CH 10 Feeback

Feedback
• 10.1 The general feedback structure

• 10.2 Some properties of negative feedback
• 10.3 The feedback voltage amplifier (Series-Shunt)
• 10.4 Systematic analysis of feedback voltage amplifiers
• 10.5 Other feedback-amplifier types
• 10.6 Summary of the feedback-analysis method
• 10.7 The stability problem
• 10.8 Effect of feedback on the amplifier poles
• 10.9 Stability study using bode plots
• 10.10 Frequency compensation
Shuenn-Yuh Lee EE/NCKU 1 Communication and Biologic IC Lab. Chapter 10

Feedback
• Feedback can be either negative (degenerative) or positive
(regenerative)
• In amplifier design, negative feedback is applied
– Desensitize the gain
– Reduce nonlinear distortion
– Reduce the effect of interference (noise)
– Control the input and output impedances
– Extend the bandwidth of the amplifier
• The basic idea of negative feedback is to trade off gain for other
desirable properties
• The use of positive feedback in the design of oscillators and
bistable circuits (Chapter 13)

Feedback (Cont.)
• We shall study the stability problems of negative-feedback
amplifiers and their potential for oscillation
• Almost all op amp circuits employ negative feedback.
• Applications
– Use of the emitter resistance RE to stabilize the bias point of bipolar
transistors and to increase the input resistance and bandwidth of a BJT
differential amplifier
– The emitter follower and the source follower employ a large amount of
negative feedback 1
RL' = RL || ro ||
g mb
Rsig + RL' 1 RL'

Rgd = Rsig , Rgs = , RC L = R ||
'
=
1 + g m RL' g m 1 + g m RL'
L
1 1
fH = =
Gm =
gm
≈
gm
1 + ( g m + g mb )Rs 1 + g m Rs
2πτ H (
2π C gd Rgd + C gs Rgs + C L RC L )
10.1 General Feedback Structure
• The open-loop amplifier gain = A; thus

• The sample xf is related to xo by the feedback factor β
x f = βxo 
 ⇒ xo = A( xs − βxo )
xi = xs − x f 
– Implicit in the above description is that the source, the load, and the
feedback network do not load the basic amplifier
• The gain of the feedback amplifier A ≡ xo = A
xs 1 + Aβ
f

General Feedback Structure (Cont.)
• The quantity Aβ is called the loop gain
– For the feedback to be negative, the loop gain Aβ should be positive
– That is, the feedback signal xf should have the same sign as xs, thus
resulting in a smaller difference signal xi
– 1+ Aβ: called he amount of feedback
– The gain with feedback will be smaller than the open-loop gain A by the
quantity 1+ Aβ
• For Aβ>>1, Af≈1/β:the gain of the feedback amplifier (closed-
loop gain) is almost entirely determined by the feedback
network
– The feedback network usually consists of passive components, which
can be chosen to be as accurate as one wishes, the advantages
Accurate, predictable, and stable gain should be apparent
xo A 1
Af ≡ = ≈
xs 1 + Aβ β
General Feedback Structure (Cont.)
• The feedback signal xf
βA
xf = xs
1 + Aβ
• For Aβ>>1, then xf ≈ xs
– The signal xi at the input of the basic amplifier is reduced to almost zero
• The difference between xs and xf, which is xi, is sometimes
referred to as the “error signal”
• The input differencing circuit is often also called a comparison
circuit (mixer)

Example 10.1
• The noninverting op amp configuration shown in Fig. 10.3(a) provides a
direct implementation of the feedback loop of Fig. 10.1
(a) Assume that the op amp has infinite input resistance and zero output
resistance. Find an expression for the feedback factor β and the ideal
closed-loop gain Af
• Solution β = v f = R1 = 1
vo R1 + R2 1 + R2 / R1
A
Af =
1 + βA
 R1 
Aβ >> 1, A  >> 1
 R1 + R2 
1 R + R2 R
Af ≅ = 1 = 1+ 2
β R1 R1
The condition can be stated as A>>Af

Example 10.1 (Cont.)
(b) Find β and R2/R1 to obtain an ideal closed-loop gain of 10 V/V
• Solution
1
𝛽𝛽 = = 0.01
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
10 = 1 + 𝑅𝑅2 /𝑅𝑅1 => 𝑅𝑅2 /𝑅𝑅1 =9
• (c) If the open-loop voltage gain A=104, find for the design in (b) the loop
gain, the amount of feedback, and the actual value of Af. By what
percentage does Af deviate from the ideal value?
• Solution
𝐴𝐴𝐴𝐴 = 104 × 0.1 = 1000
1+Aβ=1001=> ≅60 dB
𝐴𝐴 104
𝐴𝐴𝑓𝑓 = = = 9.990
1 + 𝐴𝐴𝐴𝐴 1001
Which is 0.1% below the ideal value of 10 V/V

(d) To what values must β and R2/R1 be changed to obtain a closed-loop
gain of exactly 10 V/V
• Solution
𝐴𝐴 104
𝐴𝐴𝑓𝑓 = ⇒ 10 = ⇒ 𝛽𝛽 = 0.0999𝑉𝑉/𝑉𝑉
1 + 𝐴𝐴𝐴𝐴 1 + 𝛽𝛽 ⋅ 104
1
= 0.0999 ⇒ 1 + 𝑅𝑅2 /𝑅𝑅1 = 10.01 ⇒ 𝑅𝑅2 /𝑅𝑅1 = 9.01
1 + 𝑅𝑅2 /𝑅𝑅1
• (e) If Vs=1V, find Vo, Vf, and Vi.
• Solution
𝑉𝑉𝑜𝑜
= 𝐴𝐴𝑓𝑓 = 10 ⇒ 𝑉𝑉𝑜𝑜 = 10𝑉𝑉
𝑉𝑉𝑠𝑠
1
𝑉𝑉𝑓𝑓 = 𝛽𝛽𝑉𝑉𝑜𝑜 = ⋅ 10 = 0.999𝑉𝑉
10.01
𝑉𝑉𝑖𝑖 = 𝑉𝑉𝑠𝑠 − 𝑉𝑉𝑓𝑓 = 0.001𝑉𝑉
• (f) If A decreases by 20%, what is the corresponding decrease in Af ?
A 8000
Af = = = 9.9975
1 + Aβ 1 + 1 ⋅ 8000
10.01
10 − 9.9975
error % of A f = ×100% = 0.025%
10
Exercise
• 10.1 Repeat Example 10.1, (b) to (f) for A=100 V/V.
• Solution
1
𝛽𝛽 = = 0.1
10 = 1 + 𝑅𝑅2 /𝑅𝑅1 => 𝑅𝑅2 /𝑅𝑅1 =9
• (c) If the open-loop voltage gain A=100, find for the design in (b) the loop gain, the
amount of feedback, and the actual value of Af. By what percentage does Af deviate
from the ideal value?
𝐴𝐴𝐴𝐴 = 100 × 0.1 = 10
1+Aβ=11=> ≅20.8 dB
𝐴𝐴 100
𝐴𝐴𝑓𝑓 = = = 9.091
1 + 𝐴𝐴𝐴𝐴 11
Which is -9.1% below the ideal value of 10 V/V

Exercise
(d) To what values must β and R2/R1 be changed to obtain a closed-loop gain of
exactly 10 V/V
• Solution
𝐴𝐴 100
𝐴𝐴𝑓𝑓 = ⇒ 10 = ⇒ 𝛽𝛽 = 0.09𝑉𝑉/𝑉𝑉
1 + 𝐴𝐴𝐴𝐴 1 + 𝛽𝛽 ⋅ 100
1
= 0.09 ⇒ 1 + 𝑅𝑅2 /𝑅𝑅1 = 11.11 ⇒ 𝑅𝑅2 /𝑅𝑅1 = 10.11
1 + 𝑅𝑅2 /𝑅𝑅1
• Solution
𝑉𝑉𝑜𝑜
𝑉𝑉𝑠𝑠
𝑉𝑉𝑓𝑓 = 𝛽𝛽𝑉𝑉𝑜𝑜 = 0.09 ⋅ 10 = 0.9𝑉𝑉
𝐴𝐴 80
𝐴𝐴𝑓𝑓 = = = 9.756
1 + 𝐴𝐴𝐴𝐴 1 + 0.09 ⋅ 80
10−9.756
Error % of 𝐴𝐴𝑓𝑓 = × 100% = 24%
10

Exercise (Cont.)
• 10.2 Repeat Example 10.1, (b) to (f) for Af=103 V/V. For (d) design for Af=1000 V/V,
and for (e) use Vs=0.01 V
• Solution
1
𝛽𝛽 = = 0.001
1000 = 1 + 𝑅𝑅2 /𝑅𝑅1 => 𝑅𝑅2 /𝑅𝑅1 =999
• (c) If the open-loop voltage gain A=104, find for the design in (b) the loop gain, the
amount of feedback, and the actual value of Af. By what percentage does Af deviate
from the ideal value?
• Solution
𝐴𝐴𝐴𝐴 = 1000 × 0.001 = 10
1+Aβ=11=> ≅20.8 dB
𝐴𝐴 10000
𝐴𝐴𝑓𝑓 = = = 909.1
1 + 𝐴𝐴𝐴𝐴 11
Which is -9.1% below the ideal value of 10 V/V
Exercise (Cont.)
(d) To what values must β and R2/R1 be changed to obtain a closed-loop gain of
exactly 10 V/V
• Solution
𝐴𝐴 10000
𝐴𝐴𝑓𝑓 = ⇒ 1000 = ⇒ 𝛽𝛽 = 0.0009𝑉𝑉/𝑉𝑉
1 + 𝐴𝐴𝐴𝐴 1 + 𝛽𝛽 ⋅ 10000
1
= 0.0009 ⇒ 1 + 𝑅𝑅2 /𝑅𝑅1 = 1111.11 ⇒ 𝑅𝑅2 /𝑅𝑅1 = 1110.1
1 + 𝑅𝑅2 /𝑅𝑅1
• Solution
𝑉𝑉𝑜𝑜
𝑉𝑉𝑠𝑠
𝑉𝑉𝑓𝑓 = 𝛽𝛽𝑉𝑉𝑜𝑜 = 0.0009 ⋅ 10 = 0.009𝑉𝑉
𝐴𝐴 8000
𝐴𝐴𝑓𝑓 = = = 975.6
1 + 𝐴𝐴𝐴𝐴 1 + 0.0009 ⋅ 8000
1000−975.6
Error % of 𝐴𝐴𝑓𝑓 = × 100% = 2.44%
10

Summary
• Open loop gain ≡ A
• Feedback factor ≡ β
• Loop gain ≡ Aβ
• Amount of feedback ≡ 1+Aβ
• Closed-loop gain ≡ A ≡ xo = A
xs 1 + Aβ
f
βA
• Feeback signal ≡ x f = xs
1 + Aβ
1
xi = xs
• Input signal to basic amplifier ≡ 1 + Aβ
• Closed-loop gain as a function of the ideal value
1 1 1
: A f =  
β  β  1 + 1 Aβ 1
• For large loop gain, Aβ>>1, A f ≅ β , x f ≅ xs , xi ≅ 0
10.2 Some Properties of Negative Feedback
• Gain Desensitivity
A dA dA f 1 dA
Af = ⇒ dA f = ⇒ =
1 + Aβ (1 + Aβ )2
Af 1 + Aβ A
– Demonstrated in Example 9.1, a 20% reduction in the gain of the basic
amplifier gave rise to only a 0.02% reduction in the gain of the closed-
loop amplifier
– 1+βA : desensitivity factor
• Bandwidth Extension
– For single pole assumption, where AM denotes the midband gain and ωH
is the upper 3-dB frequency
AM
A( s ) =
1 + s / ωH
A( s ) AM /(1 + AM β )
Af (s) = =
1 + βA( s ) 1 + s / ω H (1 + AM β )
– The feedback amplifier will have a midband gain of AM/(1+ AMβ) and an
upper 3-dB frequency ωHF= ωH (1+ AMβ)
Exercise
• 10.3 An amplifier with a nominal gain A=1000 V/V exhibits a gain change of 10% as
the operating temperature changes from 25oC to 75oC. If it is required to constrain
the change to 0.1% by applying negative feedback, what is the largest closed-loop
gain possible? If three of these feedback amplifiers are placed in cascade, what
overall gain and gain variability are achieved?
dA f 1 dA
=
Af 1 + βA A
1
0.1% = ×10%
1 + βA
1
= 0.01
1 + βA
A
Af = = 1000 × 0.01 = 10V / V
1 + βA
Atotal = A f 1 × A f 2 × A f 3 = 1000V / V
dA f 1 dA f 2 dA f 3
+ + = 0.3%
Af 1 Af 2 Af 3

