
CCU60_IEN &= 0x97FF; // ... disable
wrong hall event or idle interrupts.
CCU60_TCTR4 = 0x0045; // ... T12 rese
t and stop.
CCU60_T12 = zc_time + bemf_aux; // ... set the
counter time to the half of the sector time (actual zc_time).
CCU60_T12PR = (zc_time << 1) + zc_time; // ... set the
time out limit to 1.5 of the sector time.
CCU60_TCTR4 = 0x0040; // ... shadow t
ransfer request for T12.
CCU60_TCTR4 = 0x0002; // ... start T1
2.
if(bemf_aux < CCU60_CC61R) // If the commu
tation was not done yet...
{
CCU60_MCMOUTS |= 0x0080; // ... forc
e commutation.
}
if(bemf_aux < CCU60_CC62R) // If the updat
e of the pattern was not done yet...
{
CCU60_MCMOUTS = CCU60_MCMOUTS = SP_Tab_r_l[((CCU60_MCMOUT & 0x3800) >> 11
) + direction]; // ... load next commutation pattern.
}
}
else // If bemf has alre
ady been detected in this sector...
{
CCU60_T12MSEL = 0x0111; // ... change h
all sensor mode to sensorless mode.
CCU60_IEN &= 0x97FF; // ... disable
wrong hall event or idle interrupts.
CCU60_TCTR4 = 0x0045; // ... T12 rese
t and stop.
CCU60_T12 = bemf_aux - zc_time; // ... set the
counter time to the difference between the actual value and half the sector time.
CCU60_T12PR = (zc_time << 1) + zc_time; // ... set the
time out limit to 1.5 the sector time.
CCU60_CC60SR = (zc_time << 1); // ... set the
CC60R for speed measurement with the time between zero crossing points.
CCU60_CC61SR = zc_time - 5; // ... initiali
ze the CCU61 register with half of the time between zero crossings,
// so that it c
ommutates within the required time (with an advance).
CCU60_CC62SR = CCU60_CC61SR + 10; // ... initiali
ze CCU62 some time after to calculate the speed.
CCU60_TCTR4 = 0x0040; // ... shadow t
ransfer request for T12.
CCU60_TCTR4 = 0x0002; // ... start T1

Page: 8

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

86

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

2.
}
}
else // If bemf detection i
s not possible...
{
Motor_stop(); // ... stop the mot
or.
}
// USER CODE END

CCU60_ISR |= 0x2000; // clear flag CCU60_IS_WHE

}

} // End of function CCU60_viNodeI1

//****************************************************************************
// @Function void CCU60_viNodeI2(void)
//
//----------------------------------------------------------------------------
// @Description This is the interrupt service routine for the CCU60 node
// I2. If the content of the corresponding compare timer
// (configurable) equals the content of the capture/compare
// register or if a capture event occurs at the associated
// port pin, the interrupt request flag is set and an
// interrupt is triggered (only if enabled).
// Please note that you have to add application specific code
// to this function.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (NodeI2,1)

// USER CODE END

int arrexp[8]={0,5,3,1, 6,4,2,0,}; //for clockwise

unsigned int MCMData[8] = {0x190e, 0x190e, 0x3223, 0x132c, 0x2c38, 0x0d32, 0x260b,
0x190e}; //FOR CLOCLKWISE DIRECTION WORKING FINE

//int arrexp[8]={0,3,6,2,5,1,4,0,}; //for anti clockwise direction

//unsigned int MCMData[8] = {0x292c, 0x292c, 0x1A0B, 0x0B23, 0x3432, 0x250E, 0x1638
, 0x292c}; // FOR ANTI - CLOCLKWISE DIRECTION WORKING FINE

void CCU60_viNodeI2(void) interrupt CCU60_NodeI2_INT

{
// USER CODE BEGIN (NodeI2,2)

// USER CODE END

if(CCU60_IS & 0x0010) // if CCU60_IS_ICC62R

{
// Capture, Compare match rising edge detection an channel 2

// USER CODE BEGIN (NodeI2,14)

CCU60_MCMOUTS = SP_Tab_r_l[((CCU60_MCMOUT & 0x3800) >> 11) + direction]; //

... load the next commutation pattern.

Page: 9

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 87

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

speed_meas = Speed_calc(CCU60_CC60R, &hall_counter, (18*PP)); //

... calculate the speed.

if (slope == 2) //
If a falling edge of the bemf has been detected...
{
slope = 0;
// ... we need to look for rising edge.
}
if (slope == 3) //
If a rising edge of the bemf has been detected...
{
slope = 1;
// ... we need to look for falling edge.
}

if (hall_counter < 2) //
If motor speed is too small...
{
speed_meas = 0;
}
// ... reset to 0 (Only done at the first measurements of the speed).
// USER CODE END

CCU60_ISR |= 0x0010; // clear flag CCU60_IS_ICC62R

}

} // End of function CCU60_viNodeI2

//****************************************************************************
// @Function void CCU60_viNodeI3(void)
//
//----------------------------------------------------------------------------
// @Description This is the interrupt service routine for the CCU60 node
// I3. If the content of the corresponding compare timer
// (configurable) equals the content of the capture/compare
// register or if a capture event occurs at the associated
// port pin, the interrupt request flag is set and an
// interrupt is triggered (only if enabled).
// Please note that you have to add application specific code
// to this function.
//
//----------------------------------------------------------------------------
// @Returnvalue None
//
//----------------------------------------------------------------------------
// @Parameters None
//
//----------------------------------------------------------------------------
// @Date 26/08/2014
//
//****************************************************************************

// USER CODE BEGIN (NodeI3,1)

// USER CODE END

void CCU60_viNodeI3(void) interrupt CCU60_NodeI3_INT

{
// USER CODE BEGIN (NodeI3,2)
unsigned int x;
// USER CODE END

if(CCU60_IS & 0x0100) // if CCU60_IS_T13CM

{
// Timer T13 compare match detection

Page: 10

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

88

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

// USER CODE BEGIN (NodeI3,22)

bemf_aux = (CCU60_MCMOUT & 0x003F); // ...
read the present modulation pattern.

for(x=0; x<200; x++)

{
_nop_(); // No o
peration function delay
}

if((bemf_aux == BEMF_AUX_VAL11) || (bemf_aux == BEMF_AUX_VAL12)) // If p

hase A is inactive...
{
ADC1_CRPR1 = 0x0001; //
... convert channel 1 of ADC1.
}
if((bemf_aux == BEMF_AUX_VAL21) || (bemf_aux == BEMF_AUX_VAL22)) // If p
hase B is inactive...
{
ADC1_CRPR1 = 0x0004; //
... convert channel 4 of ADC1.
}
if((bemf_aux == BEMF_AUX_VAL31) || (bemf_aux == BEMF_AUX_VAL32)) // If p
hase B is inactive...
{
ADC1_CRPR1 = 0x0007; //
... convert channel 4 of ADC1.
}
ADC0_CRPR1 = 0x0003; // ...
convert DC link current.
ADC1_QINR0 = 0x0082; // ...
request DC link voltage measurement (on T13 period match).
compare_value = ((ADC1_RESR1>>3)&0x01FF)-adc_offset_resVdc; // ...
set the compare value to half the DC link voltage.
// USER CODE END

CCU60_ISR |= 0x0100; // clear flag CCU60_IS_T13CM

}

if(CCU60_IS & 0x0200) // if CCU60_IS_T13PM

{
// Timer T13 period match detection

// USER CODE BEGIN (NodeI3,21)

if(slope==1 || slope==0) // If b
emf still not detected...
{
Sensorless_calc(); //
... check the bemf and calculate the new parameters if necessary.
}

if(spd_loop==1) // If 1
ms reached...
{
speed_ref_pu = FloatTo1Q15((float)(speed_ref/MAX_SPEED)); //
... calculate the speed reference in pu (1Q15).
speed_meas_pu = FloatTo1Q15((float)(speed_meas/MAX_SPEED)); //
... calculate the measured speed in pu (1Q15).
PI_spd_out = pi_controller64(&PI_spd.yn, speed_ref_pu, speed_meas_pu); //
... compute the I_ref.
spd_loop=0; //
... reset the counter.
}
else // If 1
ms not reached...
{
spd_loop++; //

Page: 11

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 89

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.C

... increase the counter.

