Macro Models of Amplifiers, Op Amps, Comparators, Multipliers, FETs and BJTs

Amplifiers

Ideal amplifiers are controlled voltage/current sources and are two-port active networks. Ideal
amplifiers have infinite power gain. Infinite power gain can be achieved either by having zero input
voltage (short circuit at the input) or zero input current (open circuit at the input) and finite output
(voltage or current). We, therefore, identify four types of ideal amplifiers. Voltage Controlled
Voltage Source (VCVS) also known as Voltage Amplifier, is represented as

 Ii   0 0 Vi 
 =  
V 0  gf 0  I 0

The equivalent network and input-output relationship of an ideal voltage amplifier is shown in the
figure 1

Ii
+ Io +
+

Vi Vo
- -

Fig. 1: Ideal voltage amplifier

Voltage Controlled Current Source (VCCS), also known as Transconductance Amplifier is

 Ii   0 0  Vi 
 =  
 I 0 yf 0 V 0

The equivalent network and input-output relationship of an ideal transconductance amplifier is

Io
+ Ii +
Vi y fV i Vo
- -

Fig. 2: Ideal trans-conductance amplifier

Current Controlled Current Source (CCCS), also known as Current Amplifier, is represented as

1
 Vi   0 0  Ii 
 =  
 I 0 hf 0 V 0

The equivalent network and input-output relationship of an ideal current amplifier is shown in the
figure 3.

Ii Io +
+
Vi A IIi Vo
- -

Fig.3: Ideal current amplifier

Current Controlled Voltage Source (CCVS), also known as Trans-resistance Amplifier, is

 Vi   0 0  Ii 
 =  
VO   zf 0  IO 

The equivalent network and input-output relationship of an ideal trans-resistance amplifier is

Ii Io
+ +
+

Vi ZfIi Vo
- -

Fig. 4: Ideal Trans-resistance amplifier

While it is possible to realize all the four amplifiers as integrated circuits, the most dominant
amplifier commercially is Voltage Amplifier as wide band amplifier, audio amplifiers,
instrumentation amplifiers and voltage controlled amplifiers. This is more due to historical reasons
rather than any technical reasons. While amplifiers are available commercially, they are not that
widely used. Operational Amplifiers, presented in the following section, are capable of meeting
practically all requirements of analog signal processing.

Op Amps

When the transfer parameter ( g f , y f , h f or z f ) of an amplifier goes to infinity, it is called Ideal

Operational Amplifiers. They are called so as they are used to perform several mathematical

2
operations including addition, subtraction, integration and differentiation. They are referred to as
Ideal OP Amps. The input-output characteristic of an ideal Op Amp is shown in the figure 5.

Fig. 5: Input-output characteristic of an ideal Op Amp

For any finite output, the input voltage and input current are zero. Such behaviour can only be
modelled by a pair of elements called nullator and norator. Nullator represents the input element

with Vi = 0 , I i = 0 . Norator represents the output element that provides any finite output voltage

and current. The ideal Op Amp is represented by a nullator-norator pair in the figure 6.

Fig. 6: Ideal Op Amp as nullator-norator pair

Every nullator has its companion norator. A nullator along with its companion norator is called
‘nullor’. An ideal Operational Amplifier is a Differential Input Differential Output Amplifier (DIDO)
with infinite gain. An ideal Operational Amplifier with one of its output (norator) terminals
grounded is a Differential Input Single Ended Output Amplifier (DISO) with infinite gain. Similarly,
an Ideal Operational Amplifier with one of its input (nullator) and one of its output (norator)
terminals grounded is a Single Input Single Ended Output Amplifier (SISO) with infinite gain.
DIDO and DISO are four terminal elements, and SISO is a three terminal element. These three
configurations of ideal operational amplifiers are shown in the figure 7.

3
 Fig. 7: Nullator-norator representation of DIDO, DISO and SISO amplifiers

Commercial Op Amps are designed with BJTs and JFETs or BJTs and MOSFETs (which will be
introduced in later sections) as active devices. Op Amp requires single or dual DC power supply
voltages for it to satisfactorily operate in the active region, with output restricted to values in
between the power supply voltages.

Non-ideal (commercial) Op Amps have finite, but large, dc gain, gain dependent on the frequency
and signal level (non-linearity), off-set voltages and currents, finite input impedance, and non-zero
output impedance. We will now explore how these non-idealities influence the behaviour of Op
Amps. All the parameters of an Op Amp also depend on the temperature.