Some Properties of Negative Feedback (Cont.)
– The upper 3-dB frequency is increased by a factor equal to the amount
of feedback (1+βA)
– Similarly, if the open-loop gain is characterized by a dominant low-
frequency pole giving rise to a lower 3-dB frequency, ωL, then the
feedback amplifier will have a lower 3-dB frequency ωLf
AM s / ω L A s
A( s ) = = M
1 + s / ωL s + ωL
A( s ) A /(1 + AM β ) s ωL
A f (s) = = M , ω Lf =
1 + βA( s ) s + ω L /(1 + AM β ) 1 + AM β

– The amplifier bandwidth is increased by the same factor by which its
midband gain is decreased, maintaining the gain-bandwidth product
constant
Bandwidth modification as a result of feedback
Vo
Vsig

Exercise
• 10.4 Consider the noninverting op-amp circuit of Exercise 10.1. Let the open-loop
gain A have a low-frequency value of 104 and a uniform -6 dB/octave rolloff at high
frequencies with a 3-dB frequency of 100Hz. Find the low-frequency gain and the
upper 3-dB frequency of a closed-loop amplifier with R1=1kΩ and R2=9kΩ
vf R1 1 1 1
β= = = = =
vo R1 + R2 1 + R2 / R1 1 + 9 10
1
1 + βA = 1 + × 10 4 = 1001
10
f −3dB ( f ) = (1 + βA) f −3dB = 1001×100 = 100.1kHz
A 10 4
Af = = = 9.99V / V
1 + βA 1001
• Interference (Noise) reduction
– Negative feedback reduce the noise or interference
Increase the signal-to-noise ratio
– For Fig. (a) S/N=Vs/Vn
– For Fig. (b) : precede our original amplifier A1 by the clean amplifier A2
and apply negative around the overall cascade of such an amount as to
keep the overall gain constant
A1 A2 A1 S V
Vo = Vs + Vn ⇒ = s A2
1 + A1 A2 β 1 + A1 A2 β N Vn

– Example : reduce the power-supply hum
The problem arises because of the large currents that this stage draws
from the power supply and the difficulty in providing adequate power-
supply filtering inexpensively
– The hum at the output will then be reduced by the amount of the voltage
gain of this added preamplifier (A2)

Exercise
• Supplement: Consider a power-output stage with voltage gain A1=1, an input signal
Vs=1V, and a hum vn of 1V. Assume that this power stage is preceded by a small-
signal stage with gain A2=100 V/V and that overall feedback with β=1 is applied. If Vs
and Vn remain unchanged, find the signal and noise voltages at the output and hence
the improvement in S/N.
Vo A1 A2 1×100 100
= = = = 0.99 Vo = Vs
A1 A2
+ Vn
A1 S V
⇒ = s A2
Vs 1 + βA1 A2 1 + 1×1×100 101 1 + A1 A2 β 1 + A1 A2 β N Vn
Vo = 0.99V ≈ 1V
Vo A1 1 1
= = = = 0.0099
Vn 1 + βA1 A2 1 + 1×1×100 101
Vo = 0.0099 V ≈ 0.01V
1
S/N = = 100 (40dB)
0.01

• (a) The amplifier transfer characteristic without feedback
– The characteristic is piecewise linear, with the voltage gain changing
from 1000 to 100 and then to 0
– Generate a large amount of nonlinear distortion
• (b) The characteristic with negative feedback (β=0.01) applied
– The slope of the steepest segment is given by
1000 100
Af 1 = = 90.9 Af 2 = = 50
1 + 1000 × 0.01 1 + 100 × 0.01
– The order-of-magnitude change
in slope ⇒ reduced
– If the overall gain has to be
restored ⇒ Preamplifier is
required (linear)
– Negative feedback does nothing
about amplifier saturation
Example 10.2
• Give the availability of amplifier with the voltage transfer characteristic
shown in Fig. 10.7(a), design an amplifier with a gain of 200 V/V and a linear
transfer characteristic for input signals as large as 5 mV.
• Solution
𝐴𝐴
𝐴𝐴𝑓𝑓 =
1 + 𝐴𝐴𝐴𝐴
By substituting Af=2 and A=100
100
2= ⇒ 𝛽𝛽 = 0.49𝑉𝑉/𝑉𝑉
1 + 𝛽𝛽 ⋅ 100
10.3 The Feedback Voltage Amplifier
• Four configurations (types of amplifiers along with their
idealized models)
RL RS
VO = VS IO = IS
RS + RL RS + RL
RL → ∞ or RS → 0 RS → ∞ or RL → 0
VO = VS IO = I S
– Circuits sensing a voltage must exhibit a high input impedance whereas

those sensing a current must provide a low input impedance
– Circuits generating a voltage must exhibit a low output impedance while
those generating a current must provide a high output impedance
Simple Implementations of Types of Amplifiers
• (a) : a common-source stage senses and produce voltages

• (b) : a common-gate circuit serves as transimpedance amplifier,
converting the source current to a voltage at the drain
• (c) : a common-source transistor generates an output current
in response to an input voltage
• (d) : a common-gate device senses and produces currents
Four types of Amplifiers
X X
VX Vout ro V V ro
A= ⋅ = − g m1 ( ro1 // RD ) ⋅ ≈ − g m1 ( ro1 // RD ) Rm = X ⋅ out =RD ⋅ ≈ RD
Vin VX ro + 1/ g m 2 I in VX ro + 1/ g m 2
V I
AI = X ⋅ out =RD ⋅ g m 2 =g m 2 RD
= − g m1 (ro1 // RD ) ⋅ g m 2
VX I out
Gm = ⋅ I in VX
Vin VX
Sense and Return Mechanisms
• Sampling a voltage : shunt feedback
– Place a voltmeter in parallel with corresponding port
• Sampling a current : series feedback
– A current meter is inserted in series with the signal
• Addition of (a) voltages

and (b) currents
– No influence on the operation of the
open-loop amplifier
– Introduces loading effects
Feedback Topologies
• Voltage-voltage feedback (Voltage Amplifier)
– Samples the output voltage and returns the feedback signal as a voltage
• Voltage-current feedback (Transconductance Amplifier)
– Sense the output current to perform feedback
• Current-voltage feedback (Transresistance Amplifier)
– Sense the output voltage and a proportional current is returned to the
summing point at the input
• Current-current feedback (Current Amplifier)
– Sense the output current and a proportional current is returned to the
summing point at the input

Voltage Amplifier-Series-Shunt Feedback
Topology
• Amplifier an input voltage signal and provide an output voltage signal
(voltage control voltage source)
• High input impedance, low output impedance
• Represent it in terms of a Th’evenin equivalent circuit
• The feedback network should sample the output voltage, and the feedback
signal xf should be a voltage that can mixed with the source voltage in series
• Voltage-sampling series-mixing (series-shunt feedback) topology
Rif ↑ than Ri
Rof ↓ than Ro
• A transistor amplifier with series-shunt feedback (Fig. 10.8)

Voltage Amplifier (Cont.)
• An example of the voltage-sampling series-mixing feedback
topology

Analysis of the Feedback Voltage Amplifier
Utilizing the Loop Gain
• The loop-gain analysis method comprises four step
– Identify the feedback network and use it to determine the value of β
– Determine the ideal value of the closed-loop gain Af as 1/β
Find β

Analysis of the Feedback Voltage Amplifier
Utilizing the Loop Gain (Cont.)
• The loop-gain analysis method comprises four step
– Determine the loop gain by breaking the feedback at the output of the
basic amplifier
Vr
Aβ = −
Vt
– Use the value of Aβ together with that of β to determine the open-loop
A
gain A. Then determine the closed-loop gain Af A =
1 + Aβ
f
– It does not enable us to find the input and output resistances of the
feedback => systematic approach is required

Example 10.3
• For the series-shunt feedback amplifier of Fig. 10.9(b), neglect the
MOSFETs’ ro
• (a) give the feedback network (β circuit) and an expression for β. Also give
an expression for the ideal or upper-bound value of the closed-loop gain Af
• (b) find the ratio R2/R1 that results in an ideal closed-loop gain of 10 V/V. If
R1=1kΩ, what value must R2 have?
• (c) Find an expression for the loop gain Aβ.
• (d) if gm1=gm2=4 mA/V and RD1=RD2=10 kΩ, determine the values of Aβ, A,
and Af
• Solution: (a)
Vf R1
β= =
Vo R1 + R2
Vo A 1 R
Aβ >> 1 ⇒ A f = = ≅ = 1+ 2
Vs 1 + βA β R1
MOSFETs’ ro
• (b) find the ratio R2/R1 that results in an ideal closed-loop gain of 10 V/V. If
R1=1kΩ, what value must R2 have?
• (c) Find an expression for the loop gain Aβ.
• (b) A = 1 + R2 ⇒ 10 = 1 + R2 ⇒ R = 9kΩ
f 2
R1 1
• (c) To determine the loop gain, we set Vs=0 =>Aβ≡-Vr/Vt
I d 2 = g m 2Vt
RD 2
I1 = − I d 2
 1 
RD 2 + R2 +  R1 || 
 g m1 
R1
I d 1 = I1 ⇒ Vr = I d 1 RD1
1
R1 +
g m1
Vr V I I V g R RD 2
Aβ = − = − r d 1 1 r = − RD1 m1 1 gm2
Vt I d 1 I1 I d 2 I1 1 + g m1 R1  1 
RD 2 + R2 +  R1 || 
 g m1 
= ( g m1 RD1 )( g m 2 RD 2 )
1 R1
1 + g m1 R1  1 
RD 2 + R2 +  R1 || 
 g m1 
MOSFETs’ ro
• (d) if gm1=gm2=4 mA/V and RD1=RD2=10 kΩ, determine the values of Aβ, A,
and Af
• (d) Aβ = − Vr = (g R )(g R ) 1 R1
1 + g m1 R1  1 
m1 D1 m2 D2
Vt
RD 2 + R2 +  R1 || 
 g m1 
1 1
= 4 ×10 × 4 ×10 × × = 16.67
+
1 4 1×  1 
10 + 9 + 1 || 
 4
R1 1
β= = = 0.1
R1 + R2 1 + 9
Aβ 16.67
A= = = 166.7
β 0.1
A 166.7
Af = = = 9.43V / V
1 + Aβ 1 + 16.67
• It’s reasonably close to the ideal value of 10 V/V
Example 10.4
• In the series-shunt feedback amplifier of Fig. 10.13(a), the op amp has an
input resistance Rid, an open-circuit voltage gain µ, and an output resistance
ro. Find expressions for β, the ideal value of Af≡Vo/Vs, and the loop gain Aβ.
For µ=104 V/V, Rid=100 kΩ, ro=1 kΩ, RL=2 kΩ, R1=1 kΩ, R2 =100 kΩ, and
Rs=10 kΩ, find β, Af|ideal, Aβ, A, and Af
• Solution:
𝑅𝑅1 1
𝛽𝛽 = =
𝑅𝑅1 +𝑅𝑅2 1+100
= 0.0099
1 𝑅𝑅2
𝐴𝐴𝑓𝑓 = = 1 + = 101 V/V
𝛽𝛽 𝑅𝑅1
Example 10.4
• In the series-shunt feedback amplifier of Fig. 10.13(a), the op amp has an
ro. Find expressions for β, the ideal value of Af≡Vo/Vs, and the loop gain Aβ
• Solution: V1
V2
𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 𝑅𝑅𝑖𝑖𝑖𝑖

