Lecture 7: Feedback

Dr. Riad Kanan

 Introduction

 IN THIS LECTURE YOU WILL LEARN

 The general structure of the negative-feedback amplifier and
the basic principle that underlies its operation.
 The advantages of negative feedback, how these come about,
and at what cost.
 The appropriate feedback topology to employ with : voltage
type
 Introduction

 Most physical systems incorporate some sort of

feedback.
 Although theory of negative feedback was developed by
electrical engineers.
 Harold Black with Western Electric Company
(Feedback amplifier 1928).
 Feedback can be negative (degenerative) or positive
(regenerative).
 Introduction

 Feedback may be used to:

 desensitize the gain
 reduce nonlinear distortion
 reduce the effect of noise
 control the input and output resistances
 extend bandwidth
 These characteristics result, however, in loss of gain.
 “The basic idea of negative feedback is to trade-off gain for
other desirable properties.”
 Introduction

 Under certain conditions, negative feedback can be

come positive.
 This causes oscillation.
 However, positive feedback does not always lead to
instability.
 Regenerative feedback has a number of applications –
specifically, in active filtering.
 The General
Feedback Structure

 Figure 1. shows the basic structure of a feedback

amplifier – signal-flow diagram.
 Open-loop amplifier has gain A (xo = Axi).

Figure 1: General structure of the feedback amplifier. This is a signal-flow diagram,

and the quantities x represent either voltage or current signals.
 The General
Feedback Structure

 Output (xo) is fed to load as well as feedback network.

 Feedback factor (b.) defines feedback signal (xf).
 Feedback signal (xf) is subtracted from input (xi).
 This characterizes negative feedback.
 Gain of feedback amplifier is defined in (1).
 Note that (1) may be approximated at 1/b..
 As such, gain of feedback amplifier is almost entirely
determined by feedback network.
 The General
Feedback Structure

xo A
(eq10.4) gain with feedback: Af  
1
xi  
 
1   Ab  
 loop 
 gain 

amount of
feedback

1
(eq10.5) if assumed that Ab  1: Af 
2
b
 The General
Feedback Structure

3 Ab
(eq10.6) feedback signal: x f  xs
1  Ab

Ab
(eq10.7) input signal: xi  x s  x f  x s 
4 xs
1  Ab
 Ab 
5
(eq10.6) input signal: xi   1   xs
 1  Ab 
 1 
(eq10.6)
6 input signal: xi    xs
 1  Ab 
 Some Properties of
Negative Feedback

 Gain De-sensitivity
dA
dA f ddd  A 
 A  dA ff A 
AAf f  dA AAdA 1  Ab 
f
AA A
AA
dA
dA dA dA11AAbb
1  Ab  Ab 
11AAbbAAbbAA  1
A
 22   A
 1
11AAbb 
 Ab
2
 1  Ab  2

AA  A
AA 111 
 
 A
1  Ab A 1  Ab 
11AAbb AA 11AAbb 
A
AAff f
 Some Properties of
Negative Feedback

 Gain De-sensitivity
 Relative change in feedback gain =
relative change in open loop gain 
sensitivity function
 Sensitivity function usually < 1 Af A  1 

Af A  1  Ab 
Sensitivity
Function
 Some Properties of
Negative Feedback

 Gain De-sensitivity
 Example
Numbers:

xi A  A  100  30
vs R1 xo
 R1  10 k

R2 R 2  1 k
 xi A 
Some Properties of vs R1 xo

Negative Feedback
R2

A  100  30
 Gain De-sensitivity R1  10 k
 Example R 2  1 k
R2 1
b 
R1  R2 11
– Feedback (loop) gain
100 A 100
T  Ab  9 Af    10
11 1  Ab 1  9
– Sensitivity
1 1  Af 1 30
S    A f  10  0.3
1  Ab 10 10 10 100
 Some Properties of
Negative Feedback

 Bandwidth Extension

xs xi xo xo  A  s  xi
A
 A0
A s 
1  s
b H
 Some Properties of
Negative Feedback

 Bandwidth Extension

 Compute feedback gain

AA00
A s  11 ss AA00
Af  s  A f  s   H
H
 ss
1  b A  s  11 bb AA00
HH 11 bb AA00
11 ss
HH
A0 A
10 1
 
1  b A0 1  s1 b1A0b 1A  s
   1 b A 
 Some Properties of
Negative Feedback Af  s 
A0
H  1  b A0
s

 Bandwidth Extension
 A0  1 
Af  s  
1  b A0  1  s  1 b A 
  H 0 
 Low Freq  Hf 
 Gain A f 0   
 
 Feedback trades gain for increased bandwidth
 Some Properties of
Negative Feedback

 Bandwidth Extension

 Gain  bandwidth conserved

 A0 
Hf
AAff 00 Hf     H 1  b A0  
 1  b A0 
 A0 H
 Some Properties of
Negative Feedback

 Bandwidth Extension
 Example

1  106 b
1
Af  s  
11
A s  
1 s 11 1 s
4 3.64  105
 Some Properties of
Negative Feedback

 Bandwidth Extension
 Example 6
Bode Diagram
10

4
10
Magnitude (abs)

