FEEDBACK: Ch.

11
1. Introduction
Feedback can be either negative or positive.
In amplifier design, negative feedback is applied to effect one or more of the following goals:
1. Desensitize the gain: make the gain less sensitive to variations of circuit components.
2. Reduce nonlinear distortion: make the output proportional to the input (make the gain constant)
3. Reduce the effect of noise: minimize the contribution to the output of unwanted electric signals.
4. Control the input and output resistances: raise or lower by an appropriate feedback topology.
5. Extend the bandwidth of the amplifier.
All of the desirable properties above are obtained at the expense of a reduction in gain.
The gain-reduction factor is called the amount of feedback, is the factor by which the circuit is:
desensitized, Rin increases, by which the bandwidth is extended, and so on.
In short, the basic idea of negative feedback is to trade off gain for other desirable properties.
2. The General Feedback Structure
2.1 Signal-Flow Diagram
Figure 11.1 shows the basic structure of a feedback amplifier. The basic amplifier is unilateral.
Assume that the feedback does not change the gain A or xo; the feedback network does not load
the amplifier output.
A: is the open-loop gain
β: is the feedback factor
xi: is the input to the basic amplifier,
xs: is the source signal
𝑥𝑓 : is the feedback signal
𝑥𝑜
𝑥𝑜 = 𝐴𝑥𝑖 → 𝐴 =
𝑥𝑖

𝑥𝑓
𝑥𝑓 = 𝛽𝑥𝑜 → 𝛽=
𝑥𝑜
𝑥𝑖 = 𝑥𝑠 − 𝑥𝑓 The sign – make the feedback negative
 • Find the closed-loop gain or the gain with feedback 𝐴𝑓 :
𝑥𝑜
𝐴𝑓 =
𝑥𝑠
𝐴
𝐴𝑓 =
1 + 𝐴𝛽
Aβ: is called the loop gain.
For the feedback to be negative, the loop gain Aβ > 0, (i.e. xf and xs should have the same sign).
(1 + Aβ): is called the amount of feedback.
If the loop gain Aβ >> 1 (large)
1
𝐴𝑓 ≈
𝛽
• Find 𝑥𝑓 𝑣𝑠 𝑥𝑠
𝑥𝑜 𝐴 𝛽𝑥𝑜 𝐴𝛽
(𝐴𝑓 = = )×𝛽 = =
𝑥𝑠 1 + 𝐴𝛽 𝑥𝑠 1 + 𝐴𝛽
𝐴𝛽
𝑥𝑓 = 𝑥
1 + 𝐴𝛽 𝑠
For Aβ >>1 then 𝑥𝑓 = 𝑥𝑠 → 𝑥𝑖 = 𝑥𝑠 − 𝑥𝑓 = 0

The input differencing circuit is called a comparison circuit. (It is also known as a mixer.)
• Find 𝑥𝑖 𝑣𝑠 𝑥𝑠
𝐴𝛽 𝑥𝑓 1
𝑥𝑓 = 𝑥𝑠 → = 𝑥 = 𝑥𝑖
1 + 𝐴𝛽 𝐴𝛽 1 + 𝐴𝛽 𝑠
1
𝑥𝑖 = 𝑥
1 + 𝐴𝛽 𝑠
For Aβ >> 1, 𝑥𝑖 ≈ 0 𝑎𝑛𝑑 𝑥𝑠 ≈ 𝑥𝑓

1. It is the tracking of the two input terminals of an op amp (𝑥𝑠 𝑎𝑛𝑑 𝑥𝑓 ).

2. the negative feedback reduces 𝑥𝑖 by the amount of feedback (1 + Aβ).
3. this reduction of input signal results in the increased linearity of the feedback amplifier.
The feedback network usually consists:
1. passive components (resistances),
2. the advantage of negative feedback in obtaining accurate, predictable, and stable gain.
 3. the overall gain will have very little dependence on the gain of the basic amplifier, A,
4. a desirable property because the gain A is usually a function of many manufacturing and
application parameters, some of which might have wide tolerances.
5. we will consider (1/β) to be the ideal value of Af .