𝑉𝑉𝑟𝑟 = −𝜇𝜇𝑉𝑉𝑡𝑡
𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 + 𝑟𝑟𝑜𝑜 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 + 𝑅𝑅2 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠
𝑉𝑉𝑟𝑟 𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 𝑅𝑅𝑖𝑖𝑖𝑖
𝐴𝐴𝐴𝐴 = − = 𝜇𝜇
𝑉𝑉𝑡𝑡 𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 + 𝑟𝑟𝑜𝑜 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 + 𝑅𝑅2 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠
𝐴𝐴𝐴𝐴 =60, A=Aβ/β=5960 V/V, and Af=A/(1+ Aβ)=99.3

Exercise
• 10.5 For the feedback voltage amplifier of Fig. 10.9(c):
• (a) Find an expression for β.
• (b) Neglecting the MOSFET ro, find an expression for the loop gain Aβ. (Hint: Break
the loop at the gate of Q.)
• (c) Find expressions for the open-loop gain circuit A.
• (d) For gm=4 mA/V, RD=10 kΩ, R1=20 kΩ, and R2= 80 kΩ, find the values of β, Aβ, A,
and the closed-loop gain Af. What would Af be if Aβ were much greater than unity?
• Solution (a) (b)
V V V
Aβ ≡ − r = − r d
R1 Vt Vd Vt
β=
R1 + R2
=−
R1
{− g m [RD || (R1 + R2 )]}
R1 + R2
= g m [RD || (R1 + R2 )]
R1
R1 + R2
RD R1
= gm
RD + R1 + R2

Exercise (Cont.)
• 10.5 For the feedback voltage amplifier of Fig. 10.9(c):
• (c) Find expressions for the open-loop gain circuit A.
• (d) For gm=4 mA/V, RD=10 kΩ, R1=20 kΩ, and R2= 80 kΩ, find the values of β, Aβ, A,
and the closed-loop gain Af. What would Af be if Aβ were much greater than unity?
• Solution
• (c) Aβ R (R + R2 )
A= = g m [RD || (R1 + R2 )] = g m D 1
β RD + R1 + R2
(d) β = R1 20
• = = 0.2V / V
R1 + R2 20 + 80
RD R1 10 × 20
Aβ = g m =4 = 7.27
RD + R1 + R2 10 + 20 + 80
Aβ 7.27
A= = = 36.36V / V
β 0.2
A 36.36
Af = = = 4.4V / V
1 + Aβ 1 + 7.27
1 1
If Aβ were >> 1, then ⇒ A f ≅ = = 5V / V
β 0.2
10.4 Systematic Analysis of Feedback Voltage
Amplifiers
(b)
(a)
– It consists of a unilateral open-loop amplifier (the A circuit including Ri,
A,and Ro) and an ideal voltage-sampling series-mixing feedback network
(the β circuit, not load the A circuit)
– Connecting the β circuit does not change the value of A=Vo/Vi
Ideal Case
• Closed-loop voltage gain
Vo A
Af ≡ =
Vs 1 + Aβ
– A and β have reciprocal units in a dimensionless loop gain Aβ
• Input resistance
Vs Vs V V + βVo V + βAVi
Rif ≡ = = Ri s = Ri i = Ri i
I i Vi / Ri Vi Vi Vi
= Ri (1 + Aβ )
– Increase the input resistance by a factor equal to the amount of feedback
– Only depend on the method of mixing
– By intuitive : Vi=Vs-Vf=Vs/(1+βA) is quite small, thus the input current Ii
becomes small and the resistance seen by Vs become large
– General form
Z if ( s ) ≡ Z i ( s )[1 + A( s ) β ( s )]

Ideal Case (Cont.)
• Output resistance
– Reduce the Vs to zero and apply a test voltage Vx at the output
Vx − AVi
Ix =
Ro
for Vs = 0, Vi = −V f = − βVo = − βVx
Vx + AβVx V Ro
Ix = ⇒ Rof ≡ x =
Ro I x 1 + Aβ
– Reduce the output resistance by a factor equal to the amount of

feedback
– Depend only on the method of sampling
– By intuitive : the feedback samples the output voltage Vo, acting to
stabilize the value of Vo ⇒ reduce changes in the value of Vo
– General form Z o (s)
Z of ( s ) =
1 + A( s ) β ( s )
Effect of Loading
• Break the feedback loop so as to identify the open-loop system
– Calculate the open-loop gain and the input and output impedance
• Two-port network models
Series-Series Feedback Shunt-Shunt feedback
Shunt-Series Feedback
Series-Shunt Feedback
– (a) : Z model ⇒ voltage-current – (c) : H model ⇒ voltage-voltage

V1=H11I1+H12V2
V1=Z11I1+Z12I2
I2=H21I1+H22V2
V2=Z21I1+Z22I2
– (d) : G model ⇒ current-current
– (b) : Y model ⇒ current-voltage
I1=G11V1+G12I2
I1=Y11V1+Y12V2
V2=G21V1+G22I2
I2=Y21V1+Y22V2
Practical Case
• In a practical situation,the feedback network will not be an ideal
voltage-controlled voltage source
– Affect the value of A, Ri and Ro
• Block diagram of a practical series-shunt feedback amplifier

Practical Case (Cont.)
• The two-port feedback network in terms of its h parameters
– The choice of h parameters is based on the fact that this is the only
parameter set that represents the feedback network by a series network
at port 1 and a parallel network at port 2
– Since the feedback network is usually passive, its forward transmission

can be neglected, thus assume |h21|feedback network << |h21|basic amplifier
– If the basic amplifier is unilateral, a situation that prevails when
|h12|basic amplifier<< |h12 |feedback network,

• Replace h11 and h22 as R11 and R22, respectively.
1. R11 and R22 represent the loading effect of the feedback network on the
basic amplifier at the input and output, respectively
2. R11:the impedance looking into port 1 of the feedback network with port
2 short-circuits ⇒ destroy the feedback (due to parallel output)
R22:the admittance looking into port 2 of the feedback network with port
1 open-circuited ⇒ destroy the feedback (due to series input)

The loading effect is found by looking into the appropriate port of the
feedback network while the other port is open-circuited or short-circuited
so as to destroy the feedback
If the connection is a shunt one, we short-circuit the port; if it is a series
one, we open-circuit it
3. Determination of β
One applies a voltage to port 2 of the feedback network and measures
the voltage that appears at port 1 while the latter port is open-circuited
The series connection at the input suggests that β should be found with
port 1 open-circuited
Vf
β≡
Vo I1 = 0

Another presentation by H model
V V1=H11I1+H12V2
β = h12 ≡ 1
V2 I1 = 0
I2=H21I1+H22V2
4. Find the open-loop gain A according to open-loop circuit in Fig. (e)

Vo
A=
Vi
Vo A
5. Determine the closed-loop gain Af Af ≡ =
Vs 1 + Aβ
6. Determine input resistance Ri and output resistance Ro from open-loop
circuit, then find the input and output resistances with feedback
𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖 1 + 𝐴𝐴𝐴𝐴 , 𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜 ⁄(1 + 𝐴𝐴𝐴𝐴)
The actual input and output resistances with feedback

 1 1 
Rin = Rif − Rs , R out = 1  −
R R 
Shuenn-Yuh Lee EE/NCKU  of 49
L  Communication and Biologic IC Lab. Chapter 10
• Conceptual view of opening a voltage-voltage (series-shunt)
feedback loop with proper loading
I2
h22 =
V2 I1 = 0
V1
h11 =
I1 V
2 =0
h22 -1
h11
– (c) : H model ⇒ voltage-voltage

V1=H11I1+H12V2
I2=H21I1+H22V2

Summary
• A summary of the rules for finding the A circuit and β for the
feedback voltage amplifiers

Summary
• Note that
– Ri and Ro are the input and output resistances of A circuit. (Fig. 10.14(e))
– Rif and Rof are the input and output resistance of the feedback amplifier,
including RS and RL (Fig. 10.14(a))
– The actual input and output resistances of the feedback amplifier usually
exclude RS and RL. These are denoted Rin and Rout in Fig. 10.14(a) and
can be easily determined as
Rin = Rif − RS
Rout = 1
 1 1 
 −
R 
 of RL 
Rif = Rin + RS
1 1 1
Rof = RL // Rout ⇒ = +
Rof RL Rout
Example 10.5
• Fig. 10.17(a) shows an op amp connected in the noninverting configuration.
The op amp has an open-loop gain μ, a differential input resistance Rid, and
an output resistance ro. Recall that in our analysis of op-amp circuits in
Chapter 2, we neglected the effects of Rid (assume it to be infinite) and of ro
(assume it to be zero). Here we wish to use the feedback method to analyze
the circuit taking both Rid and ro into account. Find expressions for A, β, the
closed-loop gain Vo/Vs, the input resistance Rin (see Fig. 9.16(a)), and the
output resistance Rout. Also find numerical values, given μ=104, Rid=100kΩ,
ro=1kΩ, RL=2kΩ, R1=1kΩ, R2=100kΩ, and Rs=10kΩ

𝑉𝑉𝑜𝑜 𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 𝑅𝑅𝑖𝑖𝑖𝑖

𝐴𝐴 ≡ = 𝜇𝜇
𝑉𝑉𝑖𝑖 𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 + 𝑟𝑟𝑜𝑜 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 + 𝑅𝑅2 + 𝑅𝑅1
= 5967 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑡𝑡𝑡𝑡 𝑡𝑡𝑡𝑡𝑡 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣 𝑜𝑜𝑜𝑜 5960 𝑖𝑖𝑖𝑖 𝐸𝐸𝐸𝐸. 4
𝑉𝑉𝑓𝑓 𝑅𝑅1
𝛽𝛽 ≡ = = 0.0099𝑉𝑉/𝑉𝑉
𝑉𝑉𝑜𝑜 𝑅𝑅1 + 𝑅𝑅2
𝑉𝑉𝑜𝑜 1
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ≡ = = 101 𝑉𝑉/𝑉𝑉
𝑉𝑉𝑠𝑠 𝛽𝛽
𝐴𝐴𝐴𝐴=59.1, 1+𝐴𝐴𝐴𝐴=60.1
𝑉𝑉 𝐴𝐴 5967
𝐴𝐴𝑓𝑓 ≡ 𝑉𝑉𝑜𝑜 = 1+𝐴𝐴𝐴𝐴 = 60.1 = 99.3
𝑠𝑠
𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑠𝑠 + 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅1 //𝑅𝑅2 ≈ 111𝑘𝑘𝑘𝑘

𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖 1 + 𝐴𝐴𝐴𝐴 = 111 × 60.1 = 6.67𝑀𝑀𝛺𝛺
𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 − 𝑅𝑅𝑆𝑆
𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 − 𝑅𝑅𝑠𝑠 = 6.66𝑀𝑀𝛺𝛺
𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = 1
𝑅𝑅𝑜𝑜 = 𝑟𝑟𝑜𝑜 //𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 = 662𝛺𝛺 � 1 − 1
𝑅𝑅𝑜𝑜 662 𝑅𝑅𝑜𝑜𝑜𝑜 𝑅𝑅𝐿𝐿
𝑅𝑅𝑜𝑜𝑜𝑜 = = = 11𝛺𝛺
1 + 𝐴𝐴𝐴𝐴 60.1
𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 //𝑅𝑅 𝐿𝐿⇒ 𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 ≈ 11𝛺𝛺

Example 10.6
• As another example of a series-shunt feedback amplifier, consider the circuit
shown in Fig. 10.9(b) which we analyzed in Example 10.3 by determining
the loop gain Aβ. In this example we wish to first analyze the circuit using our
systematic procedure and then compare the results to those obtained in
Example 10.3. For convenience, the circuit is repeated in Fig. 10.18(a). It is
required to obtain its voltage gain Vo/Vs, input resistance Rin, and output
resistance Rout. Find numerical values for the case gm1=gm2=4 mA/V,
RD1=RD2=10 kΩ, R1=1 kΩ, and R2=9 kΩ. For simplicity, neglect ro of each of
Q1 and Q2.