2
10

 1
0
10 b
58 kHz
-2 0 2 4 6
10 10 10 10 10
Frequency (Hz)
 The Four Basic
Feedback Topologies

 Voltage Amplifiers
 Current Amplifiers
 Trans-conductance Amplifiers
 Trans-resistance Amplifiers
 Voltage Amplifiers

 voltage amplifiers – accept input voltage and yield

output voltage.
 VCVS
 Thevenin Output
 voltage-mixing / voltage-sampling – is the topology
most suitable for voltage amps.
 Is also known as series-shunt feedback.
 Provides high input resistance/low output resistance.
Figure 9.6: Block diagram of a feedback voltage amplifier. Here the appropriate
feedback topology is series–shunt.
 Voltage Amplifiers

 Increased input resistance results because Vf subtracts

from Vs, resulting in smaller signal Vi at the input.
 Low Vi causes input current to be smaller.
 This effects higher input resistance.
 Decrease output resistance results because feedback
works to keep Vo as constant as possible.
 Vo and Io change / vary together.
 This effects lower output resistance.
 Voltage Amplifiers

ii
 Input resistance 
Ri
Ro 
vs vi 
Av i vo
Rif 



g mvi
ii vs 

vs vs ii Rii Ri  v f  0 0 Rif

vf 
 b vo
bbvovo
bbAvAvi i
 bb Ai
AiiiR
Rii
Rif  Ri 1  Ab 
vs  ii Ri  b Aii Ri  0
Series mixing increases input resistance
 Voltage Amplifiers

 Output resistance  io

vo vi Ri
Rof  
 Ro Av i vo
io vs  
vo vo  Avi
io   g mvi 
Ro Ro Rof

 b vo

vi   b vo
vo 1  Ab 
Ro
Rof 
io  1  Ab
Ro
Shunt sampling decreases output resistance
Figure 2: Examples of a feedback voltage amplifier. All these circuits employ series–
shunt feedback. Note that the dc bias circuits are only partially shown.
Summary: Series-shunt
connection

Series-shunt

A
– Gain decreased Af 
1  Ab

– Input resistance increased Rif  Ri 1  Ab 

Ro
– Output resistance Rof 
decreased 1  Ab
 Summary

 1. Ri and Ro are the input and output resistances,

respectively, of the A circuit.
 2. Rif and Rof are the input and output resistances,
respectively, of the feedback amplifier, including Rs and
RL
 Current Amplifiers

 current amplifier – accepts input current to generate

output current.
 CCCS
 Norton Source
 current-mixing / current-sampling – topology is most
suitable for current amps.
 Is also known as shunt-series feedback.
 Provides low input resistance/high output resistance.
 Figure 3: (a) Block diagram of a feedback current amplifier. Here, the appropriate
feedback topology is the shunt–series. (b) Example of a feedback current amplifier.
 Transconductance
Amplifiers

 transconductance amplifier – accepts input voltage and

generates output current.
 VCCS
 Norton Source Output
 voltage-mixing / current-sampling – topology is most
suitable for transconductance amps.
 Is also known as series-series feedback.
 Provides high input resistance/high output resistance.
Figure 4: (a) Block diagram of a feedback transconductance amplifier. Here, the
appropriate feedback topology is series–series. (b) Example of a feedback
transconductance amplifier. (c) Another example.
 Transresistance
Amplifiers

 transresistance amplifier – accepts input current and

generates output voltage.
 CCVS
 Thevenin Source Output
 current-mixing / voltage-sampling – topology is most
suitable for current amps.
 Is also known as shunt-shunt feedback.
 Provides low input resistance/low output resistance.
 Figure 4: (a) Block diagram of a feedback transresistance amplifier. Here, the
appropriate feedback topology is shunt–shunt. (b), and (c) Examples of feedback
transresistance amplifiers.
 Determining Loop
Gain

Table 1
 Summary of
Feedback Analysis
Method

 Always begin analysis by determining an approximate value for

the closed-loop gain (Af).
 Assume that loop gain Ab is large.
 Af = 1/b
 This value should serve for final check on Af.
 The shunt connection at input or output will always result in
reducing the corresponding resistance.
 In utilizing negative feedback to improve the properties of an
amplifier under design, the starting point is selection of feedback
topology.
 Feedback factor (b.) may be determined as 1/Af.
 Summary

 Negative feedback is employed to make the amplifier gain less

sensitive to component variations; to control input and output
impedances; to extend bandwidth; to reduce nonlinear distortion;
and to enhance signal-to-interference ratio.
 The advantages above are obtained at the expense of a reduction
in gain and at the risk of the amplifier becoming unstable (that is,
oscillating). The latter problem is solved by careful design.
 For each of the four basic types of amplifier, there is an
appropriate feedback topology. The four topologies, together
with their analysis procedures, are summarized in Table 1.
 Summary

 The key feedback parameter are the loop gain (Ab.), which for
negative feedback must be a positive dimensionless number, and
the amount of feedback (1+Ab.). The latter directly determines
gain reduction, gain desensitivity, bandwidth extension, and
changes in input and output resistances.