}
Idc_meas = ((ADC1_RESR0<<3) & 0x7FE0)-adc_offset_resIdc; // ...
save the measured current in pu (1Q15).
PI_curr_out = pi_controller64(&PI_spd.yn, PI_spd_out, Idc_meas); // ...
compute the duty_cycle.
CCU60_CC63SR = (int)((3300*((long)(32767-PI_curr_out)))>>15); // ...
set the duty cycle to T13.
CCU60_TCTR4 = 0x4000; // ...
enable shadow transfer.
increment_free = 1; // ...
set flag for speed ramp.
// USER CODE END

CCU60_ISR |= 0x0200; // clear flag CCU60_IS_T13PM

}

} // End of function CCU60_viNodeI3

// USER CODE BEGIN (CCU60_General,10)

// USER CODE END

Page: 12

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

90

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.H

//****************************************************************************
// @Module Capture / Compare Unit 60 (CCU60)
// @Filename CCU60.H
// @Project Aircraft.dav
//----------------------------------------------------------------------------
// @Controller Infineon XE167F-96F66
//
// @Compiler Keil
//
// @Codegenerator 2.2
//
// @Description This file contains all function prototypes and macros for
// the CCU60 module.
//
//----------------------------------------------------------------------------
// @Date 26/08/2014 17:45:07
//
//****************************************************************************

// USER CODE BEGIN (CCU60_Header,1)

// USER CODE END

#ifndef _CCU60_H_
#define _CCU60_H_

//****************************************************************************
// @Project Includes
//****************************************************************************

// USER CODE BEGIN (CCU60_Header,2)

// USER CODE END

//****************************************************************************
// @Macros
//****************************************************************************

// USER CODE BEGIN (CCU60_Header,3)

// USER CODE END

//****************************************************************************
// @Defines
//****************************************************************************

// This parameter identifies CCU60 timer 12

#define CCU60_TIMER_12 12

// This parameter identifies CCU60 timer 13

#define CCU60_TIMER_13 13

// This parameter identifies CCU60 channel 0

#define CCU60_CHANNEL_0 0

// This parameter identifies CCU60 channel 1

#define CCU60_CHANNEL_1 1

// This parameter identifies CCU60 channel 2

#define CCU60_CHANNEL_2 2

// This parameter identifies CCU60 channel 3

#define CCU60_CHANNEL_3 3

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 91

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CCU60.H

// USER CODE BEGIN (CCU60_Header,4)

// USER CODE END

//****************************************************************************
// @Typedefs
//****************************************************************************

// USER CODE BEGIN (CCU60_Header,5)

// USER CODE END

//****************************************************************************
// @Imported Global Variables
//****************************************************************************

// USER CODE BEGIN (CCU60_Header,6)

// USER CODE END

//****************************************************************************
// @Global Variables
//****************************************************************************

// USER CODE BEGIN (CCU60_Header,7)

// USER CODE END

//****************************************************************************
// @Prototypes Of Global Functions
//****************************************************************************

void CCU60_vInit(void);

// USER CODE BEGIN (CCU60_Header,8)

// USER CODE END

//****************************************************************************
// @Interrupt Vectors
//****************************************************************************

#define CCU60_NodeI1_INT 0x31

#define CCU60_NodeI2_INT 0x32
#define CCU60_NodeI3_INT 0x33

// USER CODE BEGIN (CCU60_Header,9)

// USER CODE END

#endif // ifndef _CCU60_H_

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

92

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

#include "MAIN.H"

// Variable declaration

// ADC offset calculation

long C_SDATA adc_offset_acumm; // Accumulator for averaging of ADC off
set.
int C_SDATA adc_offset_resA; // Computed ADC offset for bemf A.
int C_SDATA adc_offset_resB; // Computed ADC offset for bemf B.
int C_SDATA adc_offset_resC; // Computed ADC offset for bemf C.
int C_SDATA adc_offset_resVdc; // Computed ADC offset for DC link volt
age.
int C_SDATA adc_offset_resIdc; // Computed ADC offset for DC link curr
ent.

// Declaration of bit-variables required to store motor status

C_BDATA status_word; // Full motor status word.

C_SBIT status_on C_ATBIT(status_word,0); // ON.

C_SBIT status_off C_ATBIT(status_word,1); // OFF.

C_SBIT status_overcurrent C_ATBIT(status_word,2); // Overcurrent detection.

C_SBIT status_hs_error C_ATBIT(status_word,3); // Hall sensor error detect
ion.
C_SBIT status_time_out C_ATBIT(status_word,4); // Time_out detection.
C_SBIT status_ctrap C_ATBIT(status_word,5); // Ctrap detection.
C_SBIT status_bemf C_ATBIT(status_word,6); // Status of bemf detection
.
C_SBIT status_overvoltemp C_ATBIT(status_word,7); // Overvoltage or overtempr
eature detection.

// Global variables
int C_SDATA direction; // Index-offset for reading the swi
tching pattern, depending on the specified motor direction.
int C_SDATA once; // Variable to allow motor start.
unsigned int C_SDATA sv_start_ref; // Start speed reference for motor
start.
unsigned int C_SDATA sv_start_ref0; // Copy of start reference to check
changes in end reference and recalculate ramp time.
unsigned int C_SDATA sv_end_ref; // End speed reference for motor st
art.
int C_SDATA motor_status; // Motor stop command.
int C_SDATA direction_user; // Variable which stores the desire
d direction of rotation.
unsigned long C_SDATA rampup_time; // Time to achieve the end referenc
e.
int C_SDATA trap_reset; // Set to '1' to reset the 'Latched
CTRAP' condition.
unsigned int C_SDATA speed_ref; // Desired motor speed reference.
unsigned int C_SDATA circ_index; // Index to access the circular buf
fer used in motor speed computation.
unsigned int C_SDATA speed_buffer[6*PP]; // Buffer to preserve the previous
samples of motor speed.
int C_SDATA oc_counter; // Counter to specify the time limi
t for motor overcurrent.
unsigned long C_SDATA time_step; // Sets the increment frequency of
the speed reference.
unsigned long C_SDATA counter1; // Counter to increase the speed re
ference.
unsigned long C_SDATA speed_acumm; // Accumulator to calculate the ave
rage motor speed.
int C_SDATA hall_counter; // Counter to start taking into acc
ount speed measurements.
int C_SDATA speed_counter; // Counter to start measuring avera
ge speed.
int C_SDATA increment_free; // Flag to specify that T13 interru

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 93

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

pt has occured.
int C_SDATA spd_loop; // Speed closed loop prescaler.
int C_SDATA Idc_meas; // DC link current measurement.
unsigned int C_SDATA speed_meas; // Variable to store the actual mot
or speed (sector).

// BEMF
unsigned int C_SDATA compare_value; // DC link voltage/2
int C_SDATA slope; // Slope of the bemf: Rising = 0, F
alling = 1.
int C_SDATA CH_conv; // Stores the last converted channe
l to apply the corresponding offset.
unsigned int C_SDATA zc_time; // Time between zero crossing point
s in sensorless mode.
unsigned int C_SDATA bemf_aux; // Stores the current value of T12
to change to sensorless mode.
unsigned int C_SDATA back_emf_measure; // Last measured bemf.
int C_SDATA slope_known; // Slope of the bemf is detected .
int C_SDATA bemf_counter1; // Counter for detection of bemf.

// PI variables
struct PI_cont C_SDATA PI_spd; // PI speed variable.
struct PI_cont C_SDATA PI_curr; // PI current variable.
int C_SDATA PI_spd_out; // PI speed output.
int C_SDATA PI_curr_out; // PI current output.

