VLSI Design and Testing 1 Introduction to CMOS Technology

HKBK COLLEGE OF ENGINEERING

Department of Electronics and Communication Engg.

VLSI Design and Testing(BEC602)

Introduction to CMOS Circuits

1 Introduction to CMOS Technology

Complementary Metal-Oxide-Semiconductor (CMOS) technology has become a fundamental
part of the modern integrated circuit (IC) industry.

Although CMOS is widely used today, its origins date back nearly a century.

1.1 Historical Background

The concept of the MOS field-effect transistor (MOSFET) was first proposed by J. Lilienfeld
in 1925, with a similar structure later suggested by O. Heil in 1935.

Early attempts to develop MOS transistors faced material-related challenges, leading to the
invention of the bipolar junction transistor (BJT), which became the dominant technology for
many years.

The MOS transistor gained renewed interest with the advent of the silicon planar process in
the early 1960s.

However, quality control and material challenges delayed its commercial adoption until around
1967.

Initially, only single-polarity MOS transistors (p-type or n-type) were commonly used.

CMOS, which utilizes both p-type and n-type transistors on the same substrate, was first
applied to ultra-low-power applications such as digital watches.

Due to the complexity of CMOS fabrication, it was initially less favored in general system
designs. However, as nMOS processing technology became more intricate, the relative
complexity of CMOS became less of a concern.

1.2 The Growing Importance of CMOS

The demand for low-power, high-density ICs led to a surge in CMOS adoption.

System designers faced increasing challenges with chip size and power consumption, making
CMOS an attractive alternative.

Today, CMOS has become the dominant technology for Very Large-Scale Integration (VLSI)
circuit design.

1
VLSI Design and Testing 2 MOS Transistors

2 MOS Transistors
A MOS (Metal-Oxide-Silicon) transistor is a fundamental semiconductor device used in
modern integrated circuits.

It is formed by layering conducting, insulating, and transistor-forming materials on a silicon

substrate.

The key structural elements include diffusion regions, polysilicon layers, and metal
interconnections, all separated by insulating layers.

2.1 CMOS Technology and Transistor Types

CMOS (Complementary MOS) technology utilizes two types of MOS transistors:

nMOS (n-type MOSFET) – An n-type transistor is fabricated on a p-type silicon

substrate, with two n-type diffused regions acting as the source and drain.

pMOS (p-type MOSFET) – A p-type transistor is fabricated on an n-type silicon

substrate, with two p-type diffused regions serving as the source and drain.

The doping of the silicon substrate determines the type of charge carriers:

nMOS transistors use electrons as the majority carriers (negatively charged).

pMOS transistors use holes as the majority carriers (positively charged).

2.2 Physical Structure and Components

A typical MOS transistor consists of the following components:

Source (S): One of the two terminals where current enters or exits.

Drain (D): The other terminal where current enters or exits.

Gate (G): A conducting electrode (typically polysilicon) placed over a thin insulating layer
(oxide) that controls current flow.

Substrate (Body): The silicon region in which the device is fabricated, either p-type for
nMOS or n-type for pMOS.

For an nMOS transistor, the structure consists of a p-type substrate separating two n-type
diffusion regions. A pMOS transistor has an n-type substrate with two p-type diffusion
regions. In both cases, the gate electrode sits above the channel region, separated by an
insulating oxide layer.

Dr. Sufia Banu, ECE, HKBKCE 2

VLSI Design and Testing 2 MOS Transistors

Figure 1: MOS transistor physical structures

2.3 Operation of MOS Transistors

The MOS transistor operates by controlling the flow of current between the source and drain
terminals using the gate voltage.
Applying a positive voltage to the gate of an nMOS transistor attracts electrons, creating a
conductive channel between the source and drain.
Applying a negative voltage to the gate of a pMOS transistor attracts holes, enabling
conduction between the source and drain.
The source and drain terminals are interchangeable, and their designation depends on the direction
of current flow. The gate acts as a control terminal, regulating the electrical connection between
the drain and source.

2.4 Conclusion
MOS transistors are the building blocks of digital and analog circuits. The combination of nMOS
and pMOS transistors in CMOS technology enables low-power, high-speed circuits used in

Dr. Sufia Banu, ECE, HKBKCE 3

VLSI Design and Testing 3 MOS Transistor Switches

microprocessors, memory chips, and various electronic devices.