• Solution:
V RD1 g m1 RD1 Vo g R g [R || (R1 + R2 )]

A1 = d 1 = − =− A= = A1 A2 = m1 D1 m 2 D 2
Vi 1 / g m1 + (R1 || R2 ) 1 + g m1 (R1 || R2 ) Vi 1 + g m1 (R1 || R2 )
4 × 10 × 4[10 || (1 + 9 )]
= − g m 2 [RD 2 || (R1 + R2 )]
Vo
A2 = = = 173.9 V / V
Vd 1 1 + 4(1 || 9 )

Vf R1 1
β= = = = 0.1
Vo R1 + R2 1 + 9
Vo A 173.9
Af = = = = 9.5 V / V
Vs 1 + βA 1 + 0.1×173.9
Rin = ∞
Ro = RD 2 || (R2 + R1 ) = 10 || 10 = 5 kΩ
Ro 5000
Rout = = = 272Ω << Ro
(1 + Aβ ) 1 + 0.1×173.9

Exercises
• 10.6 If the op amp of Example 10.5 has a uniform –20-dB/decade high-frequency
rolloff with f3dB=1kHz,, find the 3-dB frequency of the closed-loop gain Vo/Vs
𝑓𝑓𝐻𝐻𝐻𝐻 = 1 + 𝐴𝐴𝐴𝐴 𝑓𝑓𝐻𝐻 ≈ 1 + 5967 × 0.0099 × 1𝐾𝐾𝐾𝐾𝐾𝐾 = 60.1𝐾𝐾𝐾𝐾𝐾𝐾
• 10.7 The circuit shown in Fig. E10.17 consists of a differential stage followed by an
emitter follower, with series-shunt supplied by the resistors R1 and R2. Assume that
the dc component of Vs is zero, and that β of the BJTs is very high, find the dc
operating current of each of the three transistors and show that the dc voltage at the
output is approximately zero. Then find the values of A, β, Af≡Vo/Vs, Rin, and Rout.
Assume that the transistors have β=100
VC2
Rid3

Exercises: 10.7
• DC analysis
𝐼𝐼𝐸𝐸𝐸 = 𝐼𝐼𝐸𝐸𝐸 = 0.5𝑚𝑚𝑚𝑚
𝑉𝑉𝐶𝐶𝐶 ≈ 10.7 − 0.5 × 20 = 0.7𝑉𝑉
𝑉𝑉𝑜𝑜 = 0.7 − 𝑉𝑉𝐵𝐵𝐵𝐵𝐵 = 0𝑉𝑉
𝐼𝐼𝐸𝐸𝐸 = 5𝑚𝑚𝑚𝑚
𝑉𝑉𝑇𝑇 25
𝑟𝑟𝑒𝑒𝑒 = 𝑟𝑟𝑒𝑒𝑒 = = = 50Ω
𝐼𝐼𝐸𝐸𝐸 0.5
𝑉𝑉𝑇𝑇 25
𝑟𝑟𝑒𝑒𝑒 = = = 5Ω
𝐼𝐼𝐸𝐸𝐸 5
• β Circuit
𝑅𝑅1 1
𝛽𝛽 = = = 0.1𝑉𝑉/𝑉𝑉
𝑅𝑅1 + 𝑅𝑅2 1 + 9
𝑅𝑅11 = 𝑅𝑅1 ||𝑅𝑅2 = 1||9 = 0.9𝑘𝑘𝑘 VC2
Rid3
𝑅𝑅22 = 𝑅𝑅1 + 𝑅𝑅2 = 1 + 9 = 10

Exercises: 10.7 (Cont.)
• A Circuit • Closed loop

𝑉𝑉𝑜𝑜 20// 𝛽𝛽3 + 1 𝑟𝑟𝑒𝑒𝑒 + 2//10 2//10 𝑉𝑉𝑜𝑜 𝐴𝐴 85.7
𝐴𝐴 = = × 𝐴𝐴𝑓𝑓 = = = = 8.96𝑉𝑉/𝑉𝑉
𝑉𝑉1 10 1//9 2//10 + 𝑟𝑟𝑒𝑒𝑒 𝑉𝑉𝑠𝑠 1 + 𝛽𝛽𝛽𝛽 1 + 0.1 × 85.7
𝑟𝑟𝑒𝑒𝑒 + 𝑟𝑟𝑒𝑒𝑒 + + 𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖 1 + 𝛽𝛽𝛽𝛽 = 21 × 9.57 = 201𝑘𝑘𝑘
𝛽𝛽1 + 1 𝛽𝛽2 + 1
= 85.7𝑉𝑉/𝑉𝑉 𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 − 𝑅𝑅𝑠𝑠 = 201 − 10 = 191𝑘𝑘𝑘
𝑅𝑅𝑖𝑖 = 10 + 1//9 + 𝛽𝛽1 + 1 𝑟𝑟𝑒𝑒𝑒 + 𝑟𝑟𝑒𝑒𝑒 = 21𝑘𝑘𝑘 𝑅𝑅𝑜𝑜 181
𝑅𝑅𝑜𝑜𝑜𝑜 = = = 18.9Ω
20 1 + 𝛽𝛽𝛽𝛽 9.57
𝑅𝑅𝑜𝑜 = 2// 9 + 1 // 𝑟𝑟𝑒𝑒𝑒 + = 181Ω 1 1 1
𝛽𝛽3 + 1 = + ⇒ 𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = 19.1Ω
𝑅𝑅𝑜𝑜𝑜𝑜 𝑅𝑅𝐿𝐿 𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜

Exercises
• 10.8 For the series-shunt amplifier in Fig. 10.9(c), which has considered in Exercise
10.5, find A, β, Af, Rin, and Rout. Neglect ro of Q. Compare results to those obtained in
Exercise 10.6
• The feedback network is composed of the voltage-driver
resistors R1 and R2
• (a) The loading effect of the feed-back network at the
input is: R1||R2
• (b) The loading effect of the feed-back network at the
output is: R1+R2
For the CG amplifier
A g m  RD || ( R1 + R2 ) 
VF R1
β
= =
Vo R1 + R2
A g m  RD || ( R1 + R2 )  g m RD
=Af = =
1+ β A  R1  RD (1 + g m R1 )
1+  g 
 m D R || ( R + R )
2  1 +
R1 + R2
 R1 + R2 
1
1
Ri = 1/ g m ⇒ Rin = (1 + β A)
gm
RD || ( R1 + R2 )
Ro RD || ( R1 + R2 ) ⇒ R=
=
1+ β A
out

10.5 Other Feedback-Amplifier Types
• Three other types of feedback amplifier:
– The feedback transconductance amplifier, the feedback current amplifier,
and the feedback transresistance amplifier
• Basic principles
1. Sensing: Sample the output signal of interest. If Vo is the output signal
of interest (voltage and transresistance amplifier), the feedback is
connected in parallel (or shunt) with the amplifier output node. If Io is
the output signal of interest (current and transconductance amplifier),
the feedback is connected in series with the amplifier output node.

1. Sensing :

2. Mixing: If the input signal is a voltage (voltage and transconductance
amplifier), the signal source is represented by it’s The’venin signal
equivalent and the feedback voltage signal Vf is connected in series
with the input source Vs. If the input signal is a current (current and
transresistance amplifier), the Norton form is used to represent the
signal source, and the feedback current signal If is connected in parallel
(shunt) with the input signal source Is

2. Mixing :

Basic principles
3. Feedback:
Amplifier Type Appropriate Feedback Topology Af
Voltage Series-Shunt Vo/Vs
Transconductance Series-Series Io/Vs
Current Shunt-Series Io/Is
Transresistance Shunt-Shunt Vo/Is
The appropriate feedback topology not only stabilizes the gain of

interest, but also makes the input and output resistances more ideal
4. Input and output resistance. The increase or decrease of the input or
output resistance depends on the type of connection: Series connection
always increases the resistance and parallel (shunt) connect always
decrease the resistance by amount of feedback (1+Aβ)
Ex. Current amplifier:
Shunt connection at input: Rif = Ri (1 + βA)
Series connection at output Rof = Ro (1 + βA)
Basic principles (Cont.)
5. Dimensions of A, β, Aβ, and Af : Depending on the amplifier type, A, β,
and Af have the dimensions of V/V, A/A, V/A, or A/V. However, Aβ is
always dimensionless
6. Determine β and Af|ideal: To determine β, we set A=∞ => Af|ideal=1/β
(a) For the voltage (series) mixing case, this results in Vi=0 and hence
Ii=0
(b) For the current (shunt) mixing, A=∞ leads to Ii=0 and corresponding
Vi=0
7. Analysis using the loop gain: The loop gain (Aβ) together with the
feedback factor β can be used to determine the open-loop gain A and
hence the closed-loop gain Af. But the input and output resistances
can’t be found => We need to obtain the A circuit using the systematic
analysis approach (illustrated by examples)

Exercises
• 10.9 A feeback current amplifier is to have an ideal closed-loop gain of 10 A/A. What
is the required value of the feedback factor β? If the open-loop amplifier has a gain of
1000 A/A, an input resistance of 1kΩ, and an output resistance of 100 kΩ, find the
realized values of closed-loop gain, input resistance, and output resistance.
1
𝐴𝐴𝑓𝑓|𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = = 10 ⇒ 𝛽𝛽 =0.1
𝛽𝛽
𝐴𝐴 1000
𝐴𝐴𝑓𝑓 = = = 9.9𝐴𝐴/𝐴𝐴
1 + 𝛽𝛽𝛽𝛽 1 + 1000 × 0.1
𝑅𝑅𝑖𝑖 1000
𝑅𝑅𝑖𝑖𝑖𝑖 = =
1+𝛽𝛽𝛽𝛽 1+1000×0.1
= 9.9Ω
𝑅𝑅𝑜𝑜𝑓𝑓 = 𝑅𝑅𝑜𝑜 1 + 𝛽𝛽𝐴𝐴 = 100 × 1 + 1000 × 0.1 = 10.1𝑀𝑀Ω

Feeback Transconductance Amplifier
• Series-Series Feedback Amplifier
• (a) Ideal structure; (b) equivalent circuit
– It consists of a unilateral open-loop amplifier (A=Io/Vi is a

transconductance) and an ideal feedback network (β is a transresistance)

Ideal Case
Io A
Af ≡ =
Vs 1 + Aβ
Vs Vs V V + βI o V + βAVi
Rif ≡ = = Ri s = Ri i = Ri i
I i Vi / Ri Vi Vi Vi
= Ri (1 + Aβ )
– Increase the input resistance by a factor equal to the amount of feedback

– Only depend on the method of mixing
– Series mixing always increases the input resistance

Ideal Case (Cont.)
– Reduce the Vs to zero and apply a test current It at the output
V
Rof ≡
It
Vi = −V f = − βI o = − βI t
V = (I t − AVi )Ro = (I t + AβI t )Ro
Rof = (1 + Aβ )Ro
– Increase the output resistance by a factor equal to the amount of

feedback
– Depend only on the method of sampling
– By intuitive : the negative feedback tries to make IO constant in spite of
changes in the output voltage, which means increased output resistance
– Voltage (shunt) sampling reduces the output resistance, current (series)
sampling increases it
Practical Case
• Devise a simple method for finding A and β
• Block diagram of a practical series-series feedback amplifier

• The two-port feedback network in terms of its z parameters
– This is the only parameter set that represents the feedback network by a
series network at port 1 and a series network at port 2
– The forward transmission through the feedback network can be

neglected, thus assume |z21|feedback network << |z21|basic amplifier
– If the basic amplifier is unilateral, a situation that is obtained when
|z12|basic amplifier<< |z12 |feedback network

– The A circuit is composed of the basic amplifier augmented at the input
with Rs and z11 and augmented at the output with RL and z22
– Replace z11 and z22 as R11 and R22, respectively.
– The loading effect is found by looking into the appropriate port of the
feedback network while the other port is open-circuited or short-circuited
so as to destroy the feedback
If the connection is a shunt one, we short-circuit the port; if it is a series
one, we open-circuit it

• Determination of β β = z12 ≡
V1
I2 I1 = 0
– The feedback network samples the output current [I2=Io] and provides a
voltage [Vf=V1] that is mixed in series with the input source
– The series connection at the input suggests that β should be found with
port 1 open-circuited
• Conceptual view of opening a voltage-current (series-series)
feedback loop with proper loading V
Z11 = 1 V1
β = Z12 ≡
I1 I 2 =0 I2 I1 = 0
V2
Z 22 =
I2 I1 = 0
– (a) : Z model ⇒ voltage-current

z22 V1=Z11I1+Z12I2
z11
V2=Z21I1+Z22I2
Design Parameters

Design Parameters (Cont.)

