Feedback
Feedback
Feedback
Feedback in Amplifiers
10.1 INTRODUCTION
feedback implies transfer of energy from the output of a system to its input. If a
h e term feedba
he or the whole) of the output signal of an amplifier is fed back and superimposed on the
Pt signal, the performance of the amplifier changes significantly. The amplifier is then said
a feedback
amplifier.)
to be
Peedback can be of two types negative feedback and positive feedback. When the feed-
Lk Sinal diminishes the magnitude of the input signal, the feedback is called negative,
innerse or degenerative feedback. Negative feedback reduces the overall gain of the amplifier.
referredto
Ifthe feedback signal enhances the magnitude of the input signal, the feedback isincreased
as positive,
direct or regenerative feedback. The overall gain of the amplifier is by
positive feedback.)
offers many
Despite the reduction in the overall gain of the amplifier, negative feedback
advantages, and is therefore used extensively in electronic circuits. On the contrary, positive
There-
feedback causes an instability in the gain of the amplifier and distorts the input signal.
sometimes in narrowband
fore, it is not generally used in amplifiers. It is used in oscillators and
amplifiers and radio receivers because it increases the signal power.
he whole system in the dashed box in Fig. 10.1 constitutes the feedback amnlke
ifier. The
transfer gain A, of the feedback amplifier is theratio of the output voltage V. to thee
applied input signal voltage V, i.e.,
externally
4- S (10.1)
IfV =V,-V, we have
V = AV, = A(V,- V.
Since V=BV Eq. (10.2) gives 10.2
V.(1+Aß) =
AV, or,
AV,1+ AB
A. Thus
(10.3)
A represents the transfer.
In the absence of any feedback ß 0,= and A, =
gain
without feedback. The quantity-A Bis termed theloopga!n, because1t gives the product of
the gains of the branches making up the loop. It is also called the feedback factor, the ref
return
ratio the loop transmission.)
or
The difference between unity and the loop gain, i.e. (1 +Aß) is referred to as the retum
difference. The gain A, with feedback is sometimes called the closed-loop gain. The quantity A
is the gain without feedback, but încluding the loading effect of the feedback network. A i
is
known as the open-loop gain.
In Eq. (10.3), both A and ß can be real (positive or negative) quantities or complex funec.
tions of frequency.
If I 1+AB I> 1, we have I A, I < IA 1, which characterises negative feedback. If
1+AßI < 1, we obtain | A, l> l'A I, which gives positive feedback. If I 1+AßI =0.
1Al= o, implying that the amplifier gives an output signal even in the absence of an input
signal; the amplifier then becomes an oscillator.
The feedback introduced into an amplifier is usually expressed in decibel (dB) by the rela-
tionship
F dB of feedback = 20
log0 20 lo810 11+ Ap
20 los10+AB (10.4)
For the negative feedback, Fis negative since I 1 +A Bl>1 for such a feedback. When t
feedbackis positive, I1+ABl <1, and Fis positive.
10.3 FEEDBACK AMPLIFIER TOPOLOGIES
The output signal in a feedback amplifer can be sampled by connecting the feedback network
either in shunt with the output(eoltage sampling) or in series with the output (currentsam
pling). In eachtype of sampling, the feedback signal can be returned to the input in series with
the externally applied signal (series mixing), or in shunt with the external signal (shunt
ing). This givesfour feedback topologies: (i) voltage-series feedback or voltage sampling-series
mixing, (iü) voltage-shunt feedback or voltage sampling-shunt mixing, (ii) current-series feed:
back or current sampling-series mixing, and (iv) current shunt feedback or current sampling
shunt mixing.