3 MOS Transistor Switches

MOS (Metal-Oxide-Semiconductor) transistors act as controlled switches in digital circuits. The
gate terminal determines whether the switch is ON or OFF. We assume:
A high voltage (‘1’) is normally set to 5V and is called POWER (VDD).
A low voltage (‘0’) is normally set to 0V and is called GROUND (VSS).

The strength of a signal is measured by its ability to source or sink current.

3.1 nMOS Transistor as a Switch

An nMOS transistor operates as follows:
Gate = ‘1’ (VDD): The switch is ON (closed), allowing current flow.
Gate = ‘0’ (VSS): The switch is OFF (open), blocking current flow.

Passing Signals
Good for passing ‘0’ (LOW signal).

Imperfect for passing ‘1’ (HIGH signal is degraded due to threshold voltage drop
VDD − Vth).

Figure 2: nMOS transistor as a switch

3.2 pMOS Transistor as a Switch

A pMOS transistor operates as follows:
Gate = ‘0’ (VSS): The switch is ON (closed), allowing current flow.
Gate = ‘1’ (VDD): The switch is OFF (open), blocking current flow.

Dr. Sufia Banu, ECE, HKBKCE 4

VLSI Design and Testing 3 MOS Transistor Switches

Passing Signals
Good for passing ‘1’ (HIGH signal).

Imperfect for passing ‘0’ (LOW signal is degraded).

Figure 3: pMOS transistor as a switch

3.3 Transmission Gate (Complementary Switch)

A transmission gate consists of an nMOS and a pMOS transistor connected in parallel. It is
controlled by:
Control signal (C) applied to nMOS.

Complementary control signal (C) applied to pMOS.

Figure 4: Transmission Gate (Complementary Switch)

Advantages of Transmission Gates

Overcomes the limitations of individual nMOS and pMOS switches.
Provides low resistance for both ‘0’ and ‘1’ signals.
Used in multiplexers, flip-flops, and pass-transistor logic circuits.

Dr. Sufia Banu, ECE, HKBKCE 5

VLSI Design and Testing 4 CMOS Logic

3.4 Summary of Switch Behavior

Table 1: Summary of Switch Behavior

Switch Type Gate Signal State Passes ‘0’ Well Passes ‘1’ Well
1 (VDD) ON Yes No (degraded ‘1’)
nMOS
0 (VSS) OFF No No
0 (VSS) ON No (degraded ’0’) Yes
pMOS
1 (VDD) OFF No No
Transmission Gate C = 1, C = 0 ON Yes Yes

3.5 Conclusion
MOS transistors serve as voltage-controlled switches in digital circuits:

nMOS transistors efficiently pass ‘0’ signals but degrade ‘1’ signals.

pMOS transistors efficiently pass ‘1’ signals but degrade ‘0’ signals.

Transmission gates (TG) provide optimal switching by combining nMOS and pMOS
transistors.

These principles are widely applied in CMOS logic circuits, multiplexers, and memory
circuits.

4 CMOS Logic
4.1 The Inverter
A CMOS inverter is a fundamental building block of digital circuits, implementing the logical
NOT function.

The logical function of an inverter follows the truth table:

Table 2: Truth table of an inverter

Input (A) Output (Y)

0 1
1 0

From the table, we observe:

– When the input is ‘0‘, the output must be ‘1‘. This requires a P-SWITCH (PMOS
transistor) to connect the output to VDD.

Dr. Sufia Banu, ECE, HKBKCE 6

VLSI Design and Testing 4 CMOS Logic

– When the input is ‘1‘, the output must be ‘0‘. This requires an N-SWITCH (NMOS
transistor) to connect the output to VSS.

Based on the above observations, the inverter consists of:

– A PMOS transistor connected between the output and VDD (pull-up network).
– An NMOS transistor connected between the output and VSS (pull-down network).

When the input is ‘0‘, the PMOS transistor turns ON, pulling the output to VDD.
Simultaneously, the NMOS transistor remains OFF, ensuring no direct path between VDD and
VSS.

Conversely, when the input is ‘1‘, the NMOS transistor turns ON, pulling the output to VSS,
while the PMOS transistor remains OFF.

Transistor-Level CMOS Inverter

A fully complementary CMOS gate follows this principle, where:

– A pull-down network (NMOS transistors) connects the output to VSS when required.
– A pull-up network (PMOS transistors) connects the output to VDD when required.