// Function declaration

void ADC_OFFSET(int CH_OFFSET);

void Init_control(void);
void Motor_start(void);
void Motor_stop(void);
void Commutation_init(void);
void RampUp(void);
void Speed_ref_ramp(void);
int Speed_calc(unsigned int time_meas, int* speed_counter, int limit);
void Sensorless_calc(void);

// Modulation pattern
#define MCMOUTS_CTE_R(curhs,exphs,mcmps) (curhs << 11) | (exphs << 8) | (mcmps
)
#define MCMOUTS_CTE_L(curhs,exphs,mcmps) (curhs << 11) | (exphs << 8) | (mcmps
)

int const SP_St_Tab_r_l[16] =

// For motor start (0-7: Clockwise, 8-15: Counterclockwise)
{
0x0000, MCMOUTS_CTE_R(1,3,0x24), MCMOUTS_CTE_R(2,6,0x09), MCMOUTS_CTE_R(3,2
,0x21),
MCMOUTS_CTE_R(4,5,0x12), MCMOUTS_CTE_R(5,1,0x06), MCMOUTS_CTE_R(6,4,0x18),
0x0000,
0x0000, MCMOUTS_CTE_L(1,5,0x18), MCMOUTS_CTE_L(2,3,0x06), MCMOUTS_CTE_L(3,1
,0x12),
MCMOUTS_CTE_L(4,6,0x21), MCMOUTS_CTE_L(5,4,0x09), MCMOUTS_CTE_L(6,2,0x24),
0x0000
};
int const SP_Tab_r_l[16] =
// Normal operation (0-7: Clockwise, 8-15: Counterclockwise)
{
0x0000, MCMOUTS_CTE_R(3,2,0x21), MCMOUTS_CTE_R(6,4,0x18), MCMOUTS_CTE_R(2,6
,0x09),
MCMOUTS_CTE_R(5,1,0x06), MCMOUTS_CTE_R(1,3,0x24), MCMOUTS_CTE_R(4,5,0x12),
0x0000,
0x0000, MCMOUTS_CTE_L(5,4,0x09), MCMOUTS_CTE_L(3,1,0x12), MCMOUTS_CTE_L(1,5
,0x18),
MCMOUTS_CTE_L(6,2,0x24), MCMOUTS_CTE_L(4,6,0x21), MCMOUTS_CTE_L(2,3,0x06),
0x0000

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

94

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

};

// Function implementation

//****************************************************************************
// @Function void ADC_OFFSET(int CH_OFFSET)
// Description This function computes the ADC offset value, by taking
// 65,536 samples of the ADC measured data. This computed
// offset value is preserved in the variable adc_offset_resX.
// Parameters Channel to be converted
//----------------------------------------------------------------------------
void ADC_OFFSET(int CH_OFFSET)
{
unsigned int i = 0;
adc_offset_acumm = 0; // Reset the ac
cumulator.
switch(CH_OFFSET)
{
case 1: // Phase A.
for(i = 0; i < 65535; i++) // Until al
l samples not accumulated, be in the loop.
{
ADC1_CRPR1 = 0x0001; // Requ
est 'start sampling' command to the ADC.
while(!(ADC1_VFR & 0x0001)); // Wait
until valid conversion.
adc_offset_acumm += (ADC1_RESR0>>2 & 0x03FF); // Accu
mulate the sample, right aligned.
}
adc_offset_resA = (int)(adc_offset_acumm >> 16); // Do the a
verage of all the samples.
break;
case 4: // Phase B
for(i = 0; i < 65535; i++) // Until al
l samples not accumulated, be in the loop.
{
ADC1_CRPR1 = 0x0004; // Requ
est 'start sampling' command to the ADC.
while(!(ADC1_VFR & 0x0001)); // Wait
until valid conversion.
adc_offset_acumm += (ADC1_RESR0>>2 & 0x03FF); // Accu
mulate the sample, right aligned.
}
adc_offset_resB = (int)(adc_offset_acumm >> 16); // compute
the average of all the samples.
break;
case 7: // Phase C
for(i = 0; i < 65535; i++) // Until al
l samples not accumulated, be in the loop.
{
ADC1_CRPR1 = 0x0007; // Requ
est 'start sampling' command to the ADC.
while(!(ADC1_VFR & 0x0001)); // Wait
until valid conversion.
adc_offset_acumm += (ADC1_RESR0>>2 & 0x03FF); // Accu
mulate the sample, right aligned.
}
adc_offset_resC = (int)(adc_offset_acumm >> 16); // Do the a
verage of all the samples.
break;
case 3: // DC link curr
ent.
for(i = 0; i < 65535; i++) // Until al
l samples not accumulated, be in the loop.
{
ADC0_CRPR1 = 0x0003; // Requ
est 'start sampling' command to the ADC.
while(!(ADC0_VFR & 0x0001)); // Wait

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 95

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

until valid conversion.

adc_offset_acumm += ((ADC0_RESR0 & 0x0FFC) << 3); // Accu
mulate the sample, 1Q15.
}
adc_offset_resIdc = (int)(adc_offset_acumm >> 16); // Do the a
verage of all the samples.
break;
case 2: // DC link volt
age.
for(i = 0; i < 65535; i++) // Until al
l samples not accumulated, be in the loop.
{
ADC1_CRPR1 = 0x0002; // Requ
est 'start sampling' command to the ADC.
while(!(ADC1_VFR & 0x0002)); // Wait
until valid conversion.
adc_offset_acumm += (ADC1_RESR0>>3 & 0x01FF); // Accu
mulate the sample, 1Q15.
}
adc_offset_resVdc = (int)((adc_offset_acumm >> 16)-511); // Do the a
verage of all the samples.
break;
}
}

//****************************************************************************
// @Function void Motor_start(void)
// Description This function executes the functions required for Motor
// start-up such as initialization of motor related variables,
// initialization of motor SFRs, etc.
// It also ensures that the start reference is always lower
// than the end reference.
//----------------------------------------------------------------------------
void Motor_start(void)
{
if(once == 0) // Execute only the first t
ime.
{
once = 0x0001; // Set the flag so that thi
s code never executes until next motor start-up.
status_word = 0x0001; // Set 'status_on' flag to
convey to the user that 'Motor_start' command is issued.

if(sv_start_ref > sv_end_ref) // Start reference limitati

on.
{
sv_start_ref = sv_end_ref; // Ensure that the star
t reference never exceeds the end reference.
}

sv_start_ref0 = sv_start_ref; // Make another copy of sta

rt voltage reference to check later if the end reference changes.
speed_ref = sv_start_ref; // Set start speed referenc
e into actual speed reference.

Init_control(); // Initialisation of all th

e variables required for motor start-up.
Commutation_init(); // Initialisation of all SF
Rs required for motor start-up.
}
}

//****************************************************************************
// @Function void Init_control(void)
// Description This function initialises various variables required
// for motor start-up.
//----------------------------------------------------------------------------
void Init_control(void)

Page: 4

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

96

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

int i = 0;
for(i = 0;i<(6*PP);i++)
{
speed_buffer[i] = 0; // Initialise the speed buffer.
}
speed_acumm = 0x00000008; // Initialise speed accumulator with 8 so that divi
sion by 8 won't result in zero.
circ_index = 0x0000; // Initialise the circular index for speed measurem
ent.
speed_counter = 0x0000; // Initialise the speed counter.
hall_counter = 0x0000; // Initialise the hall counter.

counter1 = 0x00000000; // Initialise counter 1.

time_step = 0x00000000; // Initialise the time_step for the ramp-up.

slope_known = 0x0000; // Initialise the detection of the bemf slope.

bemf_counter1 = 0x0000; // Initialise bemf counter 1.
slope = 0x0000; // Initialise the slope of the bemf.

spd_loop = 0; // Initialise the speed closed loop prescaler.

PI_spd.yn = 0x00000000; // Initialise the integral part of the speed PI.
PI_curr.yn = 0x00000000; // Initialise the integral part of the current PI.

oc_counter = 0x0000; // Initialise the s/w time base for overcurrent det
ection.
}

//****************************************************************************
// @Function void Commutation_init(void)
// Description This function initialises the SFRs required for motor start-up
// such as loading of the modulation register, enabling of the timers,
// enabling of the interrupts etc.
//----------------------------------------------------------------------------
void Commutation_init(void)
{
int aux = 0;

CCU60_MCMCTR = 0x0000; // No switching or synchronization

available.

CCU60_CC61SR = 3; // Initialise CC61 register to avoi

d hall sensor spikes.
CCU60_CC62SR = CCU60_CC61SR + 5; // Initialise CC62 some time later
than CC61 to calculate speed and set the new pattern.
CCU60_T12 = CCU60_CC62SR + 1; // Initialize T12 timer with a suff
iciently larger value so that
// it doesn't generate an interrupt
immediately.

CCU60_CC63SR = 0x0CE4; // Initialize channel 3 shadow regi

ster with value corresponding to 0% duty cycle.

CCU60_TCTR4 = 0x4040; // Select the events to trigger sha

dow transfers of T12 & T13.

aux = ((P10_IN & 0x0380)>>7) + direction; // Read the hall sensor pins from t
he corresponding port.

CCU60_MCMOUTS = SP_St_Tab_r_l[aux] | 0x8080; // Load the Modulation pattern corr

esponding to the hall sensor data
// and force the shadow transfer of
the modulation pattern to the actual register.
CCU60_MCMOUTS = SP_Tab_r_l[aux]; // Load the expected hall and modul
ation patterns into shadow register.

CCU60_IEN |= 0x6000; // Enable idle and wrong hall event

Page: 5

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 97

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

interrupts.

CCU60_TCTR4 = 0x0600; // T13 Run --> Start PWM.

CCU60_MCMCTR = 0x0014; // Trigger shadow tranfers at CC61R

compare match and synchronized to T13 zero match.