A) Voltage-series Feedback
Fig. 10.2 shows schematically the voltage-seriesfeedback configuration. The emitter follower
circuit of Fig. 10.3 is an illustration of voltage-series negative feedback In Fig. 10.3, the
outpue
voltage Vo appearing across the load resistance R, is tatally returned to the input. Conse
cater. ioThens F e e
hack in Amplifiers
a
a 223
xternaly ntly,
feedback ratio B is unity here. In Fig. 10.3 the emitter-base.junction is forward
ho
th
through the resistance g The capacitance C decouples the signal source from the
bia V and behaves as a short cireuit Tor ac) The output voltage is thesampled signal. The
signal being a voltage, the feedback is voltage or shunt feedback. The feedback signal
(10.1 Sampe returned to the input in series with the external input signal voltage, so that we have
voltagng, So the feedback in the emitter follower circuit is voltage-series or uvoltage-voltage
or shunt-series feedback.
cc
(103)
sfer gain
Internal
oduct of
e returm amplifier Cin
B
A
e return
Feedback
k.antityAisA RLVo
network
3
x func
Feedback amplifier A
Fig. 10.2 Voltage-series feedback Fig. 10.3 An emitter follower circuit illustrating
configuration. voltage-series negative feedback.
ack. If
The input signal voltage to the transistor in Fig. 10.3 is V, = V, - Vo, where V, is the/
input externally applied inputsignal voltage. This shows that the feedback is negative. Sincefpr
negative feedback I 1 + ABI>1,theoverall voltage gain is
rela- A
(10.5)
A1+
AB
The gain A of the amplifier without feedback can be found by transferring the load resis-
(104 tance R from the emitter to the collector leg. Thus
en the A he R A (10.6)
(see Sec 8.9). As B = 1 here. Eq. (10.5) gives
h M
iwork
he Bz (10.7)
sam
Ah+ he R
with Equation (10.7) shows that the voltage gain is less than unity. This expression of the gain,
obtained from the concept of feedback, agrees with Eq (8.79) derived earlier.
erres Voltage-shunt Feedback
feed Fig. 10.4 depicts the block diagram of an amplifier with voltage-shunt feedback. An illustration
ng of this type of feedback is given by the circuit of Fig. 10.5. Here the transistor is in the CE
configuration and the resistor R, provides the feedback from the collector to the base of the
transistor. i'he capacitor C, isolates the sigrial source from the supply Ve and acts as a short
circuit for ac. The output voltage V, is much greaterthan the input voltage V, and 180° out of
wer
phase with it. An increase of the input signal voltage V, increases the current 1 Consequently,
both , and 1, will increase because ,
that the circuit provides
+1 The increase in , is less than that of , showing
a , negative feedback"
=
Electronics-Fundamentals and Applicot.
224 lications
+Vcc
R
w
Internal C
Vso amplifier A Rs
w
s
V
Vs
Feedback
network B
Feedback amplifier A
Fig. 10.5 Circuit showing
10.4 Block diagram of
Fig.
voltage-shunt feedback.
voltage-shunt-feedbáck.
The feedback current is
(10.8)
R R
to the output voltage V shom
since V>» V, Thus the feedback current I, is proportional
that the feedback is voltage-shunt.
Current-series Feedback
current series feedback amplifier is schematically
shown in the block diagram of Fig. 106
A
o +Vcc
Internal
amplifier A
o C
Rs V
Feedback
V RE
network B
Feedback amplifier A
the feedback signal is the voltage Ve Hence, the circuit is referred to as the current-series
feedback. Following the arguments put forward for an emitter follower, we find that the feed
OVoc
Rc
Internal
amplifier A Rs
Is
w Vo
Feedback
network B REaVa
Feedback amplifier A
Fig. 10.8 Schematic representation of
Fig. 10.9 A circuit showing current-shunt feedback.
current-shunt feedback.
(10.9)
R R
since V>> V. Neglecting the base current of the transistor T, compared with its collector
current we have Ve2 = (-1,) RE2. So, we get from Eq. (10.9)
R
A
Or, RE21,
1,(R+R=1,Rg2 R+Rga (10.10)
The feedback current 1, being proportional to the output current 1, the feedback here is
current-shunt or current-current or series-shunt.