A Y

VDD

Figure 5: CMOS Inverter

This design ensures low power dissipation, as there is no direct current path between VDD
and VSS during steady-state operation.

4.2 Combinational logic

Combinational logic in CMOS is implemented using N-SWITCHES (NMOS transistors) and
P-SWITCHES (PMOS transistors).

By arranging these transistors in series or parallel, different logical functions can be realized.

Dr. Sufia Banu, ECE, HKBKCE 7

VLSI Design and Testing 4 CMOS Logic

4.2.1 AND Function

If two N-SWITCHES are placed in series, as shown in figure below, the resulting composite
switch is closed (ON) only when both inputs are set to ‘1’.

This implements an AND function.

Figure 6: AND function using NMOS transistors in series

Similarly, the corresponding structure for P-SWITCHES is shown in figure below. The
composite switch is ON only when both inputs are set to ‘0’.

Figure 7: AND function using PMOS transistors in series

4.2.2 OR Function
When two N-SWITCHES are placed in parallel, the composite switch is ON if either input is
set to ‘1’, implementing an OR function.

In contrast, when two P-SWITCHES are placed in parallel, the composite switch is OFF if
both inputs are set to ‘1’.

Dr. Sufia Banu, ECE, HKBKCE 8

VLSI Design and Testing 4 CMOS Logic

Figure 8: OR function using NMOS and PMOS transistors in parallel

4.2.3 CMOS Combinational Gates

By combining these series and parallel switch structures, CMOS logic gates such as NAND, NOR,
and more complex logic functions can be designed.

4.3 The NAND gate

The NAND gate is a fundamental digital logic gate that implements the NOT AND
operation.
It is widely used in digital circuit design due to its universal property, meaning any logic
function can be realized using only NAND gates.

4.3.1 Construction of a Two - input CMOS NAND gate

Figure below illustrates the construction of a 2-input NAND gate using CMOS technology.
The pull-down network (PDN) is based on the AND function (A.B). The presence of a
‘0’ in the truth table requires an AND structure for the n-channel MOSFETs (NMOS).
The pull-up network (PUN) follows De Morgan’s Theorem, forming a parallel
p-channel MOSFET (PMOS) OR structure that implements:
A·B =A+B
The PMOS network is the logical dual of the NMOS network, ensuring proper functionality.

Dr. Sufia Banu, ECE, HKBKCE 9

VLSI Design and Testing 4 CMOS Logic

CMOS NAND Circuit Diagram

VDD

Y
A

Figure 9: Two - input CMOS NAND gate

Truth Table of a NAND Gate

Table 3: Truth Table of a Two - input NAND Gate

A B N1 N2 P1 P2 Y
0 0 OFF OFF ON ON 1
0 1 OFF ON ON OFF 1
1 0 ON OFF OFF ON 1
1 1 ON ON OFF OFF 0

Y =A·B
A
Y
B

4.3.2 Key Observations

1. Full Voltage Swing: - There is always a path from either VDD (logic 1) or GND (logic
0) to the output, ensuring full logic levels. - This makes CMOS logic fully restoring.

2. No Ratioing Required: - Unlike nMOS logic, where transistor sizes need careful ratioing,
CMOS gates do not require it.

3. No Static Power Dissipation: - There is never a direct path between VDD and GND for
any input combination, minimizing power consumption.

Dr. Sufia Banu, ECE, HKBKCE 10

VLSI Design and Testing 4 CMOS Logic

4. Extending to Multi-Input NAND Gates: Larger NAND gates are constructed by:
Adding NMOS transistors in series in the pull-down network.
Adding PMOS transistors in parallel in the pull-up network.

VDD

A n

Figure 10: n - input CMOS NAND gate

Y = A · B · C······ n

4.3.3 Conclusion
The CMOS NAND gate is a fundamental building block in digital logic design.
It provides advantages such as low power dissipation, full voltage swing, and
scalability.
Understanding its construction is essential for designing efficient digital circuits.

4.4 The NOR gate

The NOR gate is a fundamental digital logic gate that implements the NOT OR operation.
It is widely used in digital circuits and serves as a universal gate, meaning any logic function
can be realized using only NOR gates.

Dr. Sufia Banu, ECE, HKBKCE 11

VLSI Design and Testing 4 CMOS Logic

4.4.1 Construction of a Two - input CMOS NOR Gate

Figure below illustrates the construction of a 2-input NOR gate using CMOS technology.
The pull-down network (PDN) follows the OR function (A + B). The presence of a ‘1’ in
the truth table requires an OR structure for the n-channel MOSFETs (NMOS), meaning
NMOS transistors are in parallel.