CCU60_TCTR4 = 0x0002; // T12 Run.

CCU60_TRPCTR |= 0x8000; // Enable CCU6 trap.

//****************************************************************************
// @Function void Motor_stop(void)
// Description This function implements the functions required for motor
// stop such as switching off the PWM, all MOSFETs, stopping
// all timers etc.
//----------------------------------------------------------------------------

void Motor_stop(void)
{

CCU60_MCMCTR = 0x0000; // Enable shadow tranfers direct mode.

CCU60_MCMOUTS = 0x8080; // All transistors OFF.

CCU60_TRPCTR &= 0x7FFF; // Trap disable.

CCU60_TCTR4 = 0x0505; // T13/T12 Reset, T13/T12 Stop --> No PWM

active.

CCU60_CC63SR = 0x0CE4; // Re-initialise shadow register channel 3

with duty cycle = 0,
// so that it is ready with the correct val
ue when 'motor start' command is requested.

CCU60_TCTR4 = 0x4000; // Shadow Tranfer by T13.

CCU60_IEN &= 0x3FFF; // Wrong hall event and idle interrupts dis
abled.

status_off = 1; // Set motor off.

status_on = 0; // Clear motor on.

if(motor_status == 1) // If motor stop command detected...

{
once = 0x0000; // ... clear variable 'once' so that it
executes motor start function when user requests 'start' command.
}

if(direction_user > 1) // If 'direction_user' is higher than 1...

{
direction_user = 1; // ... ensure that 'direction_user' nev
er takes an invalid value.
}

direction = direction_user * 8; // Set the offset for the pattern based on

'direction_user'.

//****************************************************************************

Page: 6

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

98

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

// @Function void RampUp(void)

// Description This function verifies whether the user has entered a new
// 'end reference' and accordingly ramps-up the duty cycle,
// to reach the new 'end reference'.
//----------------------------------------------------------------------------
void RampUp(void)
{

int ref_difference = 0; // Initialise the differenc

e.
if(increment_free == 1) // If the T13 period match
timer interrupt has occured...
{
increment_free = 0; // ... clear the interr
upt flag.
counter1 += 64; // ... increase the cou
nter.
}

if(counter1 > time_step) // If the 'time_step' for r

amping has elapsed...
{
Speed_ref_ramp(); // ... increment the re
ference speed by one count.
counter1 = 0; // ... reset the counte
r.
}

if(sv_start_ref0 != sv_end_ref) // If the end speed referen

ce has been changed by the user...
{

C_CoLOAD_0_1(0, C_LSHIFT(sv_end_ref, 16)); // ... then calculate t

he new difference that the motor has to achieve.
C_CoSUB_0_1(0, C_LSHIFT(sv_start_ref0, 16));
C_CoABS();
ref_difference = C_CoSTOREMAS();
if(ref_difference > 1) // ... substract th
e already done increment.
{
ref_difference -= 1;
}
time_step = rampup_time / ref_difference; // ... calculate the ne
w 'time_step' for the new difference in speed.
sv_start_ref0 = sv_end_ref; // ... assign the new r
eference so it doesn't calculate again.
}

//****************************************************************************
// @Function void Speed_ref_ramp(void)
// Description This function increments/decrements the actual speed
// reference until it becomes equal to the speed end reference.
//----------------------------------------------------------------------------
void Speed_ref_ramp(void)
{
if(sv_end_ref > speed_ref) // If the end reference hasn't b
een reached...
{
speed_ref += 1; // ... increment speed refe
rence.
}
if(sv_end_ref < speed_ref) // If the actual reference is gr
eater than end reference...
{
speed_ref -= 1; // ... decrement the speed
reference.

Page: 7

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 99

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

//****************************************************************************
// @Function void Speed_calc(void)
// Description This function computes the actual motor speed from the
// measured time between two hall events or zero crossing points.
//----------------------------------------------------------------------------
int Speed_calc(unsigned int time_meas, int* speed_counter, int limit)
{
unsigned int time = 0;
unsigned long speed_cal = 0;

speed_acumm += time_meas; // Accumulate the new measu

red time.

speed_acumm -= speed_buffer[circ_index]; // Subtract out the oldest

sample from the accumulator.

speed_buffer[circ_index] = time_meas; // Preserve the latest samp

le in the buffer.

if(*speed_counter < limit) // If motor is starting...

{
*speed_counter = *speed_counter + 1; // ... not enough sampl
es.
time = time_meas; // ... use the measured
sample directly, instead of the average.
speed_cal = SPEED_CONSTANT/(6 * PP); // ... load the divisio
n factor to calculate motor speed in RPM.
}
else // If motor has passed ramp
up stage....
{ // ... the accumulator
has sufficient samples.
if((int)((long)speed_acumm >> 16) > 0) // If the accumulated t
ime is too big...
{
time = (int)(speed_acumm >> 3); // ... scale it dow
n to a smaller number (lose precision).
speed_cal = SPEED_CONSTANT >> 3; // ... load the div
ision factor to calculate motor speed in RPM.
}
else
{ // If the accumulated t
ime is not too big...
time = (int) (speed_acumm); // ... don't scale
it down so that no precision is lost.
speed_cal = SPEED_CONSTANT; // ... load the div
ision factor to calculate motor speed.
}
}
circ_index++; // Increment index for circ
ular buffer.
if(circ_index > ((6 * PP) - 1)) // If the end of circular b
uffer is reached...
{
circ_index = 0; // ... reset the it to
'0'.
}

return ((int)(speed_cal / time)); // Compute motor speed.

//****************************************************************************
// @Function void Sensorless_calc(void)

Page: 8

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

100

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

// Description This function maintains the state operation of the motor

// in case the hall sensors don't work.
//----------------------------------------------------------------------------
void Sensorless_calc(void)
{
CH_conv = ADC1_RESR0>>12 & 0x0007; // Store th
e channel converted.
switch(CH_conv) // Check wh
ich phase is being measured.
{
case 1: // Phas
e A.
back_emf_measure = (ADC1_RESR0>>2 & 0x03FF)-adc_offset_resA; //
Substract the corresponding offset of phase A.
break;
case 4: // Phas
e B.
back_emf_measure = (ADC1_RESR0>>2 & 0x03FF)-adc_offset_resB; //
Substract the corresponding offset of phase B.
break;
case 7: // Phas
e C.
back_emf_measure = (ADC1_RESR0>>2 & 0x03FF)-adc_offset_resC; //
Substract the corresponding offset of phase C.
break;
}
if(!slope_known) // If the s
lope is not known yet...
{
if(slope == 0) // If w
e are looking for a rising edge...
{
if(back_emf_measure >= compare_value) //
If the bemf is greater than half of bus voltage...
{
bemf_counter1++; //
... increment samples counter.
if(bemf_counter1 == 4) //
If enough valid bemf samples detected...
{
slope = 1;
// ... look for a falling edge.
zc_time = CCU60_T12;
// ... save the captured time between two bemf zero crossings (in sensorless mode).
bemf_counter1 = 0;
// ... reset the counter.
}
}
}
if(slope == 1) // If w
e are looking for a rising edge...
{
if(back_emf_measure < compare_value) //
If the bemf is lower than half of bus voltage...
{
bemf_counter1++; //
... increment samples counter.
if(bemf_counter1 == 4) //
If enough valid bemf samples detected...
{
slope = 0;
// ... look for a falling edge.
zc_time = CCU60_T12;
// ... save the captured time between two bemf zero crossings (in sensorless mode).

bemf_counter1 = 0;
// ... reset the counter.
}

Page: 9

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 101

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

}
}
if(zc_time > 70) // If t
he value has sense (not just after the commutation)...
{
slope_known = 1; //
... the slope is already known.
}
}
else // If the s
lope is known...
{
if(slope == 0) // If w
e are looking for a rising edge...
{
if(back_emf_measure >= compare_value) //
If the bemf is greater than half of bus voltage...
{

bemf_counter1++;
// ... increase 'bemf_counter'.

if(bemf_counter1 == 2)
// If two valid samples are detected...
{
zc_time = CCU60_T12;
// ... save the captured time between two zero crossings (in sensorless mode).

status_bemf = 1;
// ... set the bemf detection status flag.
bemf_counter1 = 0;
// ... reset sample counter.
slope = 3;
// ... slope = 3, so that next we start looking for falling edge.
if(status_hs_error)
// If sensorless mode...
{
CCU60_CC60SR = zc_time;
// ... set the CC60R for speed measurement with the time between zero cross
ing points.
CCU60_CC61SR = (zc_time>>1) - 5;
// ... initialize the CCU61 register with half of the time between zero cro
ssings,

// so that it commutates within the required time (with an advance).