Summary
feedback transconductance amplifier (series-series)

Summary (Cont.)
• Note that
– Ri and Ro are the input and output resistance of the A circuit which can
be determined by breaking the output loop and looking between Y and Y’
– Rif and Rof are the input and output resistance of the feedback amplifier
exclude RS and RL. These are denoted Rin and Rout in Fig. 9.20(a) and
can be easily determined as
Rin = Rif − RS
Rout = Rof − RL

Example 10.7
• Fig. 10.32(a) shows a feedback transconductance amplifier utilizing an op
amp together with an NMOS transistor. The feedback network consists of a
resistance RF that senses the output current Io (recall that the drain and
source currents of the MOSFET are equal) and provides a feedback voltage
that is subtracted from Vs by means of the differencing action of the op-amp
input. Observe that the feedback topology is series-series, which is uniquely
appropriate for the transconductance amplifiers.
The op amp has a gian µ=1000 V/V, a differential input resistance
Rid=100kΩ, and an output resistaor ro1= 1kΩ, The MOSFET gm=2mA/V,
ro2=20kΩ, and RF=1kΩ
• (a) Find closed-loop gain Af|iedal ≡Io/Vs and β≡Vf/Io
𝐼𝐼𝑜𝑜 1
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ≡ = = 1𝑚𝑚𝑚𝑚/𝑉𝑉
𝑉𝑉𝑠𝑠 𝑅𝑅𝐹𝐹
1
𝛽𝛽 = = 𝑅𝑅𝐹𝐹 = 1𝑉𝑉/𝑚𝑚𝑚𝑚

Example 10.7
• (b) Find the A circuit and determine A, Ri and Ro
𝑅𝑅𝟏𝟏𝟏𝟏 ≡ 𝑅𝑅𝑭𝑭 = 𝟏𝟏𝟏𝟏Ω

𝑅𝑅𝟐𝟐𝟐𝟐 ≡ 𝑅𝑅𝑭𝑭 = 𝟏𝟏𝟏𝟏Ω
𝑅𝑅𝑖𝑖𝑖𝑖 𝜇𝜇𝑉𝑉𝑖𝑖𝑖𝑖 𝑟𝑟𝑜𝑜𝑜
𝑉𝑉𝑖𝑖𝑖𝑖 = 𝑉𝑉𝑖𝑖 , 𝐼𝐼 = , 𝐼𝐼𝑜𝑜 = 𝐼𝐼
𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝐹𝐹 1 𝑅𝑅𝐹𝐹 + 𝑟𝑟𝑜𝑜𝑜
+ 𝑅𝑅𝐹𝐹 ||𝑟𝑟𝑜𝑜𝑜
𝑔𝑔𝑚𝑚
𝐼𝐼 𝑅𝑅 𝜇𝜇 𝑟𝑟𝑜𝑜𝑜 100 1000 20
𝐴𝐴 ≡ 𝑜𝑜 = 𝑖𝑖𝑖𝑖 1 = 1 = 649.2𝑚𝑚𝑚𝑚/𝑉𝑉
𝑉𝑉𝑖𝑖 𝑅𝑅𝑖𝑖𝑖𝑖 +𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐹𝐹 ||𝑟𝑟𝑜𝑜𝑜 𝑅𝑅𝐹𝐹 +𝑟𝑟𝑜𝑜𝑜 100+1 + 1||20 1+20
𝑔𝑔𝑚𝑚 2
𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝐹𝐹 = 100 + 1 = 101𝑘𝑘Ω
𝑅𝑅𝑜𝑜 = 𝑟𝑟𝑜𝑜𝑜 + 1 + 𝑔𝑔𝑚𝑚 𝑟𝑟𝑜𝑜𝑜 𝑅𝑅𝐹𝐹 = 20 + 1 + 2 × 20 × 1 = 61𝑘𝑘Ω
Example 10.7
• (c) Find Af, Rin and Rout
𝐴𝐴𝛽𝛽 = 649.2 × 1 = 649.2, 1+ Aβ =1+649.2=650.2

𝐼𝐼 𝐴𝐴 649.2 mA
𝐴𝐴𝑓𝑓 ≡ 𝑜𝑜 = = = 0.998
𝑉𝑉𝑠𝑠 1+𝐴𝐴𝛽𝛽 650.2 V
𝑅𝑅𝑖𝑖𝑛𝑛 = 𝑅𝑅𝑖𝑖𝑓𝑓 = 1+ 𝐴𝐴𝛽𝛽 𝑅𝑅𝑖𝑖 = 650.2 × 101 = 65.7𝑀𝑀Ω
𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜𝑓𝑓 = 1+ 𝐴𝐴𝛽𝛽 𝑅𝑅𝑜𝑜 = 650.2 × 61 = 39.7𝑀𝑀Ω

Exercises
• 10.10 If in the circuit of Example 10.7, the gain µ is reduced from 1000 V/V to 100
V/V, find the new values of A, Af, Rin, and Rout
𝐼𝐼𝑜𝑜 𝑅𝑅𝑖𝑖𝑖𝑖 𝜇𝜇 𝑟𝑟𝑜𝑜𝑜 100 100 20

𝐴𝐴 ≡ = = = 64.9𝑚𝑚𝑚𝑚/𝑉𝑉
𝑉𝑉𝑖𝑖 𝑅𝑅𝑖𝑖𝑖𝑖 +𝑅𝑅𝐹𝐹 1 + 𝑅𝑅𝐹𝐹 ||𝑟𝑟𝑜𝑜𝑜 𝑅𝑅𝐹𝐹 +𝑟𝑟𝑜𝑜𝑜 100+1 1+ 1||20 1+20
𝑔𝑔𝑚𝑚 2
𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝐹𝐹 = 100 + 1 = 101𝑘𝑘Ω
𝑅𝑅𝑜𝑜 = 𝑟𝑟𝑜𝑜𝑜 + 1 + 𝑔𝑔𝑚𝑚 𝑟𝑟𝑜𝑜𝑜 𝑅𝑅𝐹𝐹 = 20 + 1 + 2 × 20 × 1 = 61𝑘𝑘Ω
𝐴𝐴𝛽𝛽 = 64.9 × 1 = 64.9, 1+ Aβ =1+64.9=65.9
𝐴𝐴𝑓𝑓 ≡ 𝑜𝑜 = = = 0.985
𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 = 1+ 𝐴𝐴𝛽𝛽 𝑅𝑅𝑖𝑖 = 65.9 × 101 = 6.7𝑀𝑀Ω
𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜𝑜𝑜 = 1+ 𝐴𝐴𝛽𝛽 𝑅𝑅𝑜𝑜 = 65.9 × 61 = 4𝑀𝑀Ω
Exercises
• 10.11 If we want to double the ideal Af of the feedback transconductance amplifier of
Example 10.7, what value should RF be changed to? What should A and Af become?
𝐼𝐼𝑜𝑜 1 𝑚𝑚𝑚𝑚 𝑚𝑚𝑚𝑚

𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ≡ = =1 ⇒ 𝐴𝐴𝑓𝑓 = 2 ⇒ 𝑅𝑅𝐹𝐹 = 500Ω = 𝛽𝛽
𝑉𝑉𝑠𝑠 𝑅𝑅𝐹𝐹 𝑉𝑉 𝑉𝑉
𝐼𝐼 𝑅𝑅 𝜇𝜇 𝑟𝑟𝑜𝑜𝑜 100 100 20
𝐴𝐴 ≡ 𝑜𝑜 = 𝑖𝑖𝑖𝑖 1 = 1 = 982.7𝑚𝑚𝑚𝑚/𝑉𝑉
𝑉𝑉𝑖𝑖 𝑅𝑅𝑖𝑖𝑖𝑖 +𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐹𝐹 ||𝑟𝑟𝑜𝑜𝑜 𝑅𝑅𝐹𝐹 +𝑟𝑟𝑜𝑜𝑜 100+0.5 + 0.5||20 0.5+20
𝑔𝑔𝑚𝑚 2
𝐴𝐴𝛽𝛽 = 982.7 × 0.5 = 491.35, 1+ Aβ =1+491.35=492.35
𝐴𝐴𝑓𝑓 ≡ 𝑜𝑜 = = = 1.996

Example 10.8
• Because negative feedback extends the amplifier bandwidth, it is commonly
used in the design of broadband amplifiers. One such amplifier is the
MC1553. Part of the circuit of the MC1553 is shown in Fig. 10.24 (a). The
circuit shown (called a feedback triple) is composed of three gain stages
with series-series feedback provided by the network composed of RE1, RF,
and RE2.
• Obtain that the feedback network samples the emitter current Io of Q3, and
thus Io is the output quantity of the feedback amplifier. However, practically
speaking, Io is rather difficult to utilize. Thus it is usual to take as the output Ic,
the collector current of Q3. This current is of course almost equal to Io; Ic=αIo.
Thus, as a transconductance amplifier with Ic as the output current, the
output resistance of interest is that labeled Rout in Fig. 10.24(a). In some
applications, Ic is passed through a load resistance, such as RC3, and the
voltage Vo is taken as the output. Assume that the bias circuit, which is not
shown, established IC1=0.6mA, IC2=1mA, and IC3=4mA. Also assume that for
all three transistors, hfe=100 and ro=∞.
• (a) Find an approximate expression and value for the ideal closed-loop gain
Af=Io/Vs and hence for β, Ic/Vs, and Vo/Vs
𝐼𝐼𝐶𝐶 𝐼𝐼𝑜𝑜
≅ = 84 mA/V
𝑉𝑉𝑆𝑆 𝑉𝑉𝑆𝑆
𝑉𝑉𝑜𝑜 −𝐼𝐼𝐶𝐶 𝑅𝑅𝐶𝐶𝐶
=
𝑉𝑉𝑆𝑆 𝑉𝑉𝑆𝑆
= −86 × 0.6 = −50.4
𝑉𝑉𝑓𝑓 𝑅𝑅𝐸𝐸𝐸 𝑉𝑉𝑓𝑓

= 𝐼𝐼𝑜𝑜 𝑅𝑅 ⇒ 𝐼𝐼𝑜𝑜 = 𝑅𝑅 𝑅𝑅𝐸𝐸1 + 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐸𝐸𝐸
𝑅𝑅𝐸𝐸𝐸 𝐸𝐸1 +𝑅𝑅𝐹𝐹 +𝑅𝑅𝐸𝐸𝐸 𝐸𝐸𝐸𝑅𝑅𝐸𝐸𝐸
Substituting Vf=VS
𝐼𝐼𝑜𝑜 1 𝑅𝑅𝐸𝐸1 𝑅𝑅𝐹𝐹
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ≡ = 1+ +
𝑉𝑉𝑆𝑆 𝑅𝑅𝐸𝐸1 𝑅𝑅𝐸𝐸𝐸 𝑅𝑅𝐸𝐸𝐸
1 100 640
=
100
1+
100
+
100
=84 mA/V
1 𝑅𝑅𝐸𝐸1 𝑅𝑅𝐹𝐹
𝛽𝛽 = = 𝑅𝑅𝐸𝐸1 � 1 + + = 0.0119 𝑉𝑉/𝑚𝑚𝑚𝑚
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑅𝑅𝐸𝐸𝐸 𝑅𝑅𝐸𝐸𝐸
(b) Use feedback analysis to find A, β, Af, Vo/Vs, Rin, and Rout. For the
calculation Rout, assume that ro of Q3 is 25kΩ.
4mA
1mA
0.6mA