Non-ideal Op Amps, henceforth referred to as Op Amps, have finite gain ranging from 103 to 106
(60 dB to 120dB). Op Amps have several independent parasitic capacitances associated with
internal nodes, with each capacitance reducing the gain at 20dB/decade. The gain transfer
function of an Op Amp can be approximated in the range where magnitude of the gain is more
than one (above 0dB) is

 s 
A0 1 + 
A=  ωz 
 s  s  s 
1 +  1 +  1 + 
 ω p1   ω p 2   ω p 3 

where A0 is the dc gain, and ω p1 < ω p 2 < ω p 3 are the corner frequencies. If ω z is absent, each

corner frequency causes, when fully effective, 20 dB/decade attenuation and 900 phase shift. At

the corner frequency, the magnitude reduces by 3 dB and phase shift becomes 450 (at ω1 as long

as ω2 and ω3 are much greater than ω1 ). At ω2 the phase shift can become 90
0
+ 450 , and at

ω3 it can become 180

phase shift becomes 180 . At frequencies greater than ω3 , phase shift approaches 270 . If ω z is
0 0

present the Op Amp will be designed for it to be greater than ω p1 . If ω z is present the phase shift

may never reach 180O, depending on its value in relation to the other corner frequencies.

All the active devices of an Op Amp may go into saturation as the output voltage levels approach

∆V0 ∆V0
the supply voltages of the Op Amp, making the gain ( ) go toward zero. is maximum
∆Vi ∆Vi

4
  V + +V −  +
when the output voltage is at   , where V is the positive supply voltage and V− is the
 2 
∆V0
negative supply voltage. approaches zero as the output approaches V + or V − .
∆Vi
The mismatches at the inputs cause input offset voltage, which could be of the order of a few milli
volts to a few tens of milli volts, and input offset current, which could be a few nano amperes in
case of Op Amps designed with BJTs. If the Op Amp is a voltage amplifier, the input impedance
can be from a few MΩ to a few hundred MΩ. The output impedance can be few hundred Ohms. The
output current of majority Op Amps is limited to 20 mA.

The maximum rate of change of output of an Op Amp is called ‘slew rate’ and is expressed as
Volts/µsec. It is one to 20 Volts/µsec with the commercially available Op Amps. It arises out of
limiting current available for charging the capacitor at the output. All commercially available Op
Amps are protected against short circuit at the output. This limits the output current to 20 to 50
mA.

Op Amps are also characterized by their gain-bandwidth (first corner frequency) (GB) product, a
parameter that we will use extensively when we design Op Amp based circuits. It is typically1MHz
to 20 MHz.

TL081C is an operational amplifier introduced in 1977 by Texas Instruments and costs about $
0.20 when procured in multiples of thousands. It is meant for use in the commercial temperature
range (0 to 70O C). The important parameters of TL081C are

Parameter Value

1. Total Supply Voltage 7 to 36 Volts

All the parameters are defined for +15V

2. Gain-Bandwidth Product at 25OC 3 MHz

3. Slew Rate 13 V/sec

4. CMRR 70 dB

5. Input Offset Voltage 20mV(max)

6. Input Offset Voltage Temperature Coefficient 18V/OC

7. Input Offset Current 2 nA (max)

8. Input Bias Current 10 nA (max)

9. Input Resistance 1012 Ohms

10. Output Resistance 200 Ohms

5
Comparators
Comparator is an interface component. Its input is analog and its output is digital (1/0 or
high/low). The output of the comparator changes its state when the input voltage crosses, while
decreasing or increasing, a reference value. The input-output characteristics of ideal comparators
are shown in the figure 10.

Vi
+

-
V ref

a. Voltage Comparator

IR
- Vo

+
Ii

b. Current comparator

Fig. 10: Ideal comparator

An ideal comparator, as can be seen from the characteristic, will have infinite gain in the active
region while transiting from one state to the other.

LM311 is a comparator made available by most of the IC manufacturers, and costs about $ 0.20
when procured in multiples of thousands. It is meant for use in the commercial temperature range
(0 to 70O C). The important parameters of LM311 are

Parameter Value

1. Total Supply Voltage 3.5 to 30 Volts

2. Rise Time 115 ns

3. Common Mode Input Voltage 0.5 to 28 V

4. Input Offset Voltage 7.5 mV(max)

5. Input Offset Current 70 nA (max)

6. Input Bias Current 300 nA (max)

6
Even when the forward gain is infinity, the behaviour of a comparator differs from that of an ideal
one because of input offset voltage (7.5 mV in case of 311). The output will be as shown in the
figure 11.