CCU60_CC62SR = CCU60_CC61SR + 10;
// ... initialize CCU62 some time after to calculate the speed.
CCU60_T12PR = (zc_time>>1) + zc_time;
// ... set time-out to 1.5 times the sector time.
CCU60_TCTR4 = 0x0045;
// ... T12 Reset and Stop (updated CC61R, CC62R and T12PR).
CCU60_TCTR4 = 0x0002;
// ... start T12 again.
}
}
}
}
else if(slope == 1) // If w
e are looking for a falling edge...
{
if(back_emf_measure < compare_value) //
If the bemf is lower than half of bus voltage...
{
bemf_counter1++;
// ... increase 'bemf_counter'.
if(bemf_counter1 == 2)
// If two valid samples are detected...

Page: 10

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

102

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.C

{
zc_time = CCU60_T12;
// ... save the captured time between two zero crossings (in sensorless mode).

status_bemf = 1;
// ... set the bemf detection status flag.
bemf_counter1 = 0;
// ... reset sample counter.
slope = 2;
// ... slope = 2, so that next we start looking for rising edge.
if(status_hs_error)
// If sensorless mode...
{
CCU60_CC60SR = zc_time;
// ... set the CC60R for speed measurement with the time between zero cross
ing points.
CCU60_CC61SR = (zc_time>>1) - 5;
// ... initialize the CCU61 register with half of the time between zero cro
ssings,

// so that it commutates within the required time (with an advance).

CCU60_CC62SR = CCU60_CC61SR + 10;
// ... initialize CCU62 some time after to calculate the speed.
CCU60_T12PR = (zc_time>>1) + zc_time;
// ... set time-out to 1.5 times the sector time.
CCU60_TCTR4 = 0x0045;
// ... T12 Reset and Stop (updated CC61R, CC62R and T12PR).
CCU60_TCTR4 = 0x0002;
// ... start T12 again.
}
}
}
}
}
}

Page: 11

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 103

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.H

// Import global variables

// Status word
extern C_BDATA status_word; // Full status motor state.

extern BIT status_ctrap; // Ctrap detection.

// Global variables
extern int C_SDATA direction; // Index-offset for reading the swi
tching pattern, depending on the specified motor direction.
extern int C_SDATA once; // Counter to specify the time limi
t for motor Overcurrent.
extern unsigned int C_SDATA sv_start_ref; // Start speed reference for motor
start.
extern unsigned int C_SDATA sv_end_ref; // End speed reference for motor st
art.
extern int C_SDATA motor_status; // Motor stop command.
extern int C_SDATA direction_user; // User desired direction.
extern unsigned long C_SDATA rampup_time; // Time to achieve the end referenc
e.
extern int C_SDATA trap_reset; // Set to '1' to reset the 'Latched
CTRAP' condition.

// PI variables
extern struct PI_cont C_SDATA PI_spd; // PI speed variable.
extern struct PI_cont C_SDATA PI_curr; // PI current variable.

// Import functions
extern void Motor_start(void);
extern void Motor_stop(void);

// Function inlines

//*************************************************************************************
************
// @Function void INIT_PI_VAR(void)
// Description This function initialises the PI gain & Limit variables with the co
rresponding
// user values in 'CONFIG.H'.
//-------------------------------------------------------------------------------------
------------
INLINE void INIT_PI_VAR(void)
{
PI_spd.ymax = PI_SPEED_YMAX; // Initialise the PI upper limit variable.
PI_spd.ymin = PI_SPEED_YMIN; // Initialise the PI lower limit variable.
PI_spd.kp = PI_SPEED_KP; // Initialise the PI proportional gain variable
.
PI_spd.ki = PI_SPEED_KI; // Initialise the PI integral gain variable.

PI_curr.ymax = PI_CURR_YMAX; // Initialise the PI upper limit variable.

PI_curr.ymin = PI_CURR_YMIN; // Initialise the PI lower limit variable.
PI_curr.kp = PI_CURR_KP; // Initialise the PI proportional gain variable
.
PI_curr.ki = PI_CURR_KI; // Initialise the PI integral gain variable.
}

//*************************************************************************************
************
// @Function void MOTOR_START_PARAMS(void)
// Description This Function loads the values for the motor start.
//-------------------------------------------------------------------------------------
------------
INLINE void MOTOR_START_PARAMS(void)
{
trap_reset = 1; // Set the 'trap_reset' flag so that 't
rap' flag is cleared under power ON.
status_ctrap = 0; // Clear the trap flag to allow the mot
or to run.

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

104

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.H

sv_end_ref = MOTOR_END_REF; // Initialise the end reference with pr

edefined value.
sv_start_ref = MOTOR_START_REF; // Initialise the start reference with
predefined value.
rampup_time = RAMP_UP_TIME; // Initialise the ramp-up time with pre
defined value.
direction_user = MOTOR_DIRECTION; // Initialise the desired direction.
direction = direction_user * 8; // Initialise the motor direction and m
ultiplies by 8 to select the apropiate commutation pattern.
motor_status = 0; // Clear the motor status flag to allow
start.
}

//*************************************************************************************
************
//*************************************************************************************
************
// @Function void INIT_CTRL_VAR(void)
// Description This function initialises the various intermediate variables that a
re required to
// be set to pre-defined default values before any motor related opera
tion.
//-------------------------------------------------------------------------------------
------------
INLINE void INIT_CTRL_VAR(void)
{
direction = 0x0; // Initialise default motor direction to clockwise.
once = 0x0; // Clear the variable 'once' so that by default mot
or stop is assumed.
status_word = 0x0002; // Initialise status flags to reflect motor stopped
condition.
INIT_PI_VAR(); // Execute PI initialisation routine.
MOTOR_START_PARAMS(); // Execute control panel dependent variable initial
isation routine.
}

//*************************************************************************************
************
// @Function void TRAP_RESET(void)
// Description This Function will reset the trap flag if requested.
//-------------------------------------------------------------------------------------
------------
INLINE void TRAP_RESET(void)
{
if (trap_reset == 1) // If trap reset requested...
{
trap_reset = 0; // ... acknowledge the request.
status_ctrap = 0; // ... reset the trap flag.
}
}

//*************************************************************************************
************
// @Function void MOTOR_OPERATION(void)
// Description This Function will start or stop the motor depending on present inp
uts.
//-------------------------------------------------------------------------------------
------------
INLINE void MOTOR_OPERATION(void)
{
if (status_ctrap) // If trap flag is set...
{
motor_status = 1; // ... stop de motor.
}
else // If trap flag is not set...

{
if (motor_status == 0) // If the start of the motor is requested...

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 105

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FUNC.H

{
Motor_start(); // ... start the motor.
}
else // If the start of the motor is not requested..
.
{
Motor_stop(); // ... stop the motor.
}
}
}

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

106

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\CONFIG.H

// Define parameters and configuration

// Speed PI
#define PI_SPEED_YMAX 16368 // Corresponding to an ADC read of 7.5
A (shifted to 1Q15) [ADC range 0 A - 15 A, nominal current = 7 A]
#define PI_SPEED_YMIN 0 // The minimum output of the speed PI i
s 0 A
#define PI_SPEED_KP 358 // Kp has to be divided by 64, to scale
it for the output calculation, where it is multiplied by 64 (0.7 in 1Q15 divided by
64)
#define PI_SPEED_KI 229 // Ki parameter of the speed PI control
ler in per unit (0.007 in 1Q15)

// Current PI
#define PI_CURR_YMAX 32112 // Corresponding to 98% of duty cycle [
0.98 to 1Q15]
#define PI_CURR_YMIN 0 // Corresponding to 0% of duty cycle
#define PI_CURR_KP 205 // Kp has to be divided by 64, to scale
it for the output calculation, where it is multiplied by 64 (0.4 in 1Q15 divided by
64)
#define PI_CURR_KI 16 // Ki parameter of the current PI contr
oller in per unit (0.0005 in 1Q15)

// Motor parameters
#define PP 4 // Pole pairs
#define MAX_SPEED 9000 // Max. speed that the motor can achiev
e without field weakening (No load speed)

// User configuration
#define MOTOR_DIRECTION 0 // Clockwise = 0, Counterclockwise = 1
#define RAMP_UP_TIME 1600000 // Time to achieve the speed end refere
nce (value = time*64*fpwm = 1s*64*25000s-1
#define MOTOR_START_REF 600 // Start reference of the motor speed
#define MOTOR_END_REF 3000 // End reference of the motor speed
#define TIME_CURRENT_MAX 1000 // Maximum time overcurrent (up to 10 A
) is allowed before stopping the motor [TIMER_CURRENT_MAX=tmax/T13CM=40ms/40us]