A Circuit
hfe=100
4mA
1mA
Vc2
0.6mA
Vc1
(a) 𝑉𝑉𝑐𝑐𝑐 −𝛼𝛼 𝑅𝑅𝐶𝐶𝐶 //(ℎ𝑓𝑓𝑓𝑓 + 1)𝑟𝑟𝑒𝑒𝑒

= = −14.92𝑉𝑉/𝑉𝑉
α =
100 /101 0.99 𝑉𝑉𝑖𝑖 𝑟𝑟𝑒𝑒𝑒 + 𝑅𝑅𝐸𝐸𝐸 // 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐸𝐸𝐸
VT 25 × 0.99 𝑉𝑉𝑐𝑐𝑐 −𝛼𝛼 𝑅𝑅𝐶𝐶𝐶 //(ℎ𝑓𝑓𝑓𝑓 + 1) 𝑟𝑟𝑒𝑒𝑒 + 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐸𝐸𝐸 //𝑅𝑅𝐸𝐸𝐸
re1 = = 41.3Ω = = −131.2𝑉𝑉/𝑉𝑉
IC1 α 0.6 𝑉𝑉𝑐𝑐𝑐 𝑟𝑟𝑒𝑒𝑒
VT 25 × 0.99 𝐼𝐼𝑜𝑜 𝐼𝐼𝑒𝑒𝑒 1
=
re 2 = = 24.8Ω = = = 10.6𝑚𝑚𝑚𝑚/𝑉𝑉
IC 2 α 1 𝑉𝑉𝑐𝑐𝑐 𝑉𝑉𝑏𝑏𝑏 𝑟𝑟𝑒𝑒𝑒 + 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐸𝐸𝐸 //𝑅𝑅𝐸𝐸𝐸
VT 25 × 0.99 𝐼𝐼𝑜𝑜
r= = = 6.2Ω 𝐴𝐴 ≡ = −14.92 × −131.2 × 10.6 × 10−3 = 20.7𝐴𝐴/𝑉𝑉
IC 3 α
e3
4 𝑉𝑉𝑖𝑖

𝑉𝑉𝑓𝑓 𝑅𝑅𝐸𝐸𝐸
𝛽𝛽 ≡ = × 𝑅𝑅𝐸𝐸𝐸 = 11.9Ω
𝐼𝐼𝑜𝑜 𝑅𝑅𝐸𝐸𝐸 + 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐸𝐸𝐸
𝐼𝐼𝑜𝑜 𝐴𝐴 20.7
𝐴𝐴𝑓𝑓 ≡ = = = 83.7𝑚𝑚𝑚𝑚/𝑉𝑉
𝑉𝑉𝑠𝑠 1 + 𝐴𝐴𝐴𝐴 247.3
𝑉𝑉𝑜𝑜 −𝐼𝐼𝑐𝑐 𝑅𝑅𝐶𝐶𝐶
(c) = = −𝐴𝐴𝑓𝑓 𝑅𝑅𝐶𝐶𝐶 = −50.2𝑉𝑉/𝑉𝑉
𝑉𝑉𝑠𝑠 𝑉𝑉𝑠𝑠
𝑅𝑅𝑖𝑖 = ℎ𝑓𝑓𝑓𝑓 + 1 𝑟𝑟𝑒𝑒1 + 𝑅𝑅𝐸𝐸1 // 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐸𝐸2 = 13.11𝑘𝑘𝑘𝑘

𝑅𝑅𝑖𝑖𝑛𝑛 = 𝑅𝑅𝑖𝑖𝑓𝑓 = 𝑅𝑅𝑖𝑖 1 + 𝐴𝐴𝐴𝐴 = 3.24𝑀𝑀𝑀𝑀
𝑅𝑅𝐶𝐶2
𝑅𝑅𝑜𝑜 = 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝐸𝐸1 //𝑅𝑅𝐸𝐸2 + 𝑟𝑟𝑒𝑒3 + = 143.9𝛺𝛺
ℎ𝑓𝑓𝑓𝑓 + 1
𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜 1 + 𝐴𝐴𝐴𝐴 = 35.6𝑘𝑘𝑘𝑘
𝑟𝑟𝜋𝜋3
𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑟𝑟𝑜𝑜 + 𝑅𝑅𝑜𝑜𝑜𝑜 // 𝑟𝑟𝜋𝜋3 + 𝑅𝑅𝐶𝐶2 1 + 𝑔𝑔𝑚𝑚3 𝑟𝑟𝑜𝑜3
𝑟𝑟𝜋𝜋3 + 𝑅𝑅𝐶𝐶2
0.625
= 25 + 35.6// 0.625 + 5 1 + 160 × 25
0.625 + 5
= 2.19𝑀𝑀𝑀𝑀
(d)
參考第七章59頁講義
𝑟𝑟𝜋𝜋𝜋
𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑟𝑟𝑜𝑜 + 𝑅𝑅𝑜𝑜𝑜𝑜 // 𝑟𝑟𝜋𝜋𝜋 + 𝑅𝑅𝐶𝐶𝐶 1 + 𝑔𝑔𝑚𝑚𝑚 𝑟𝑟𝑜𝑜𝑜
𝑟𝑟𝜋𝜋𝜋 + 𝑅𝑅𝐶𝐶𝐶
• Consider the ro effect, find the output impedance of CE amplifier with emitter
resistor
I x ⋅ Re ⋅ rπ
Vπ = −
Rb + Re + rπ
{
Vx + g mVπ ro =I x ro +  Re // ( rπ + Rb )  }
 Re ⋅ rπ 
Vx = I x ro +  Re // ( rπ + Rb )  + g m ro 
 Rb + R e + rπ 
Vx Re ⋅ ( rπ + Rb ) Re ⋅ rπ
Ro = =+ ro  Re // ( rπ + Rb )  + g m ro
Ix Rb + Re + rπ Re ⋅ ( rπ + Rb )
 rπ 
ro +  Re // ( rπ + Rb )  1 + g m ro
Ro = 
 ( rπ + Rb ) 

Exercises
• D10.12 For the feedback triple in Fig. 10.24(a), analyzed in Example 10.8, modify the
value of RF to obtain a closed-loop transconductance Io/Vs of approximately 100
mA/V. Assume that the loop gain remains large. What is the new value of RF? For
this value, what is the approximate value of the voltage gain if the output voltage is
taken at the collector of Q3?
V f' RE 2
β≡ '
= × RE1
I o RE 2 + RF + RE1
Io A 1 R + RF + RE1
Af ≡ = ≅= E 2 = 100mA / V
Vs 1 + Aβ β RE 2 × RE1
100 + RF + 100
⇒ = 0.1 ⇒ RF = 800Ω
100 ×100
Vo − I c RC 3
= = − Af RC 3 = −100mA / V × 600Ω = −60V / V
Vs Vs

Exercises
• 10.13 Determine the loop gain of the feedback amplifier of Fig. 10.24(a). Set Vs=0,
break the loop between the collector of Q1 and the base of Q2, apply a voltage Vi to
the base of Q2, and connect a resistance equal to rπ2 between the collector Q1 and
ground. Find Aβ as (-Vr/Vt) where Vr≡Vc1.
Vr I c 2 I b 3 I e 3 I f I e1 I c1 Vr
Aβ = − = −
Vt Vt I c 2 I b 3 I e 3 I f I e1 I c1
I c 2 = g m 2Vt = 40Vt
RC 2 5
I b3 = I c 2 = Ic2 = 0.347 I c 2
Ri 3 + RC 2 5 + 9.42
Ri 3 = (β + 1){re 3 + RE 2 || (RF + (RE1 || re1 ))}
= 101{0.00625 + [0.1 || 0.64 + (0.1 || 0.0471)]}
= 9.42

Exercises (Cont.)
Vr I c 2 I b 3 I e3 I f I e1 I c1 Vr
Aβ = − = −
Vt Vt I c 2 I b 3 I e3 I f I e1 I c1
I c 2 = 40Vt
I b 3 = 0.347 I c 2
I e3 = (β + 1)I b 3 = 101I b 3
RE 2
I f = I e3
RE 2 + RF + (RE1 || re1 )
100
= I e3 = 0.13I e 3
100 + 640 + (100 || 41.7 )
RE1 100
I e1 = I f = If = 0.706 I f
RE1 + re1 100 + 41.7
I c1 = αI e1 = 0.99 I e1
Vr = − I c1 (RC1 || rπ 2 ) = − I c1 (9 || 2.5) = −1.957 I c1
Vr
Aβ = − = 249.3
Vt
Feedback Transresistance Amplifier
• Shunt-Shunt Feedback Amplifier
• Ideal structure
– The A circuit has an input resistance Ri, a transrresistance A, and an
output resistance Ro, the β circuit is a voltage-controlled current source,
and β is transconductance

Ideal Case
Vo A
Af ≡ =
I s 1 + Aβ
Vi Vi Vi
Rif ≡ = =
I s I i + βVo I i + βAI i
Ri
=
(1 + Aβ )
Vt − AI i
I=
Ro
AIi for I s = 0, I i = − I f = − βVo = − βVt
Vt + AβVt Vt Ro
I= ⇒ Rof ≡ =
Ro I 1 + Aβ
Practical Case
• Block diagram of a practical shunt-shunt feedback amplifier

• The two-port feedback network in terms of its y parameters
shunt network at port 1 and a shunt network at port 2
– If the basic amplifier is unilateral. The first assumption is justified when
the reverse y parameters of the basic amplifier and of the feedback
network satisfy the condition |y12|basic amplifier<< |y12|feedback network
– The second assumption is justified when the forward y parameters
satisfy the condition : |y21|feedback network << |y21|basic amplifier
• Conceptual view of opening a current-voltage (shunt-shunt)
I
feedback loop with proper loading Y = I1 11
1
V 1 V2 = 0 β = Y12 ≡
V2 V1 = 0
I
Y22 = 2
V2 V1 = 0
– (b) : Y model ⇒ current-voltage

I1=Y11V1+Y12V2
Y11-1 98
Y -1 I2=Y21V1+Y22V2
22
Shuenn-Yuh Lee EE/NCKU Communication and Biologic IC Lab. Chapter 10
Design Parameters


Summary
voltage-sampling shunt-mixing (shunt-shunt) case of Fig. 9.25

Summary (Cont.)
• Note that
exclude RS and RL. These are denoted Rin and Rout in Fig. 8.19 and can
be easily determined as
I1
Y11 =
Rin = 1 ⇐ Rif = Rin||Rs V1 V
 1 1  2 =0
 −
R  I2
 if RS  Y22 =
V2
Rout = 1 ⇐ Rof = Rout||RL V1 = 0
 1 1 
 − 
R R 
 of L 

Example 10.9
• Figure 10.27(a) shows a feedback transresistance amplifier. It is formed by
connecting a resistance RF in the negative-feedback path of a voltage
amplifier with gain µ, an input resistance Rid, and an output resistance ro.
The amplifier µ can be implemented with an op amps, a simple differential
amplifier, a single-ended inverting amplifier, or, in the limit, a single-transistor
CE or CS amplifier. The latter case will be considered in Exercise 10.15. Of
course, the higher the gain µ, the more ideal the characteristics of the
feedback transresistance amplifier will be, simply because of the
concomitant increase in loop gain. Using the following numerical values:
case µ=104 V/V, Rid=100 kΩ, ro=1 kΩ, RF=10 kΩ, and Rs=RL=2 kΩ.