Vi +
∞ Vo
V ref -

Fig. 11: Comparator with infinite forward gain and with offset voltage

The commercial comparators are designed to have a forward gain of the order of 100 rather than
the high values (> 106) associated with Op Amp. If the forward gain of the comparator is finite,

 Vu − VL 
the active transition region will be   instead of zero in case of infinite gain. The input-
 A0 
output relationship is shown in the figure 12.

Vi +
A Vo
V ref -

Vu − VL
Vi offset + ⇒ input referred offset
Ao

Fig, 12: Comparator with finite gain and Offset voltage

Multipliers

Multipliers are analog components that provide multiplication of two input voltages or currents.
Multipliers can perform a variety of signal processing applications including mixing, modulating,
and demodulating. They can also act as voltage controlled amplifiers, filters and oscillators, and
phase detectors. Multipliers are available commercially in several forms including voltage
controlled amplifier, current controlled amplifier and digitally controlled amplifier. A multiplier is
shown schematically in the figure 13.

7
 Fig. 13: Multiplier

The input-output relationship of a multiplier can be written as

V0 = Voffset + K X VX + KY VY + K 0VX VY

where Voffset is the DC offset voltage and K X VX and KYVY are feed through components and K 0
is the multiplier constant and has the dimension per volts.

A precision multiplier like MPY634 has a transfer function

 ( X − X 2 )(Y1 − Y2 ) 
V0 = A  1 − ( Z1 − Z 2 )  volts with -10V < X, Y < 10V
 10 
The dynamic ranges of inputs and output are compatible with one another. Some of the important
characteristics of MPY634 are

Parameter Value

1. Bandwidth 6 MHz (min)

2. Slew rate 20V/sec

3. Output Offset Voltage +100 mV(max)

4. Output Short Circuit Current 30 mA

Transconducting Devices: FETs and BJTs

An ideal voltage controlled current source (VCCS) is also known as ‘transconductance amplifier’.
The input-output relationship a transconductance amplifier is given by

 Ii   0 0   Vi   0 0  Vi 
 =  =  
 I 0   yf 0  V 0   gm 0 V 0

8
If a transconductor is designed with a given value of gm then it becomes an amplifier with a fixed
gain. gm has to be made a function of one of two variables available to us, namely ‘output current’
or ‘input voltage’. A device that provides a transfer parameter that is linearly varying with output
current or input voltage will meet all analog signal processing functions. The device should,
therefore, provide a gm as given by

∂I 0
gm = = KVi or KI 0
∂Vi
The first relationship gives

∂I 0 = KVi ∂Vi
where VT is known as Threshold Voltage.
I 0 = K (Vi − VT ) 2
giving a square law relationship between output current and input voltage.

The second relationship gives

∂I 0
= K ∂Vi
I0
ln I 0 = KVi where IS is known as Reverse Saturation Current

(
I 0 = I S ε KVi − 1)
giving an exponential relationship between the output current and input voltage. There are two
semiconductor devices, Field Effect Transistors (FETs) and Bipolar Junction Transistors (BJTs) that
exhibit these relationships. FET exhibits a square law relationship in the region above threshold
voltage, and an exponential relationship in sub-threshold region (Vi < VT). BJT exhibits
exponential relationship.

The field-effect transistor was first patented by Julius Edgar Lilienfeld in 1926 and by Oskar Heil in
1934, but practical semiconducting devices (the JFET) were developed only much later. The
MSOFET (Metal Oxide on Semiconductor Field Effect Transistor), which largely superseded the JFET
and had a more profound effect on electronic development, was invented by Dawon Kahng and
Martin Atalla in 1960. The bipolar point-contact transistor was invented in December 1947 at the
Bell Telephone Laboratories by John Bardeen and Walter Brattain under the direction of William
Shockley. The junction version known as the bipolar junction transistor, invented by Shockley in
1948, enjoyed three decades as the device of choice in the design of discrete and integrated
circuits.