2.2 Systematic Analysis of Feedback Voltage Amplifiers

Disassembles the feedback voltage-amplifier circuit so as to obtain
1. The “A circuit,” from which we can determine
i. the open-loop gain A,
ii. the input and output resistances, and
2. the “β circuit” to find the feedback factor β.
3. using the feedback formulas to find the characteristic parameters of the feedback amplifier
i. the closed-loop gain 𝐴𝑓 and
ii. the input and output resistances with feedback.
There are two cases
1. The Ideal Case: in which the feedback network does not load the basic amplifier.
2. The Practical (General) Case: in which the feedback network loads the basic amplifier, and
also there is a finite source resistance Rs and a finite load resistance RL.
A. The Ideal Case
series–shunt is the appropriate feedback topology for a voltage amplifier.

The ideal structure of the series–shunt feedback amplifier is shown in Fig. (a). It consists of
1. the A circuit:
i. a unilateral open-loop amplifier and
ii. an ideal voltage-sampling.
iii. It has an input resistance Ri,
iv. an open-circuit voltage gain A,
v. and an output resistance Ro.
vi. Assume that the source is ideal 𝑅𝑠 = 0
vii. 𝑅𝐿 = ∞ (no load resistance).
2. the β circuit: is a voltage-mixing feedback network. (β circuit does not load the A circuit, i.e.
connecting the β circuit does not change the value of A (defined as A ≡ Vo/Vi)).
Therefore, the closed-loop voltage gain 𝐴𝑓 is given by
𝑉𝑜 𝐴
𝐴𝑓 = =
𝑉𝑠 1 + 𝐴𝛽
The equivalent-circuit model of the series–shunt feedback amplifier is shown in Fig. (b).
𝑅𝑖𝑓 𝑎𝑛𝑑 𝑅𝑜𝑓 are the input and output resistances of the feedback amp. Respectively. 𝐴𝑓 is the
open-circuit voltage gain of the feedback amplifier,
1. Find 𝑅𝑖𝑓 = 𝑉𝑠 /𝐼𝑖
𝑉𝑠
𝑉𝑖 =
1 + 𝐴𝛽
𝑉𝑖 𝑉𝑠
𝐼𝑖 = =
𝑅𝑖 (1 + 𝐴𝛽)𝑅𝑖
𝑉𝑠
𝑅𝑖𝑓 = = (1 + 𝐴𝛽)𝑅𝑖
𝐼𝑖
• The series-mixing feedback at the input increases the amplifier input resistance by (1+Aβ),
a highly desirable property for a voltage amplifier.
2. Find 𝑅𝑜𝑓 = 𝑉𝑥 /𝐼𝑥

In Fig. a, set Vs = 0 and apply a test voltage Vx between the output terminals, as shown in Fig. c.
𝑉𝑥
𝑅𝑜𝑓 =
𝐼𝑥
𝑉𝑥 − 𝐴𝑉𝑖
𝐼𝑥 =
𝑅𝑜
𝑉𝑖 = −𝑉𝑓

𝑉𝑓 = 𝛽𝑉𝑜 = 𝛽𝑉𝑥

𝑉𝑖 = −𝛽𝑉𝑥
𝑉𝑥 (1 + 𝐴𝛽)
𝐼𝑥 =
𝑅𝑜
𝑉𝑥 𝑅𝑜
𝑅𝑜𝑓 = =
𝐼𝑥 (1 + 𝐴𝛽)
The shunt sampling (or voltage sampling) at the output results in a decrease in the amplifier output
resistance by a factor of (1+Aβ),

B. The Practical (General) Case

In a practical series–shunt feedback amplifier, the feedback network will not be an ideal voltage-
controlled voltage source.
the feedback network is usually resistive and hence will load the basic amplifier and thus affect
the values of A, Ri, and Ro.
Given a series–shunt feedback amplifier of Fig. (a), find the A circuit and the β circuit.