// Speed calculation
#define SPEED_CONSTANT 15468750 // SPEED_CONSTANT=(60s/min)/(Step size
of T12)

// Sensorless
#define BEMF_AUX_VAL11 0x18 // Phase A inactive (B+ C-)
#define BEMF_AUX_VAL12 0x24 // Phase A inactive (C+ B-)
#define BEMF_AUX_VAL21 0x12 // Phase B inactive (A+ C-)
#define BEMF_AUX_VAL22 0x21 // Phase B inactive (C+ A-)
#define BEMF_AUX_VAL31 0x06 // Phase C inactive (A+ B-)
#define BEMF_AUX_VAL32 0x09 // Phase C inactive (B+ A-)

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 107

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\DSPLIB_KEIL.H

/**********************************************************************************
* Module Header File
* Filename DspLib_Keil.h
* Project DSP library for C166S V2 Core
*----------------------------------------------------------------------------------
* Compiler: Keil
*
* Version: V 1.0
*
* Description:
* This header file lists
* 1. Typedef statements
* 2. All the functions of DSP Library
*----------------------------------------------------------------------------------
* Date: March 2008
**********************************************************************************/

#ifndef _DSPLIB_KEIL_H_
#define _DSPLIB_KEIL_H_

// #include <XC161.h> /* special function register XC161 */

typedef short DataS; //16-bit data

typedef long DataL; //32-bit data

//*********************************************************************************
// This contains the user defined data types for the complex functions
//-----------------------------------------------------------------------------

typedef struct
{
short real;
short imag;
}CplxS;

typedef struct
{
long real;
long imag;
}CplxL;

//****************** Arithmetic Operations ****************************************

extern void CplxAdd_16(
CplxS* X, // first 1Q15 data
CplxS* Y, // second 1Q15 data
CplxS* R // output
);
extern void CplxSub_16(
CplxS* X, // first 1Q15 data
CplxS* Y, // second 1Q15 data
CplxS* R // output
);
extern void CplxMul_16(
CplxS* X, // first 1Q15 data
CplxS* Y, // second 1Q15 data
CplxS* R // result
);
extern DataL Mul_32(
DataL* X, // first 1Q15 data
DataL* Y // second 1Q15 data
);

//***************************** FIR Filters ***************************************

extern DataS Fir_16(
DataS* H, // pointer to filter coefficient vector

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

108

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\DSPLIB_KEIL.H

DataS* IN, // the new input sample

DataS N_h, // Filter order
DataS* D_buffer // pointer to delay buffer
);
extern DataS Fir_32(
DataL* H, // pointer to filter coefficient vector
DataS* IN, // the new input sample
DataS N_h, // Filter order
DataS* D_buffer // pointer to delay buffer
);
extern void Fir_dec(
DataS* x, // pointer to input x
DataS* h, // pointer to coefficent vector h
DataS* y, // pointer to output y
DataS* d_buffer, // pointer to delay buffer
DataS N_x, // size of x
DataS N_h, // filter order
DataS D // decimating factor
);
extern void Fir_inter(
DataS* x, // pointer to input x
DataS* h, // pointer to coefficent vector h
DataS* y, // pointer to output y
DataS* d_buffer, // pointer to delay buffer
DataS N_x, // size of x
DataS N_h, // filter order
DataS I // interpolating factor
);
extern void Fir_lat(
DataS* x, // pointer to input x
DataS* K, // pointer to lattice coefficent vector K
DataS* y, // pointer to output y
DataS* u, // pointer to delay buffer
DataS N_x, // size of x
DataS M // filter nummber of stages
);
extern void Fir_sym(
DataS* x, // pointer to input x
DataS* h, // pointer to coefficent vector h
DataS* y, // pointer to output y
DataS* d_buffer, // pointer to delay buffer
DataS N_x, // size of x
DataS N_h // half of filter order
);
extern void Fir_cplx(
CplxS* x, // pointer to complex input x
CplxS* h, // pointer to complex coefficent vector h
CplxS* y, // pointer to complex output y
DataS* d_buffer, // pointer to delay buffer
DataS N_x, // size of x
DataS N_h // half of filter order
);
extern void Fir_16_Blk(
DataS* H, // pointer to filter coefficient vector
DataS* IN, // the new input sample
DataS* R, //Pointer to output buffer
DataS* D_buffer, // pointer to delay buffer
DataS N_h, // Filter order
DataS N_x
);

//***************************** IIR Filters ***************************************

extern DataS IIR_1(
DataS* B_A, // pointer to filter coefficients
DataS* IN, // the new input sample
DataS N, // Filter order for input X
DataS* x_y // delay buffer
);

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 109

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\DSPLIB_KEIL.H

extern DataS IIR_2(

DataS* h, // pointer to filter coefficients
DataS* IN, // the new input sample
DataS N, // Filter order
DataS* u // state variable vector
);
extern void IIR_bi_1(
DataS* x, // pointer to complex input x
DataS* h, // pointer to complex coefficent vector h
DataS* y, // pointer to complex output y
DataS* u_w, // pointer to state veriables
DataS N_x, // size of x
DataS N_biq // number of the biquads
);
extern DataS IIR_bi1_smpl(
DataS* h, // pointer to complex coefficent vector h
DataS* x, // pointer to complex input x
DataS N_biq, // number of the biquads
DataS* u_w // pointer to state veriables
);
extern void IIR_bi_2(
DataS* x, // pointer to complex input x
DataS* h, // pointer to complex coefficent vector h
DataS* y, // pointer to complex output y
DataS* u, // pointer to state veriables
DataS N_x, // size of x
DataS N_biq // number of the biquads
);
extern DataS IIR_bi2_smpl(
DataS* h, // pointer to complex coefficent vector h
DataS* x, // pointer to complex input x
DataS N_biq, // number of the biquads
DataS* u // pointer to state veriables
);
extern void IIR_bi2_32(
DataS* x, // pointer to complex input x
DataL* h, // pointer to complex 32-bit coefficent vector h
DataS* y, // pointer to complex output y
DataS* u, // pointer to state veriables
DataS N_x, // size of x
DataS N_biq // number of the biquads
);
extern void IIR_lat(
DataS* x, // pointer to input x
DataS* K, // pointer to lattice coefficent vector K
DataS* y, // pointer to output y
DataS* u, // pointer to delay buffer
DataS N_x, // size of x
DataS N_h // filter order
);
extern void IIR_bi_24(
DataS* x, // pointer to complex input x
DataS* h, // pointer to complex coefficent vector h
DataS* y, // pointer to complex output y
DataS* u, // pointer to state veriables
DataS N_x, // size of x
DataS N_biq // number of the biquads
);
extern void IIR_bi_2_Blk(
DataS* IN, // the new input sample
DataS* h, // pointer to filter coefficients
DataS* Y, // the Output sample
DataS* u_w, // state variable vector
DataS N_biq , // number of the biquads
DataS N_x // number of input and output samples
);
//***************************** Mathematical functions ****************************
extern DataS P_series(

Page: 3

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

110

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\DSPLIB_KEIL.H

DataS* a, // pointer to coefficients

DataS* IN, // pointer to the input value
DataS N // the order of series
);
extern void Windowing(
DataS* h, // pointer to coefficients
DataS* x, // pointer to the input value vector
DataS* y, // poniter to output vector
DataS N // the length of the window
);
extern DataS Sine(
DataS x //input
);
extern DataS div_q15(
DataS x, // dividend
DataS y // divisor
);
extern DataS div_q31(
DataL x, // dividend
DataS y // divisor
);
extern DataS Sqrt( DataS x );

//***************************** Matrix ********************************************

extern void Matrix_mul(
DataS* x_1, // pointer to 1. input matrix
DataS* x_2, // pointer to 2. input matrix
DataS* y, // poniter to output matrix
DataS row1, // number of rows in matrix 1
DataS col1, // number of columns in matrix 1
DataS row2, // number of rows in matrix 2
DataS col2 // number of columns in matrix 2
);
extern void Matrix_trans(
DataS* x, // pointer to input matrix
DataS* y, // poniter to output matrix
DataS row, // number of rows in matrix
DataS col // number of columns in matrix
);