• (a) Find the ideal closed-loop transresistance Vo/Is of the feedback factor β
𝐼𝐼𝑓𝑓 = 𝐼𝐼𝑆𝑆
𝑉𝑉𝑜𝑜 = 𝑉𝑉𝑖𝑖 − 𝐼𝐼𝑓𝑓 𝑅𝑅𝐹𝐹 = 0 − 𝐼𝐼𝑆𝑆 𝑅𝑅𝐹𝐹
𝑉𝑉
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ≡ 𝑜𝑜 = −𝑅𝑅𝐹𝐹 = −10 𝑘𝑘Ω = −10 𝑉𝑉/𝑚𝑚𝑚𝑚
𝐼𝐼𝑆𝑆
1
𝛽𝛽 = = −0.1 𝑚𝑚𝑚𝑚/𝑉𝑉

• (b) Find the A circuit and hence A, Ri, and Ro
𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 𝑅𝑅11 𝑅𝑅𝑠𝑠 = 𝑅𝑅𝑖𝑖𝑖𝑖 𝑅𝑅𝐹𝐹 𝑅𝑅𝑠𝑠 = 100 10 2 = 1.639 𝑘𝑘Ω
𝑅𝑅𝑜𝑜 = 𝑟𝑟𝑜𝑜 𝑅𝑅22 𝑅𝑅𝐿𝐿 = 𝑟𝑟𝑜𝑜 𝑅𝑅𝐹𝐹 𝑅𝑅𝐿𝐿 = 1 10 2 = 625Ω
𝑉𝑉𝑜𝑜 𝑉𝑉𝑖𝑖𝑖𝑖 𝑉𝑉𝑜𝑜 𝑅𝑅𝐹𝐹 ||𝑅𝑅𝐿𝐿 10||2
𝐴𝐴 ≡ = = −𝜇𝜇𝑅𝑅𝑖𝑖 = −104 × 1.639 × = −10244 V/mA
𝐼𝐼𝑖𝑖 𝐼𝐼𝑖𝑖 𝑉𝑉𝑖𝑖𝑖𝑖 𝑟𝑟𝑜𝑜 + 𝑅𝑅𝐹𝐹 ||𝑅𝑅𝐿𝐿 1 + 10||2

• (c) Find expressions for the loop gain, Af, Rif, Rin, Rof, and Rout
𝐴𝐴𝐴𝐴 = −10244 × −0.1 = 1024.4

1 + 𝐴𝐴𝐴𝐴 = 1025.4
𝐴𝐴 −10244 V
𝐴𝐴𝑓𝑓 = = = −9.990
1 + 𝐴𝐴𝐴𝐴 1025.4 mA
𝑅𝑅𝑖𝑖 1.639𝑘𝑘Ω
𝑅𝑅𝑖𝑖𝑖𝑖 = = = 1.6 𝛺𝛺
1 + 𝐴𝐴𝐴𝐴 1025.4
1 1
𝑅𝑅𝑖𝑖𝑖𝑖 = = ≈ 1.6 𝛺𝛺
1 1 1 1
− −
𝑅𝑅𝑖𝑖𝑖𝑖 𝑅𝑅𝑠𝑠 1.6 2000
𝑅𝑅𝑜𝑜 625
𝑅𝑅𝑜𝑜𝑜𝑜 = = = 0.61 𝛺𝛺
1 + 𝐴𝐴𝐴𝐴 1025.4
1 1
𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = = ≈ 0.61 𝛺𝛺
1 1 1 1
− −
𝑅𝑅𝑜𝑜𝑜𝑜 𝑅𝑅𝐿𝐿 0.61 2000

• (d) If instead of a current source Is having a source resistance Rs=2kΩ, the
amplifier is fed from a voltage source Vs having a source resistance Rs=2kΩ,
find an expression for and the value of the voltage gain Vo/Vs
𝑉𝑉𝑠𝑠
𝑉𝑉𝑜𝑜 = 𝐴𝐴𝑓𝑓 𝐼𝐼𝑠𝑠 = 𝐴𝐴𝑓𝑓
𝑅𝑅𝑠𝑠
𝑉𝑉𝑜𝑜 𝐴𝐴𝑓𝑓 −9.990𝑘𝑘𝑘𝑘
= =− = −4.995𝑉𝑉/𝑉𝑉
𝑉𝑉𝑠𝑠 𝑅𝑅𝑠𝑠 2𝑘𝑘𝑘𝑘

Exercises
• 10.14 Determine the loop gain of the amplifier of Fig. 10.27(a) directly. Set Is=0,
replace the amplifier µ with its equivalent circuit, and break the loop at the amplifier
input, ensuring that a resistance equal to Rid is connected across Rs. Show that
µRL (Rid || Rs )
Aβ =
ro [RL + RF + (Rid || Rs )] + RL [RF + (Rid || Rs )]
• Evaluate Aβ using the numerical values given in Example 10.9
𝑅𝑅𝐿𝐿 || 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠

𝑉𝑉𝑟𝑟 = −𝜇𝜇𝑉𝑉𝑡𝑡
𝑟𝑟𝑜𝑜 + 𝑅𝑅𝐿𝐿 ||𝑅𝑅𝐹𝐹 + 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠
𝑅𝑅𝐿𝐿 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠
= −𝜇𝜇𝑉𝑉𝑡𝑡
𝑟𝑟𝑜𝑜 𝑅𝑅𝐿𝐿 + 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠 + 𝑅𝑅𝐿𝐿 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠
𝑉𝑉𝑟𝑟 𝜇𝜇𝑅𝑅𝐿𝐿 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠
𝐴𝐴𝐴𝐴 ≡ − =
𝑉𝑉𝑡𝑡 𝑟𝑟𝑜𝑜 𝑅𝑅𝐿𝐿 + 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠 + 𝑅𝑅𝐿𝐿 𝑅𝑅𝐹𝐹 + 𝑅𝑅𝑖𝑖𝑖𝑖 ||𝑅𝑅𝑠𝑠
104 × 2 × 1.96
𝐴𝐴𝐴𝐴 = = 1035
1 2 + 10 + 1.96 + 2 10 + 1.96

Exercises
• 10.15 For the transresistance amplifier in Fig. E10.15, replace the MOSFET with its
equivalent-circuit model and use feedback analysis to show the following:
(a) For large loop gain (which cannot be achieved here), Af ≡ Vo/Is=-Rf.
I If 1 1
β =f = =
− ⇒ Af == −Rf
Vo −I f × Rf Rf β
(b) A = Vo = I S × ( Rs || R f ) g m ( ro || R f ) = − ( R || R ) g ( r || R )
s f m o f
IS IS
A − ( Rs || R f ) g m ( ro || R f )
= Af =
1 + Aβ 1 + ( Rs || R f ) g m ( ro || R f ) / R f
(c)
Rif = RS // Rin
( Rs || R f )
1 + ( Rs || R f ) g m ( ro || R f ) / R f
1 1 g m ( ro || R f ) 1 1 g m ( ro || R f ) 1 1
= + = + + = +
Rif ( Rs || R f ) Rf Rs R f Rf Rs Rin
Rf
Rf
⇒ Rin =
1 + g m ( ro || R f ) Rf

Exercises (Cont.)
equivalent-circuit model and us feedback analysis to show the following:
ro // R f
(d) Rof =
1 + ( Rs || R f ) g m ( ro || R f ) / R f
1
=
1
+
( Rs || R f ) g m 1 1 ( Rs || R f ) g m
= + + =
1
Rof ro // R f Rf ro R f Rf Rf
ro //
1 + g m ( Rs || R f )
Rf
R= R= ro ||
1 + g m ( Rs || R f )
of out

Exercises (Cont.)
equivalent-circuit model and us feedback analysis to show the following:
(e) For gm=5 mA/V, ro=20 kΩ, Rf=10 kΩ, and Rs=1 kΩ, find A, β, Aβ, Af, Ri, Ro, Rif, Rin,
Rof, and Rout
A =− ( Rs || R f ) g m ( ro || R f ) =−(1//10) × 5 × ( 20 //10 ) =−30.3k Ω
1
β=
− =
−0.1mA / V
Rf
( −30.3)( −0.1) =
Aβ = 3.03
A −30.3
Af = = = −7.52k Ω
1 + Aβ 1 + 3.03
=i RS // R=
R f
= 0.909k Ω
1//10 = 909Ω
= =
Ro ro // R f 20=
//10 6.67k Ω
Ri 909
=
R = = 226Ω
1 + Aβ 1 + 3.03
if
Rf 10
R= = = 0.291k Ω
= 291Ω
1 + g m ( ro || R f ) 1 + 5 × ( 20 //10 )
in
Rf
Rf 10
Rof ro=
|| 20 ||
1 + g m ( Rs || R f ) 1 + 5 × (1//10 ) Rf
= 20 ||1.803
= 1.65=
k Ω Rout
Feedback Current Amplifier
• Shunt-Series Feedback Amplifier
• Ideal structure

Ideal Situation
Io A
Af ≡ =
I s 1 + Aβ
Vi Vi Vi
Rif ≡ = =
I s I i + βI o I i + βAI i
Ri
=
(1 + Aβ )
V
Rof ≡
It
AIi I i = − I f = − βI o = − βI t
V = (I t − AI i )Ro = (I t + AβI t )Ro
Rof = (1 + Aβ )Ro

Practical Case
• Block diagram of a practical shunt-series feedback amplifier

Practical Situation (Cont.)
• The two-port feedback network in terms of its g parameters
parallel network at port 1 and a series network at port 2
– If the basic amplifier is unilateral. The first assumption is justified when
the reverse g parameters of the basic amplifier and of the feedback
network satisfy the condition |g12|basic amplifier<< |g12 |feedback network
– The second assumption is justified when the forward y parameters
satisfy the condition : |g21|feedback network << |g21|basic amplifier
• Conceptual view of opening a current-current (shunt-series)
feedback loop with proper loading I
G11 = 1 I V1 I 2 =0 β = G12 ≡ 1
I2 V1 = 0
V
G22 = 2
I2 V1 = 0
– (d) : G model ⇒ current-current

I1=G11V1+G12I2
G11-1 G22
V2=G21V1+G22I2
Design Parameters


Summary
• A summary of the rules for finding the A circuit and βfor the
voltage-sampling shunt-mixing (shunt-shunt) case of Fig. 8.23

Summary (Cont.)
• Note that
exclude RS and RL. These are denoted Rin and Rout in Fig. 8.19 and can
be easily determined as
Rin = 1
 1 1 
 −
R 
 if RS 
Rout = Rof − RL
• Comment
– Input at gate(source) : gate (base) mode : sense current
source (emitter) mode : sense voltage
– Output at drain (collector) : drain (collector) (shunt) mode : sense voltage
source (emitter) (series) mode : sense current
Example 10.10
• Figure 10.27 shows a feedback current amplifier formed by cascading an
inverting voltage amplifier µ with a MOSFET Q. The output current Io is the
drain current of Q. The feedback network, consisting of resistors R1 and R2,
sense an exactly equal current, namely, the source current of Q, and
provides a feedback current signal that is mixed with Is at the input node.
Note that the bias arrangement is not shown.