Stability of Gain is
The overall gain of the negative feedback amplifier
O
A 1+ AB (10.11
where I1+A Bl > 1. If lA Bl >>1, Eq. (10.11)gives
A (10.12)
Thus the overall gain of the amplifier is determined by the feedback network and noa
the gain A of the internal amplifier. As the feedback network usually consists of stable passi
passive
elements, the stability of the gain is improved. Supply voltage variations, changes of Daran
eters of the active device, aging and temperature changes will not have significant effects on d
on
have marked effects on A.
A
although they can
If the loop gain in a negative feedback amplifier is not large, the stability of the gain e
can
be found by differentiating Eq. (10.11) with respect to A. Thus we get
dA 1 dA
1+ Aß A
(10.13)
A
The ratio of the fractional change in the overall gain to the fractional change in the gain
of the internal amplifier gives the sensitivity S of the transfer gain to changes in the internal
amplifier gain. Thus
(10.14)
D 1+Aß (10.15)
(10.16)
As IDI> 1 for negative feedback, the percentage change in A,is less than that in A. The
overall gain stability is thus better.
pecrease in Nonlinear Distortion
rhe nonlinearity in the transfer characteristic of the amplifier active device distorts the outpu
signal when a large-amplitude signal is applied to the amplifier input. Let V and V,betne
For
signal component ofthe utput voltage and the distortion output voltage, respectively
simplicity, we assume here that the input signal voltage is sinusoidal and that the distorao
voltage is the second harmonic.
he
IFV, is the applied inputsignalvoltage, we have V, = AV, in the absence offeedback. 1
a*
voltage V, is introduced by the nonlinearity of the active device, it can be assumed to
Feedback in Amplifiers 227
gainA.
Vp
NA 1+ Aß1+ Aß (10.17)
Since 11+A Bl> 1 for negative feedback, the distortion voltage is reducedat the output
by a factor 1(1 + A B). Note thatt the signal component of the output voltage remains un
changed because the amplitude of the applied input signal is enhanced by (1 +A B); this en-
hancement can be achieved by using a preamplifier. Clearly, the dynamic range or the signal
handling capacity of the amplifier is increased.
(i)Effect on Input and Output Impedances
(a) Input impedance: In the series-mixing feedback
of the
circuit Fig. 10.10,
ternal amplifier is
input impedance of the in
Internal
amplifier A
, (10.18)
if o of
The input impedance of the feedback amplifier
1S Feedback
network B
(10.19) Feedback amplifier, A,
The externally applied input signal voltage is Fig. 10.10 Calculation of impedances for
V=V, +V=V, + pV.=V,+A pV (10.20) voltage-series feedback.
since V = AV, where A is the voltage gain of the internal amplifier including the load resis-
tance R. Equation (10.20) yields
V, = V, (1+ A B) (10.21)
So, Eq. (10.19) gives
Z V, (1+ A) 2,1+Aß) (10.22)
(10.24)
have
From Kirchhoffs current law, we
we obtain
tance R. From Eqs. (10.25) and
=,(1 +A B). (10.27
Equation (10.24) gives
V Z (10.28)
Zy 1.(1+ AB) 1+ Ap
feedback, the input impedance af
where Eq. (10.23) is used. As I 1+ABI> 1 for negative
than that in the absence offeedback
is less
negative feedback amplifier with shunt mixing
(6) Output impedance: To calculate the output
zero and
impedance, we put the input signal voltage to
Then connect a
disconnect the load resistance R. we
Z 14A, (10.31)
gited.i.
in Fig.
10.13. The current I supplied by the source V is
10.
shown
as V V
I A, -A,BI or, I1+A,B) Z
output impedance with feedback is
So, the
(10.32)
2,A, feed-
feedback, the output impedance for current-shunt
Since 11+A, PI>1for negative
than that without feedback.
hack is greater
Distortion
Reduction of Frequency stage of an RC-coupled
9.3, we have found that the low-frequency gain of one
In Sec.
amplifier is
(10.33)
A1-jifIf
is the mid-frequency gain and f, is the lower half-power frequency. The high-fre-
where A
quency gain is
Am (10.34)
A, 1+fIA
f being the upper half-power frequeney.low and
With feedback, the overall gains at high frequencies are
A (10.35)
A 1+ A
Ap 1+AAB
(10.36)
and
reduced
The plot of the magmitude of the gain with frequency of an RC-coupled amplifier
is depicted in Fig. 10.14.
prgsence and in the absence of feedback
Decrease of Phase Distortion
When the load or the coupling impedance of an amplifier contains a reactive part, the c
the amplifier is a complex quantity with a magnitude and a phase angle. So, the gain A gain
A IAI 20.