//***************************** Adaptive Filters **********************************

extern DataS Adap_filter_16(
DataS* h, // pointer to coeffcient vector h with 16 bits
DataS* IN, // pointer to the new input sample
DataS* D, // pointer to expected signal at time n
DataS* error, // pointer to error signal
DataS N_h, // filter order
DataS Step, // adaptive gain
DataS* d_buffer // delay buffer
);

extern DataS Adap_filter_32(

DataL* h, // pointer to coeffcient vector h with 32 bits
DataS* IN, // pointer to the new input sample
DataS* D, // pointer to expected signal at time n
DataS* error, // pointer to error signal
DataS N_h, // filter order
DataS Step, // adaptive gain
DataS* d_buffer // delay buffer
);

//***************************** FFT Transform *************************************

extern DataS FloatTo1Q15(
float x // floating input value
);
extern float Q15toFloat(
DataS x //1Q15 input

Page: 4

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 111

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\DSPLIB_KEIL.H

);
extern DataS Bit_reverse(
DataS* X, // index input vector
DataS N // the size of the vector
);
extern void real_DIT_FFT(
DataS* x, // inpu vector
DataS* index, // bit reversed input indecies
DataS exp, // exponent of vector size
DataS* table, // trigonomic function table
DataS* X // output of FFT
);
extern void real_DIF_IFFT(
DataS* x, // output of IFFT
DataS* index, // bit reversed input indecies
DataS exp, // exponent of vector size
DataS* table, // trigonomic function table
DataS* X // inpu vector
);
extern void real_DIT_IFFT(
DataS* x, // output of IFFT
DataS* index, // bit reversed input indecies
DataS exp, // exponent of vector size
DataS* table, // trigonomic function table
DataS* X // inpu vector
);
//***************************** Statistical functions **************************
extern void Auto_raw(
DataS* x, // Pointer to input vector in 1Q15 format
DataS* y, // Pointer to output vector in 1Q15 format
DataS N_x, // Size of input vector
DataS N_y // Size of output vector, N_y <= N_x
);
extern void Auto_bias(
DataS* x, // Pointer to input vector in 1Q15 format
DataS* y, // Pointer to output vector in 1Q15 format
DataS N_x, // Size of input vector
DataS N_y // Size of output vector, N_y <= N_x
);
extern void Auto_unbias(
DataS* x, // Pointer to input vector in 1Q15 format
DataS* y, // Pointer to output vector in 1Q15 format
DataS N_x, // Size of input vector
DataS N_y // Size of output vector, N_y <= N_x
);
extern void Cross_raw(
DataS* x1, // Pointer to the first input vector in 1Q15 format
DataS* x2, // Pointer to the second input vector in 1Q15 forma
t
DataS* y, // Pointer to output vector in 1Q15 format
DataS N_x1, // Size of first input vector
DataS N_x2 // Size of second input vector, N_x2 <= N_x1
);
extern void Cross_bias(
DataS* x1, // Pointer to the first input vector in 1Q15 format
DataS* x2, // Pointer to the second input vector in 1Q15 forma
t
DataS* y, // Pointer to output vector in 1Q15 format
DataS N_x1, // Size of first input vector
DataS N_x2 // Size of second input vector, N_x2 <= N_x1
);
extern void Cross_unbias(
DataS* x1, // Pointer to the first input vector in 1Q15 format
DataS* x2, // Pointer to the second input vector in 1Q15 forma
t
DataS* y, // Pointer to output vector in 1Q15 format
DataS N_x1, // Size of first input vector
DataS N_x2 // Size of second input vector, N_x2 <= N_x1

Page: 5

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

112

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\DSPLIB_KEIL.H

);
extern void Convol(
DataS* x, // pointer to vector x
DataS* h, // pointer to input vector x
DataS* y, // pointer to output
DataS* d_buffer, // pointer to delay buffer
DataS N_x, // size of vector x
DataS N_h // size of vector h
);
//------------------- END OF FILE ----------------------------------------------
#endif

Page: 6

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 113

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FloatTo1Q15.c

/******************************************************************************
; Module: FloatTo1Q15
; Filename: FloatTo1Q15.c
; Project: DSP library for C166S V2 Core
;------------------------------------------------------------------------------
; Compiler: Keil
;
; Version: V1.0
;
; Description: Changing the float format to 1Q15 format
;
; Date: Sept 2007
;
; Copyright: Infineon Technologies AG
;*****************************************************************************/

/******************************************************************************
; DataS FloatTo1Q15(float X);
;
; INPUTS:
; X the input value in float format
;
; RETURN:
; Y the output value in 1Q15 format
;
; ALGORITHM: According to the IEEE-754 format
;
; REGISTER USAGE:
; 1. From .c file to .asm file:
; (R8)=mmmmmmmmmmmmmmmm
; (R9)=seeeeeeeemmmmmmm (s:sign; e:exponent; m:mantissa)
;
; 2. From .asm file to .c file:
; (R4)=y
;
; Assumption:
;
;*****************************************************************************/

#include "DspLib_Keil.h"

DataS FloatTo1Q15(float x)
{
__asm
{
//read the exponent bits in R3
MOV R3,R9 ;(R3)=R9
SHL R3,#1h ;(R3)<<1
SHR R3,#8h ;(R3)>>8 = e
SUB R3,#7fh
JMPR cc_ULT,next ;if((R3)<127), jump
//if x=1
MOV R4,#07fffh ;(R4)=32767, in the case x=1
//read sign bit
MOV R2,R9 ;(R2)=(R9)
SHR R2,#0fh ;(R2)=(R2)>>15 =s
JMPR cc_Z,ende
//if x=-1
MOV R4,#8000h ;(R4)=32768, in the case x=-1

RET

next:
//read the mantissa in R4
MOV R4,R9 ;(R4)=(R9)
SHL R4,#9h ;(R4)<<9
SHR R4,#1h ;(R4)>>1
SHR R8,#8h ;(R8)>>8

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

114

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\FloatTo1Q15.c

ADD R4,R8 ;(R4)=(R4)+(R8)

SHR R4,#1h ;(R4)=(R4)>>1
ADD R4,#4000h ;(R4)=(R4)+2^14

//output in 1Q15 format

ADD R3,#1h ;(R3)=(R3)+1
NEG R3 ;(R3)=-(R3)
CMP R3,#0fh
JMPR cc_SLT,next1;if (R3)<15, jump
MOV R4,#0h;

RET

next1:
SHR R4,R3
//read the sign bit R2
MOV R2,R9 ;(R2)=(R9)
SHR R2,#0fh ;(R2)=(R2)>>15 =s
JMPR cc_Z,ende
NEG R4

ende:

RET

}
}
//------------------- END OF FILE ----------------------------------------------

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

 115

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\PI_CTRL.C

//*****************************************************************************/
// @Module PI Controller
// @Filename PI_CTRL.C
// @Controller XC16x Series
//
//
// @Description This file contains functions for PI Control algorithms
//
//****************************************************************************

//****************************************************************************
// @Function
//
// int pi_controller64(long *pi_parameter,int reference,int actual)
//
//----------------------------------------------------------------------------
// @Description
//
// PI-Controller
// derived from transfer function G = kp + 1/(p*Ti)
//
// e(k) = reference - actual T0 =
// yn(k+1)= yn(k) + ki * e(k) ki = T0/Ti
// y(k+1) = yn(k+1) + kp * e(k) * 64
//----------------------------------------------------------------------------
// Computing time 42 CPU-cycle
//
//----------------------------------------------------------------------------
// @Returnvalue
//
//----------------------------------------------------------------------------
// @Parameters
//
//--------------------------- Arguments -------------------------------------
// reference : reference value
// actual : actual value
//
// struct pi_parameter
// {
// int yn; Integral buffer
// int kp; Proportional Constant
// int ki; Integral Constant
// int ymax; Limit value max
// int ymin; Limit value min
// };
//---------------------------------------------------------------------------
// Register modified:
//
// R1,R3,R4,R5,R6,R7,R8,R9,R10
//
//----------------------------------------------------------------------------
//
//****************************************************************************

sfr MCW = 0xFFDC; //MAC Control Word

int pi_controller64(long *pi_parameter,int reference,int actual)

{
long a;
#pragma asm

mov R12,MCW ;Save MCW register

mov MCW,#1536 ;Set saturation and shift left
mov R11,ZEROS ;Load zero in R11
CoLOAD R11,R9 ;Load Accumulator (High) with R9 (reference)
CoSUB R11,R10 ;error = reference - actual
CoSTORE R9 ,MAS ;Load error in R9
mov R3,[R8+] ;