At present discrete MOSFETS are not commercially made available because of problems associated
with electro static discharge. JFET are available, but their use is not popular in signal processing.
MOSFETs and MOSFET technology is dominant in both digital and analog integrated circuits. While
discrete BJTs were commercially made available for several decades, their usage at present in
signal processing functions has practically stopped. This was mainly due to ready availability of Op
Amps and the requirement of smaller footprints for the electronic systems. Discrete semiconductor
devices at present are mainly available as power devices including Power MOSFETs and IGBTs.

9
Integrated circuits today are predominantly use CMOS (Complementary MOSFET) technology and
BiCMOS technology to a limited extent. All digital integrated circuits are manufactured using
CMOS technology, and some mixed signal circuits are made with BiCMOS technology. Some Op
Amps based on bipolar devices
vices are still produced today because of their popularity with the users.

Knowledge
nowledge of behaviour (macro model) of FETs and BJTs will be useful to an electronics engineer
in designing some power electronic and mixed mode signal circuits.

Field Effect Transistors: FET is a four terminal device with one terminal as ‘source’, the second
terminal as substrate ‘drain’, the third terminal ‘gate’ and the fourth terminal ‘substrate’. The
voltage applied to the gate with respect to the substrate controls the current between the ‘source’
and the ‘drain’. The gate is isolated from the channel by a reverse biased junction or by an
insulator known as gate oxide. The FET with reverse biased junction between the
th gate and channel
is known as JFET (Junction FET). The FET with the insulation between gate and channel is called a
MOSFET (Multi Oxide Semiconductor FET) or insulated Gate FET. The channels in JFET and MOSFET
can be either n-type
type channel or p
p-type channel.
l. In the case of JFET since it is a junction that
isolates the gate from channel, the channel is always controlled in depletion mode. In the case of
MOSFET, the isolation between channel and gate is through insulation, the channel can be
controlled by either
her polarity of gate voltage, that is, control is achieved through either depletion or
enhancement. Therefore, there are six types of FETs, normally,
- n-channel JFET
- p-channel JFET
- n-channel
channel depletion mode FET
- p-channel
channel depletion mode FET
- n-channel enhancement mode FET
- p-channel
channel enhancement mode FET
Enhancement mode FETs are the preferred devices as they are normally off
off-devices
devices (no channel
between source and drain) when no voltage is applied to gate with respect to the substrate.

Depletion mode FET technology would have been the natural choice for analog ICs because the
devices are in the active region with zero DC bias. However, as the enhancement mode FET
technology is used mainly for digital ICs, the same technology is also used for analog ICs in view
of higher reliabilities, yields and used for a single technology for mixed signal processing ICs. The
structure of n-channel
channel enhancement MSOFET is shown in the figure 14.

Fig. 14: n
n-channel enhancement MOSFET

10
 channel enhancement): A voltage applied to the gate with respect to the substrate
FET (n-channel
which is positive to cause a channel to exist between source and drain is known as threshold

voltage, VT when a voltage higher than current VT = VGS − VT is applied

saturation through the channel increases in proportion to square of the voltage

I DS = K (VGS − VT )
2
. The large signal macro model of the device is shown in figure 15.

D
G
+
V GS IDS=K(V GS-V T) 2
-
S

Fig. 15: Large signal macro model of FET

Small signal macro model of FET is shown in the figure 16.

D
G
+
∆V GS gm∆V GS
-
S

Fig. 16: Small signal macro model of FET

Fig. 17: Nullator

Bipolar Junction Transistor:: BJT consists of a heavily doped (n+ or p+) region called ‘emitter’
where carriers are injected onto another region called ‘base’ by forward biased emitter based
junction which is made up of p or n and is very thin so as to prevent any recombination of injected
carriers. Injected carriers are immediately collected by the ‘collector’ (n or p region) aided by
electric filed created by the reverse biased collector
collector-base junction. The structure of a BJT is
illustrated in the figure 17.

11
 Fig. 17: Structure of a BJT
The collector current IC = α x emitter current (IE) with a being very near to 1 (typically equal to
0.995)

I E = I EO ∈( BE
V VT )

I C = α I EO ∈( BE
V VT )

∂IC I KT
= g m = C ;VT = (for IC = 1 mA gm = 40 mS)
∂VBE VT q

α
= β (about 200)
1−α

∂VBE ∂VBE ∂VBE

The small signal micro model of a BJT is shown in the figure 18.

1
C
B + gm(1-α)
Vi gmV i
-

Fig. 18: Small signal model of a BJT

The ideal nullator-norator model of BJT is given in the figure 19.

Fig. 18: Nullator-norator model of BJT

12