• The amplifier µ can be implemented in a variety of ways, including by means
of an op amps, a differential amplifier, or a single-ended inverting amplifier.
The simple approach is to implement µ with a CS amplifier. However, in such
a case the loop gain will be very limited. Assume that the amplifier µ has an
ro1
• Use the following numerical case: µ=1000 V/V, Rid =Rs=∞, Rid=∞, R1=10kΩ,
R2=90kΩ, and for Q: gm=5mA/V and ro|Q=20kΩ

• Use the following numerical case: µ=1000 V/V, Rid =Rs=∞, Rid=∞, R1=10kΩ,
R2=90kΩ, and for Q: gm=5mA/V and ro|Q=20kΩ
• (a) Find the ideal closed-loop gain Io/Is and the feedback factor β
𝑅𝑅
𝐼𝐼𝑜𝑜 = −𝐼𝐼𝑆𝑆 1 + 𝑅𝑅2
1
𝐼𝐼𝑜𝑜 𝑅𝑅2
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ≡ =− 1+ = −10 𝐴𝐴/𝐴𝐴
𝐼𝐼𝑠𝑠 𝑅𝑅1
1 1
𝛽𝛽 ≡ =− = −0.1
𝐴𝐴𝑓𝑓 |𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑅𝑅2
1 + 𝑅𝑅
1

• (b) Find the A circuit and derive expression for A, Ri, and Ro
IQ
𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑠𝑠 𝑅𝑅𝑖𝑖𝑖𝑖 𝑅𝑅1 + 𝑅𝑅2 = 100 𝑘𝑘Ω

𝑉𝑉𝑖𝑖 = 𝐼𝐼𝑖𝑖 𝑅𝑅𝑖𝑖
1
𝐼𝐼𝑄𝑄 = −𝜇𝜇𝑉𝑉𝑖𝑖
1⁄𝑔𝑔𝑚𝑚 + 𝑅𝑅1 ||𝑅𝑅2 ||𝑟𝑟𝑜𝑜𝑜
𝑟𝑟𝑜𝑜𝑜
𝐼𝐼𝑜𝑜 = 𝐼𝐼𝑄𝑄
𝑟𝑟𝑜𝑜𝑜 + 𝑅𝑅1 ||𝑅𝑅2
𝐼𝐼𝑜𝑜 𝑉𝑉𝑖𝑖 𝐼𝐼𝑄𝑄 𝐼𝐼𝑜𝑜 𝑅𝑅𝑖𝑖 𝑟𝑟𝑜𝑜𝑜
𝐴𝐴 ≡ = = −𝜇𝜇
𝐼𝐼𝑖𝑖 𝐼𝐼𝑖𝑖 𝑉𝑉𝑖𝑖 𝐼𝐼𝑄𝑄 1⁄𝑔𝑔𝑚𝑚 + 𝑅𝑅1 ||𝑅𝑅2 ||𝑟𝑟𝑜𝑜𝑜 𝑟𝑟𝑜𝑜𝑜 + 𝑅𝑅1 ||𝑅𝑅2
𝑅𝑅𝑖𝑖 𝑅𝑅1 𝑅𝑅2 𝑟𝑟𝑜𝑜𝑜
= −𝜇𝜇 = −10.764 𝐴𝐴/𝐴𝐴
𝑅𝑅1 ||𝑅𝑅2 1⁄𝑔𝑔𝑚𝑚 + 𝑅𝑅1 𝑅𝑅2 𝑟𝑟𝑜𝑜𝑜
𝑅𝑅𝑜𝑜 = 𝑟𝑟𝑜𝑜𝑜 + 1 + 𝑔𝑔𝑚𝑚 𝑟𝑟𝑜𝑜𝑜 𝑅𝑅1 ||𝑅𝑅2 = 929 𝑘𝑘Ω
• (c) Find Af, Rif, Rin, Ro, Rof, and
𝐴𝐴 10764
𝐴𝐴𝑓𝑓 = =− = −9.991𝐴𝐴/𝐴𝐴
1 + 𝐴𝐴𝐴𝐴 1 + 1076.4
𝑅𝑅𝑖𝑖 100𝑘𝑘𝑘𝑘
𝑅𝑅𝑖𝑖𝑖𝑖 = = = 92.8𝛺𝛺
1 + 𝐴𝐴𝐴𝐴 1 + 1076.4
𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑅𝑅𝑠𝑠 = ∞ ⇒ 𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 = 92.8𝛺𝛺
𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜𝑜𝑜 = 1 + 𝐴𝐴𝐴𝐴 𝑅𝑅𝑜𝑜 = 1076.4 × 929 = 1𝐺𝐺𝛺𝛺

Exercises (Cont.)
• 10.16 For the amplifier in Example 10.10, find the values of Af, Rin, and Rout when the
value of µ is 10 times lower, that is when µ=100.
𝑅𝑅𝑖𝑖 𝑅𝑅1 ||𝑅𝑅2 ||𝑟𝑟𝑜𝑜𝑜

𝐴𝐴 = −𝜇𝜇
𝑅𝑅1 ||𝑅𝑅2 1⁄𝑔𝑔𝑚𝑚 + 𝑅𝑅1 ||𝑅𝑅2 ||𝑟𝑟𝑜𝑜𝑜
for 𝜇𝜇 = 100, 𝑅𝑅1 = 10𝑘𝑘𝑘, 𝑅𝑅2 = 90𝑘𝑘𝑘
𝑔𝑔𝑚𝑚 = 5𝑚𝑚𝑚𝑚/𝑉𝑉, 𝑟𝑟𝑜𝑜𝑜 = 20𝑘𝑘𝑘
𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑠𝑠 ||𝑅𝑅𝑖𝑖𝑖𝑖 || 𝑅𝑅1 + 𝑅𝑅2 = ∞||∞||100 = 100𝑘𝑘𝑘
100 10||90||20
𝐴𝐴 = −100 = −1076.4𝐴𝐴/𝐴𝐴
10||90 0.2 + 10||90||20
𝑅𝑅1 10
𝛽𝛽 = − =− = −0.1𝐴𝐴/𝐴𝐴
𝑅𝑅1 + 𝑅𝑅2 10 + 90
1076.4
𝐴𝐴𝑓𝑓 = − = −9.91𝐴𝐴/𝐴𝐴
1 + 1076.4 × 0.1
𝑅𝑅𝑖𝑖 100𝑘𝑘𝑘
𝑅𝑅𝑖𝑖𝑖𝑖 = = = 920Ω
1 + 𝐴𝐴𝐴𝐴 1 + 107.64
𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 = 920Ω
𝑅𝑅𝑜𝑜 = 𝑟𝑟𝑜𝑜𝑜 + 1 + 𝑔𝑔𝑚𝑚 𝑟𝑟𝑜𝑜𝑜 𝑅𝑅1 ||𝑅𝑅2 = 929𝑘𝑘𝑘
𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜 1 + 𝐴𝐴𝐴𝐴 = 929 × 108.64 = 101𝑀𝑀𝑀

Exercises (Cont.)
• 10.17 If in the circuit in Fig. 10.30(a), R2 is short-circuited, find the ideal value of Af.
For the case Rs=Rid=∞, give expressions for Ri, Ro, A, β, Af, Rin, and Rout
𝑅𝑅1
𝛽𝛽 = − � = −1
𝑅𝑅1 + 𝑅𝑅2 𝑅𝑅
2 =0
𝐴𝐴𝑓𝑓 = −1𝐴𝐴/𝐴𝐴
𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑠𝑠 ||𝑅𝑅𝑖𝑖𝑖𝑖 || 𝑅𝑅1 + 𝑅𝑅2 = ∞||∞||𝑅𝑅1 = 𝑅𝑅1
𝑅𝑅𝑜𝑜 = 𝑟𝑟𝑜𝑜𝑜 + 1 + 𝑔𝑔𝑚𝑚 𝑟𝑟𝑜𝑜𝑜 𝑅𝑅1 ||𝑅𝑅2 = 𝑟𝑟𝑜𝑜𝑜
𝑅𝑅𝑖𝑖 𝑟𝑟𝑜𝑜𝑜 𝑅𝑅1
𝐴𝐴 = −𝜇𝜇 = −𝜇𝜇 = −𝜇𝜇𝑔𝑔𝑚𝑚 𝑅𝑅1
1⁄𝑔𝑔𝑚𝑚 + 𝑅𝑅1 ||𝑅𝑅2 ||𝑟𝑟𝑜𝑜𝑜 𝑟𝑟𝑜𝑜𝑜 + 𝑅𝑅1 ||𝑅𝑅2 1⁄𝑔𝑔𝑚𝑚
𝐴𝐴 𝜇𝜇𝑔𝑔𝑚𝑚 𝑅𝑅1
𝐴𝐴𝑓𝑓 = =
1 + 𝐴𝐴𝐴𝐴 1 + 𝜇𝜇𝑔𝑔𝑚𝑚 𝑅𝑅1
𝑅𝑅𝑖𝑖 𝑅𝑅1
𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 = =
1 + 𝐴𝐴𝐴𝐴 1 + 𝜇𝜇𝑔𝑔𝑚𝑚 𝑅𝑅1
𝐹𝐹𝐹𝐹𝐹𝐹 𝜇𝜇𝑔𝑔𝑚𝑚 𝑅𝑅1 >> 1 ⇒ 𝑅𝑅𝑖𝑖𝑖𝑖 ≅ 1⁄𝜇𝜇𝑔𝑔𝑚𝑚
𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜 1 + 𝐴𝐴𝐴𝐴 = 𝑟𝑟𝑜𝑜𝑜 × 1 + 𝜇𝜇𝑔𝑔𝑚𝑚 𝑅𝑅1 ≅ 𝜇𝜇 𝑔𝑔𝑚𝑚 𝑟𝑟𝑜𝑜𝑜 𝑅𝑅1


CH 10 Feeback

Uploaded by

Copyright:

Available Formats

CH 10 Feeback

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CH 10 Feeback

Uploaded by

Copyright:

Available Formats

Feedback

• 10.1 The general feedback structure

Shuenn-Yuh Lee EE/NCKU 1 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 2 Communication and Biologic IC Lab. Chapter 10

Rsig + RL' 1 RL'

• The open-loop amplifier gain = A; thus

Shuenn-Yuh Lee EE/NCKU 4 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 6 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 7 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 8 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 10 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 11 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 13 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 16 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 17 Communication and Biologic IC Lab. Chapter 10

Bandwidth modification as a result of feedback

Shuenn-Yuh Lee EE/NCKU 18 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 20 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 21 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 22 Communication and Biologic IC Lab. Chapter 10

– Circuits sensing a voltage must exhibit a high input impedance whereas

• (a) : a common-source stage senses and produce voltages

• Addition of (a) voltages

Shuenn-Yuh Lee EE/NCKU 29 Communication and Biologic IC Lab. Chapter 10

• A transistor amplifier with series-shunt feedback (Fig. 10.8)

Shuenn-Yuh Lee EE/NCKU 31 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 32 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 33 Communication and Biologic IC Lab. Chapter 10

𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 𝑅𝑅1 // 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅𝑠𝑠 𝑅𝑅𝑖𝑖𝑖𝑖

Shuenn-Yuh Lee EE/NCKU 38 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 39 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 42 Communication and Biologic IC Lab. Chapter 10

– Reduce the output resistance by a factor equal to the amount of

Series-Series Feedback Shunt-Shunt feedback

– (a) : Z model ⇒ voltage-current – (c) : H model ⇒ voltage-voltage

Shuenn-Yuh Lee EE/NCKU 45 Communication and Biologic IC Lab. Chapter 10

– Since the feedback network is usually passive, its forward transmission

Shuenn-Yuh Lee EE/NCKU 46 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 47 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 48 Communication and Biologic IC Lab. Chapter 10

4. Find the open-loop gain A according to open-loop circuit in Fig. (e)

𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖 1 + 𝐴𝐴𝐴𝐴 , 𝑅𝑅𝑜𝑜𝑜𝑜 = 𝑅𝑅𝑜𝑜 ⁄(1 + 𝐴𝐴𝐴𝐴)

The actual input and output resistances with feedback

– (c) : H model ⇒ voltage-voltage

Shuenn-Yuh Lee EE/NCKU 50 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 51 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 53 Communication and Biologic IC Lab. Chapter 10

𝑉𝑉𝑜𝑜 𝑅𝑅𝐿𝐿 // 𝑅𝑅2 + 𝑅𝑅1 𝑅𝑅𝑖𝑖𝑖𝑖

𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑠𝑠 + 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑅𝑅1 //𝑅𝑅2 ≈ 111𝑘𝑘𝑘𝑘

Shuenn-Yuh Lee EE/NCKU 55 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 56 Communication and Biologic IC Lab. Chapter 10

V RD1 g m1 RD1 Vo g R g [R || (R1 + R2 )]

Shuenn-Yuh Lee EE/NCKU 57 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 58 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 59 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 60 Communication and Biologic IC Lab. Chapter 10

• A Circuit • Closed loop

Shuenn-Yuh Lee EE/NCKU 61 Communication and Biologic IC Lab. Chapter 10

Shuenn-Yuh Lee EE/NCKU 62 Communication and Biologic IC Lab. Chapter 10