Page: 1

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

116

C:\Users\Pablo\Documents\Universidad\PFC\Software\Definitivo\PI_CTRL.C

CoLOAD R3,[R8+] ;Load yn (integral buffer) in accumulator

mov R1,R8 ;Save parameters addres in R1
mov R5,[R1+] ;Load Kp (proportional Constant) in R5
mov R6,[R1+] ;Load Ki = T0/Ti (integral Constant) in R6
CoMAC R6,R9 ;yn = Ki * error + yn
mov R6,[R1+] ;Load ymax (limit value max)
mov R7,[R1+] ;Load ymin (limit value min)
CoMIN R11,R6 ;Limit max yn
CoMAX R11,R7 ;Limit min yn
CoSTORE R4,MAH ;Store yn-high in R4
CoSTORE R3,MAL ;Store yn-low in R3
mov [-R8],R4 ;Store R4 in integral buffer(High)
mov [-R8],R3 ;Store R3 in integral buffer(Low)
CoMUL R5,R9 ;Kp * error
CoSHL #6 ;64 * Kp * error
CoADD R3,R4 ;y = yn + (64 * Kp * error)
CoMIN R11,R6 ;Limit max y
CoMAX R11,R7 ;Limit min y
CoSTORE R4,MAS ;Store y-high in R4 (return register)
mov MCW,R12 ;Restore MCW register

#pragma endasm

a = reference;
a = actual;
a = *pi_parameter;

Page: 2

APPENDIX A. SOFTWARE STRUCTURE AND FULL CODE

Appendix B.

Flowcharts

B.1. Flowcharts of the control process

In this appendix are displayed the flowcharts of the control process, to

provide an easier way to understand the whole process and the purpose of
different variables and functions.
The flowcharts are sorted by order of appearance if starting the code by
the main.c.
The index of these flowcharts can be found in the list of figures.

117
118 B.1. FLOWCHARTS OF THE CONTROL PROCESS

Main

MAIN_vinit()

ADC_OFFSET()

INIT_CTRL_VAR()

TRAP_RESET()

MOTOR_OPERATION () Motor_stop

RampUp() status_overvoltemp = 1

NO YES
Overvoltage or
overtemperature?

Figure B.1.: Flowchart of Main function

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 119

Main_vInit()

Disable global interrupts

PSW_IEN = 0

MAIN_vUnlockProtecReg()

MAIN_vChangeFreq()

IO_vInit()

CCU60_vInit()

ADC0_vInit()

ADC1_vInit()

MAIN_vLockProtecReg()

Enable global interrupts

PSW_IEN = 1

Return

Figure B.2.: Flowchart of Main_vInit

APPENDIX B. FLOWCHARTS
120 B.1. FLOWCHARTS OF THE CONTROL PROCESS

TRAP_RESET()

trap_reset
requested?
NO

YES

trap_reset = 0
status_ctrap = 0

Return

Figure B.3.: Flowchart of TRAP_RESET

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 121

MOTOR_OPERATION()

Is a trap YES
latched?
Set stop
command
NO

NO Is the motor
ready to start?

YES

Motor_stop() Motor_start()

Return

Figure B.4.: Flowchart of MOTOR_OPERATION

APPENDIX B. FLOWCHARTS
122 B.1. FLOWCHARTS OF THE CONTROL PROCESS

Motor_start()

NO
Is the first time?

YES

once = 1
status_on = 1

Is the start reference NO

higher than the end
reference?

YES

sv_start_ref =
sv_end_ref

sv_start_ref0 = sv_start_ref
speed_ref = sv_start_ref

Init_control()

Commutation_init

Return

Figure B.5.: Flowchart of Motor_start

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 123

Motor_stop()

Disable synchronisation and

commutation request

Disable Trap

Reset T13 and T12

Load 0 % Duty Cycle in T13

Disable Wrong Hall Event

interrupt

Update status_word

Reset variable once

Update direction

return

Figure B.6.: Flowchart of Motor_stop

APPENDIX B. FLOWCHARTS
124 B.1. FLOWCHARTS OF THE CONTROL PROCESS

Init_control

Initialise the speed buffer and

speed related variables

Initialise bemf variables

Initialise control loop variables

Initialise overcurrent counter

Return

Figure B.7.: Flowchart of Init_control

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 125

Commutation_init()

Load CC61 and CC62 register values

Load T13 with a 0 % Duty Cycle

Load the actual pattern

Load the expected pattern

Enable WHE interrupts

Run T13

Select shadow transfer at CC61 compare event,

synchronised to T13 zero match

Run T12

Enable Trap

Return

Figure B.8.: Flowchart of Commutation_init

APPENDIX B. FLOWCHARTS
126 B.1. FLOWCHARTS OF THE CONTROL PROCESS

RampUp()

NO
Is the period time
reached?

YES

Increment counter

NO Is the time step

reached?

YES

Speed_ref_ramp()

Reset counter

NO
Is there a new end
reference?

YES

Calculate the new

difference

Calculate new time

step

Save new end

reference

Return

Figure B.9.: Flowchart of RampUp

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 127

Speed_ref_ramp

NO
Is the end
reference bigger?

YES

Increase
speed
reference

NO
Is the end
reference lower?

YES

Decrease
speed
reference

Return

Figure B.10.: Flowchart of Speed_ref_ramp

APPENDIX B. FLOWCHARTS
128 B.1. FLOWCHARTS OF THE CONTROL PROCESS

CCU60_viNodeI1

NO
Is CCU60_IS_T12PM?

YES

status_timeout = 1

Motor_stop()

NO
Is CCU60_IS_TRPF

YES

status_ctrap = 1

Motor_stop()

NO
Is CCU60_IS_WHE?

YES

status_hs_error = 1

YES
Is possible to change to
sensorless?

NO
1
Motor_stop()

Return

Figure B.11.: Flowchart of CCU60_viNodeI1

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 129

Save actual
time T12

YES Is BEMF already

detected in this
sector?

NO

Reconfigure
Reconfigure registers
registers

Reset T12 to half

Reset T12 to actual sector time + actual
value - half sector time value

Set CC61 register for Is the YES

next commutation commutation
done?

NO

Force
commutation

Is the next YES

pattern
loaded?

NO

Load next
pattern

Return

Figure B.12.: Flowchart of Transition

APPENDIX B. FLOWCHARTS
130 B.1. FLOWCHARTS OF THE CONTROL PROCESS

CCU60_viNodeI2

NO
Is CCU60_IS_ICC62R

YES

Load next
pattern

Speed_calc

Have we detected YES

falling edge before?

Set Rising
NO edge
detection

YES
Have we detected a
rising edge before?
Set falling
NO edge
detection

Is the speed large NO

enough

Reset speed
YES
to 0

Return

Figure B.13.: Flowchart of CCU60_viNodeI2

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 131

Speed_calc()

Accumulate
new measure

Substract old
sample

Is the motor YES

starting?

NO

Calculate speed Calculate speed

based on the based on this
average sector

Update the
buffer index

Return

Figure B.14.: Flowchart of Speed_calc

APPENDIX B. FLOWCHARTS
132 B.1. FLOWCHARTS OF THE CONTROL PROCESS

CCU60_viNodeI3

T13 Compare NO
Match?
YES

Select non-fed phase to

convert

Convert selected phase

and DC current

Request DC link voltage

conversion and set the
compare value to half
the DC link voltage

NO
T13 Period
Match

YES

Is BEMF already NO
detected?

Sensorless_calc
YES

Do the
control loops

Update Duty
Cycle

Return

Figure B.15.: Flowchart of CCU60_viNodeI3

APPENDIX B. FLOWCHARTS
B.1. FLOWCHARTS OF THE CONTROL PROCESS 133

Sensorless_calc()

Apply the offset to the

converted channel

NO Is the slope
known?
YES
NO Is BEMF ZC
detected in the NO
Rising
assumed slope? edge?

YES YES

BEMF NO NO BEMF
NO Is a detected? detected?
reasonable
value? YES YES

YES Status_bemf = 1 Status_bemf = 1

Slope_known = 1
Save zc time Save zc time
Slope = 3 Slope = 2

Sensorless NO NO Sensorless
mode? mode?

YES YES

Set the compare registers for Set the compare registers for
the next commutation and the next commutation and
speed measurement speed measurement

Return

Figure B.16.: Flowchart of Sensorless_calc

APPENDIX B. FLOWCHARTS
134 B.1. FLOWCHARTS OF THE CONTROL PROCESS

ADC0_viSRN0

NO Value out of the

limits?
YES

Is it between a NO
tolerance limit?
YES

Increase the counter

NO Maximal time
exceeded?

YES

Status_overcurrent = 1

Motor_stop()

Return

Figure B.17.: Flowchart of ADC0_viSRN0

APPENDIX B. FLOWCHARTS