The pull-up network (PUN) follows De Morgan’s Theorem, forming a series p-channel
MOSFET (PMOS) AND structure that implements:

A+B =A·B

The PMOS network is the logical dual of the NMOS network, ensuring proper functionality.

CMOS NOR Circuit Diagram

B
Y

Figure 11: Two-input CMOS NOR gate

Truth Table of a NOR Gate

Table 4: Truth Table of a Two-input NOR Gate

A B N1 N2 P1 P2 Y
0 0 OFF OFF ON ON 1
0 1 OFF ON ON OFF 0
1 0 ON OFF OFF ON 0
1 1 ON ON OFF OFF 0

Y =A+B

Dr. Sufia Banu, ECE, HKBKCE 12

VLSI Design and Testing 4 CMOS Logic

A
Y
B

4.4.2 Key Observations

1. Full Voltage Swing: - There is always a path from either VDD (logic 1) or GND (logic
0) to the output, ensuring full logic levels.

2. No Ratioing Required: - Unlike nMOS logic, where transistor sizes need careful ratioing,
CMOS gates do not require it.

3. No Static Power Dissipation: - There is never a direct path between VDD and GND for
any input combination, minimizing power consumption.

4. Extending to Multi-Input NOR Gates: - Larger NOR gates are constructed by:

Adding NMOS transistors in parallel in the pull-down network.

Adding PMOS transistors in series in the pull-up network.

VDD

A n

Figure 12: n - input CMOS NOR gate

Y = A + B + C + ········ + n

Dr. Sufia Banu, ECE, HKBKCE 13

VLSI Design and Testing 4 CMOS Logic

4.4.3 Conclusion
The CMOS NOR gate is a fundamental building block in digital logic design.

It provides advantages such as low power dissipation, full voltage swing, and
scalability.
Understanding its construction is essential for designing efficient digital circuits.

4.5 Compound gates

A compound gate is formed by combining series and parallel transistor structures to
implement a complex Boolean function in CMOS logic.
Instead of using multiple basic gates, a compound gate directly maps the logic function to a
transistor-level implementation, improving speed and reducing power consumption.

4.5.1 Example 1: CMOS Implementation of F = (A.B) + (C.D)

The function F = (A.B) + (C.D) can be implemented using CMOS transistors by following these
steps:
Pull - Down Network (PDN): The nmos network will consist of:
– A and B in series, let us call this as S1.
– C and D in series, let us call this as S2.
– S1 and S2 in parallel.
Pull - Up Network (PUN): The pmos network will consist of:
– A and B in parallel, let us call this as P1.
– C and D in parallel, let us call this as P2.
– P1 and P2 in series.

Figure 13: CMOS implementation of F = (A.B) + (C.D)

Dr. Sufia Banu, ECE, HKBKCE 14

VLSI Design and Testing 4 CMOS Logic

4.5.2 Example 2: CMOS Implementation of F = (A + B + C).D

The function F = (A + B + C).D can be implemented using CMOS transistors by following these
steps:
Pull - Down Network (PDN): The nmos network will consist of:

– A, B, and C in parallel, let us call this as P1.

– D in series with P1.

Pull - Up Network (PUN): The pmos network will consist of:

– A, B, and C in series, let us call this as S1.

– D in parallel with S1.

VDD

Figure 14: CMOS implementation of F = (A + B + C).D

4.6 Multiplexers
Complementary switches may be used to select between a number of inputs, thus forming a
multiplexer function.

Figure below shows a connection diagram for a 2 - input multiplexer.

Dr. Sufia Banu, ECE, HKBKCE 15

VLSI Design and Testing 4 CMOS Logic

Figure 15: 2 - input Multiplexer

As the switches have to pass ‘0’s and ‘1’s equally well, complementary switches with n- and
p-transistors are used.

The truth table for the structure is shown in table below:

Table 5: Truth Table of a Two-input Multiplexer

S S A B OUTPUT
0 1 X 0 0(B)
0 1 X 1 1(B)
1 0 0 X 0(A)
1 0 1 X 1(A)

The complementary switch is also called a transmission gate or pass gate (complementary).

A commonly used circuit symbol for the transmission gate is shown in figure below.