With feedback, the gain is
(10.39
A
A 1+ AB (10.40)
The feedback ratio ß can be real or a complex quantity. Taking f to be real for simplieity
we obtain
IAIL AIL
1+BIAIZ9 1+ßIAl(cos0+ j sin )
IAIZ
1+ßIAl cos 0)+ jBIAlsin
IAIL
IBIL0 (10.41)
where TBl =
1+BIAlcos) +(BIA lsin ) (10.42)
and = tan-1
BIAlsin (10.43)
1+BIAIcos 0
Writing
A I A l 20, (10.44)
we obtain from Eq. (10.41)
AI
AB (10.45)
and , - (10.46)
Clearly, negative feedback decreases the phase angle of the gain by 6. Since the phase
angle by which the output signal leads the input signal diminishes, the phase distortion is
reduced.
(vLowering of Output Noise
fv, is the noise voltage at the input of an amplifier of voltage gain A in the absence of feedback,
the noise voltage at the amplifier output is
VonAV (10.47)
In the presence of feedback withfeedback fraction B, the gain reduces to A, =A/(1 +A B,
so that the output noise voltage becomes
AV,
Von 14 Aß Van (10.48)
1+ Aß
gative feedback |1+A
FoNote that thehe
reducea. N o t e Bl> 1. There
The efore, Vn<Van Thus
signal-to-noise
pack reduces
n g a t i v ef e e d b a r
the ratio at the
signal as well as the amplifier
231
the
geor the dc load current
changes. voltage Rectifier
ystem
contains (i) a rectifier with filtThe and
Filter Voltage
RLV
() voltage regulator
block
a
regulator, as shown
diagram of Fig. 1010wn in
the Fig. 10.15 Block
rectifier
Therectifier transtorms the diagram of a regulated
The output of the
oltage). The input alternating power supply.
0r alled the rectifier contains current (or voltage) into
voThe ails of the ripple. The filter following the
fluctuating a direct
current
rectifier and the filter the rectifier stage components of current (or
The output firom the filter can
action have smoothens out the
been discussed in ripples.
tance. The voltage vary with the Secs. 6.3
changes the ac supply through 6.6.
for circuit is explainedregulator tends to eliminate in
voltage the
help of the diagramthis fluctuation. The principle of a load
or
with the
woliage regulator.A fraction of
fxed reference voltage VR s V, of the output voltage V,Fig.of the
regu-
10.16, which is known as a series
eter P, BV, is made nearly appearing across the Zener
equal to VR The voltage diode Z. By
regulator is compared with a
amplifier comprising the transistor and difference (BV-adjusting the potentiom-
aDDlied to the base of the transistorT, the load resistance R TheV) is
by the amplified
exceeds the desired value, B Vo becomes T9» cálled the pass transistor. If theamplified voltage is
transistor T, to a more positive value. larger than V and drives the base output voltage V
thereby enhancing the voltage drop acrossConsequently, the collector eurrent ofvoltage
of the
emitter junction of T. Hence the R. This reduces the forward biasingT,ofincreases,
the base-
emitter current 1 of
current and hence the T, decreases.
output
change in V is countered by a voltage Vo decreases. Thus a negativeConsequently, the load
feedback occurs:
steady. The capacitor C bypasseschange
in the current I, any
the high-frequency causing V to remain more or less
further improving the components of the output voltage, thus
constancy
lated power supply, also called the
of the output voltage. The
percentage regulation of a regu
stabilised power supply, is much less than that of a simple
rectifier. This means that the internal resistance in
small, the system behaving like an ideal voltage source. R, the equivalent circuit of Fig 6.9 is
very
Rectifier
and filter
output Vo
voltage
BV
VA