Figure 16: Symbols of Transmission gate

Dr. Sufia Banu, ECE, HKBKCE 16

VLSI Design and Testing 4 CMOS Logic

The multiplexer connection in terms of this symbol and transistor symbols is shown in figure
below.

Figure 17: 2 - input Multiplexer using Transmission gates

4.7 Memory
Memory elements are essential components in digital circuits for storing and retaining data.

Using basic CMOS structures, we can construct a simple memory element such as a flip-flop
with minimal components.

4.7.1 Flip-Flop Using a Multiplexer and Inverters

A simple flip-flop can be designed using a 2-input multiplexer and two inverters, as shown
in figure below.

Figure 18: Flip - flop

This circuit operates based on a load signal (LD) to control data storage.

Dr. Sufia Banu, ECE, HKBKCE 17

VLSI Design and Testing 4 CMOS Logic

Write Mode (LD = 1)

When the load signal is high (LD = 1):

– The output Q is directly set to the input D.

– This allows new data to be stored in the flip-flop.

This operation is illustrated in figure below.

Figure 19: Flip - flop in Write Mode

Hold Mode (LD = 0)

When the load signal is low (LD = 0):

– The multiplexer switches to a feedback loop, connecting the output back to itself through
the inverters.
– This feedback maintains the stored value, effectively holding the previous state of Q,
while the input D is ignored.

This operation is illustrated in figure below.

Figure 20: Flip - flop in Hold Mode

4.7.2 Conclusion
This simple circuit demonstrates how memory elements can be built using fundamental CMOS
components.

Flip-flops like this serve as the foundation for registers, latches, and memory units in
digital systems.

Dr. Sufia Banu, ECE, HKBKCE 18

VLSI Design and Testing 5 Alternate Circuit representations

5 Alternate Circuit representations

Generally, a design can be expressed in terms of:
1. Behavioral representation
2. Structural representation
3. Physical representation

5.1 Behavioral representation

A behavioral representation defines how a system or circuit responds to a given set of
inputs.
This representation focuses on the functionality of a system rather than its implementation
details, making it independent of the underlying technology.

5.1.1 Behavioral Specification at the Logic Level

At the logic level, the behavior of a digital circuit can be described using Boolean functions.
For example, the behavior of a logic gate can be expressed as:

F = ((A + B + C) · D)

This Boolean function specifies the logical operation without indicating how it is implemented
in hardware or its performance characteristics.

5.1.2 Higher-Level Behavioral Descriptions

Behavioral representation can also be expressed at a higher level using arithmetic or logical
operations.
For instance, an addition operation can be described in a high-level language as:
sum = a + b

Here, no specific method of addition is implied, and the word length is assumed to be that of
the machine.

5.1.3 Behavioral Representation of Sequential Circuits

For sequential circuits, behavior can be described using conditional statements.
Consider a flip-flop, where the output is updated based on the load signal (LD):
IF ( LD == 1)
THEN Q = D;

This representation, however, may be ambiguous because it could also describe a

multiplexer, which selects inputs without necessarily storing a state.

Dr. Sufia Banu, ECE, HKBKCE 19

VLSI Design and Testing 5 Alternate Circuit representations

5.1.4 Higher Levels of Behavioral Specification

More abstract behavioral descriptions can specify:

– Types of registers used in a design.

– Data transfers between these registers.

These descriptions provide even less information about implementation details but help define
how the system should function.

Eventually, behavior can be described as an algorithm in a high-level programming language.

5.1.5 Importance in Modern Design Systems

The objective of modern digital design tools is to convert high-level behavioral
specifications into optimized hardware designs efficiently.

This process ensures:

– Faster design time.

– Increased accuracy and reliability.

By using behavioral representations, designers can develop complex digital systems while
focusing on functionality before deciding on implementation details.

5.2 Structural representation

A structural specification defines how components are interconnected to perform a function
or achieve a specific behavior.

Unlike behavioral descriptions, which focus on logical operations, structural descriptions

specify the physical arrangement of circuit elements.

One example of a structural description language is MODEL, developed by Lattice Logic Ltd.
This language provides a formal way to define circuit components and their interconnections.

5.2.1 Structural Representation in MODEL

In MODEL, circuit elements such as transistors are explicitly defined along with their
connections.

The general syntax follows this format:

Part <circuit_ name > ( < inputs >) -> <outputs >
<Compone nt_Type > <drain > <gate > <source >
End

Dr. Sufia Banu, ECE, HKBKCE 20

VLSI Design and Testing 5 Alternate Circuit representations

Example 1: Inverter Description in MODEL

Part inv ( in) -> out

Nfet out in vss
Pfet out in vdd
End

The first line defines a part named inv with input in and output out.

The Nfet transistor has its drain = out, gate = in, and source = vss.

The Pfet transistor has its drain = out, gate = in, and source = vdd.

Example 2: 2-Input NAND Gate in MODEL

Nfet out b i1
Pfet out a vdd
Pfet out b vdd
End

An internal signal i1 is declared to facilitate the connection between transistors.

Two Nfet transistors form a series connection, while two Pfet transistors are in parallel,
implementing a NAND function.

The corresponding Boolean equation is:

out =∼ (a&b)

vdd
a b

out
b
i1
a

vss
Figure 21: Graphical representation of structural description for a Two - input CMOS NAND gate

Dr. Sufia Banu, ECE, HKBKCE 21

VLSI Design and Testing 5 Alternate Circuit representations

5.2.2 Advantages of Structural Representation

1. Explicit Connectivity: The entire transistor-level structure is specified.

2. Performance Optimization: Allows modifications such as transistor sizing and capacitance

adjustments.

3. Hierarchical Design: Smaller components can be combined to build complex circuits.

Example 3: Adding Performance Parameters

b 100

End

Transistor Sizing: size = 2 increases the size of Pfets, affecting speed and power.

Capacitance Values: Specified in arbitrary units to account for circuit delay effects.

5.2.3 Structural Representation of Complex Circuits

By using smaller predefined components, we can create more complex circuits.

Example 4: Transmission Gate in MODEL

Part tg (a, c, cb) -> b

Nfet a c b
Pfet a cb b
End

A transmission gate consists of an Nfet and a Pfet, controlled by complementary signals (c and
cb).

Example 5: Flip-Flop (D Latch) Using Structural Components

Dr. Sufia Banu, ECE, HKBKCE 22

VLSI Design and Testing 5 Alternate Circuit representations

End

Figure 22: Schematic representation of CMOS flip-flop

The flip-flop (D latch) is implemented using transmission gates (tg) and inverters (inv).

This hierarchical design approach enables scalability and reuse of components.

5.2.4 Structural vs. Behavioral Representation

Table 6: Comparison of Structural and Behavioral Representation

Feature Structural Representation Behavioral Representation

Focus Transistor-level connectivity Logical function
Example Nfet out in vss out = (a & b)
Performance Details Yes (e.g., capacitance, size) No
Flexibility Parameterized descriptions possible Limited parameterization
Readability More detailed, hardware-specific More abstract, easier to understand

Behavioral descriptions ensure correct logic implementation, but structural descriptions define
real circuit performance.

5.2.5 Combining Structural and Graphical Representations

Structural descriptions (MODEL) provide a text-based, parameterized method for
defining circuits.

Graphical descriptions (schematics) visually represent circuit connectivity.

Emerging design tools integrate both approaches for flexibility and efficiency.

Dr. Sufia Banu, ECE, HKBKCE 23

VLSI Design and Testing 5 Alternate Circuit representations

Example: Parameterized Inverter

Part inv ( in) [n] - > out

Nfet out in vss size = n
Pfet out in vdd size = 2 * n
End

The size parameter n allows dynamic transistor scaling, useful for design automation.

5.2.6 Conclusion
Structural representation is essential in circuit design for detailed connectivity,
performance optimization, and hierarchical modeling.

By combining structural and behavioral descriptions, engineers can achieve both logical
correctness and physical efficiency in circuit design.

5.3 Physical representation

The physical specification for a circuit is used to define how a particular part must be
constructed to yield a specific structure and, consequently, a defined behavior.

In an Integrated Circuit (IC) process, the lowest level of physical specification is the
photo-mask information, which is crucial for the various processing steps during
fabrication.

At this stage, we focus on a simplified model for the physical nature of a CMOS circuit.

5.3.1 Transistor Physical Representation

A typical physical representation for a transistor involves two rectangles, representing the
lithography required for the transistor’s fabrication.

These rectangles have precise dimensions defined by the design rules, which are based on the
specific process being used.

These rules often change for different processes, and the corresponding dimensions may not
change linearly.

Rather than focusing on these complex rules, we use a single symbol to represent a transistor
in a non-metric format, maintaining the essential physical nature of the transistor.

n-Transistor representation
The physical symbol for an n-transistor is shown in figure below.

In n-transistor, two process levels are overlaid: one for the gate connection and another for the
source and drain.

These symbols are placed on a grid where:

Dr. Sufia Banu, ECE, HKBKCE 24

VLSI Design and Testing 5 Alternate Circuit representations

– The center grid point is for the gate.

– The grid point to the right (or above) is the drain.
– The grid point to the left (or below) is the source.

These grid points can be visualized as part of a schematic layout.

Figure 23: Physical Representation of n-Transistor

p-Transistor representation
Similarly, a p-transistor uses a similar symbol, as shown in figure below. The “horizontal”
transistor layout is used here, with the gate, source, and drain points similarly defined on the
grid.

Figure 24: Physical Representation of p-Transistor

Dr. Sufia Banu, ECE, HKBKCE 25

VLSI Design and Testing 5 Alternate Circuit representations

5.3.2 Physical Symbolic Layout for an Inverter

A symbolic layout for an inverter can be constructed using the transistor symbols.

Figure 25: Physical Representation of CMOS inverter

It resembles the schematic layout but requires careful consideration of the layers in which
connections are made.
The interaction of these layers is summarized in table below:
– OK denotes that a connection is possible between two layers.
– X signifies that a direct connection is not allowed, requiring a “contact” (C) to connect
the two layers.

Table 7: Physical Layer Interactions in CMOS Design

Physical Layer n-Diffusion p-Diffusion Polysilicon Aluminum

n-Diffusion OK X Transistor OK (C)
p-Diffusion X OK Transistor OK (C)
Polysilicon Transistor Transistor OK OK (C)
Aluminum OK (C) OK (C) OK (C) OK

5.3.3 Transmission Gate Layout

The symbolic layout for a transmission gate is shown in figure below.
This layout is composed of the overlaid n- and p-transistor symbols, with grid points
connecting the appropriate terminals.

Dr. Sufia Banu, ECE, HKBKCE 26

VLSI Design and Testing 5 Alternate Circuit representations

Figure 26: Symbolic Layout for Transmission Gate

5.3.4 Building a Flip-Flop

Using the principles described, a physical sub-assembly for a flip-flop can be constructed.

Figure below shows the symbolic layout for a flip-flop.

Figure 27: Symbolic Layout for Flip-Flop

This layout combines multiple transmission gates and inverters, with appropriate connections
for Vss and Vdd supplies.

Dr. Sufia Banu, ECE, HKBKCE 27

VLSI Design and Testing 6 CMOS-nMOS comparison

5.3.5 Conclusion
CMOS IC design involves several critical steps:

1. Defining the behavior of the circuit.

2. Designing the logic that implements the behavior.

3. Translating this logic into a transistor-level description.

4. Finally, creating a physical layout for the designed logic.

6 CMOS-nMOS comparison

Table 8: Comparison of CMOS and nMOS Logic

Feature CMOS nMOS

Logic Levels Fully restored logic; output settles at Output does not settle at GND, leading
VDD or VSS (GND). to degraded noise margin.
Transition Rise and fall times are of the same or- Rise times are inherently slower than fall
Times der. times.
Transmission Passes both logic levels well; output can Pass transistor transfers logic ‘0’ well, but
Gates drive other transmission gates. logic ‘1’ is degraded. Cannot drive a sec-
ond pass transistor.
Power Dissi- Almost zero static power dissipation; Power dissipated in the circuit even when
pation power dissipated only during logic tran- output is stable, in addition to switching
sition. losses.
Precharging Both n-type and p-type devices can With enhancement-mode transistors, the
Characteris- precharge a bus to VDD or VSS. best achievable precharge is (VDD − Vt).
tics Bootstrapping or hot clocking is often re-
quired.
Power Supply Voltage required to switch a gate is a Somewhat dependent on supply voltage;
fixed percentage of VDD; variable range fixed.
from 1.5V to 15V.
Packing Den- Requires 2N devices for N -input com- Requires (N + 1) devices for N -input
sity plementary static gates; fewer for dy- gates.
namic gates.
Pull-up to Load-to-driver ratio typically 2:1. Load-to-enhancement-driver ratio opti-
Pull-down mized for logic ‘0’ level and minimal cur-
Ratio rent consumption.
Layout Encourages regular layout styles. Depletion load and different driver tran-
sistor sizes inhibit layout regularity.

Dr. Sufia Banu, ECE, HKBKCE 28