Present this cct without derivation, as in Fig. (b). We make the following observations.
1. The A circuit is obtained by augmenting the basic amplifier at the input with Rs and R11, and
at the output RL and R22. R11 and R22 represent the loading effect of the feedback network on
the basic amplifier at the input and the output, respectively.
2. R11 is the resistance looking into port 1 of the feedback network while port 2 is short-circuited.
R22 is the resistance looking into port 2 of the feedback network while port 1 is open-circuited.
As shown in Fig. (c).
Since the feedback network is connected in shunt with the output, shorting its port 2 destroys the
feedback.
Similarly, because the feedback network is connected in series with the input, opening its port 1
destroys the feedback.
It follows that the loading effect of the feedback network is obtained by looking into its
appropriate port while the other port is open-circuited or short-circuited so as to destroy the
feedback.

3. The feedback factor β is the transmission from port 2 to port 1 of the feedback network, with
port 1 open-circuited (which destroys the feedback). Reference to Fig. (d) shows that

𝑉𝑓
𝛽= |
𝑉𝑜 𝐼
1 =0

4. The open-loop
gain A can be
obtained from the
A circuit in Fig. (e) as
𝑉𝑜
𝐴=
𝑉𝑖

5. The values of A and

β can be used to
determine the closed-
loop gain 𝐴𝑓 ,
𝑉𝑜 𝐴
𝐴𝑓 = =
𝑉𝑠 1 + 𝐴𝛽
6. The open-loop input resistance Ri and output resistance Ro can be determined from the A circuit
[see Fig. (e)]. Ri and Ro can be used to determine the input and output resistances with feedback,
𝑉𝑠
𝑅𝑖𝑓 = = (1 + 𝐴𝛽)𝑅𝑖
𝐼𝑖
𝑅𝑜
𝑅𝑜𝑓 =
(1 + 𝐴𝛽)
From Fig. (a) we see that 𝑅𝑖𝑓 is the resistance seen by the ideal signal source Vs. The actual input
resistance of the feedback amplifier 𝑅𝑖𝑛 excludes Rs and is found from 𝑅𝑖𝑓 ,

𝑅𝑖𝑛 = 𝑅𝑖𝑓 − 𝑅𝑠

Similarly, 𝑅𝑜𝑓 is the output resistance of the feedback amplifier including 𝑅𝐿 . The actual output
resistance excludes 𝑅𝐿 (see Fig. (a)) and is found from 𝑅𝑜𝑓 ,
1
𝑅𝑜𝑢𝑡 =
1 1
(𝑅 − 𝑅 )
𝑜𝑓 𝐿

A final and important note:

1. Fig. (b) is only approximately equivalent to the original circuit in Fig. (a). The approximation
is a result of neglecting the small forward transmission in the feedback network relative to the
much larger forward transmission in the basic amplifier.
2. Assume also that the basic amplifier is unilateral (i.e., it does not have internal feedback; all
the feedback occurs in the feedback network and is represented by the feedback factor β.
Example 11.4
Figure (a) shows an op amp connected in the
noninverting configuration.
The op amp has: an
open-circuit voltage gain μ = 104, Rid = 100 kΩ, and
ro = 1 kΩ.
Using the feedback method to analyze the circuit.
a. Determine the feedback network?
b. what is the type of this feedback amp.
c. Find expressions for A, β, the closed-loop gain
Vo/Vs, Rin, and Rout in Fig. a .
d. Find numerical values, given μ =104, Rid = 100 kΩ, ro = 1 kΩ, RL = 2 kΩ, R1 = 1 kΩ, R2 =1
MΩ, and Rs = 10 kΩ.
Solution

1. the feedback network consists of R2 and R1.

2. Series-shunt (current-voltage) feedback amp.
Because the network samples the output voltage Vo
and provides a voltage signal (across R1) that is
mixed in series with the input source Vs. 1 2
3. Find R11 (The loading effect of the feedback
network at the input side): by short-circuiting port
2 of the feedback network (because it is connected
in shunt) and looking into port 1. 𝑅11 = 𝑅1 ∥ 𝑅2
4. Find R22 (The loading effect of the feedback network at the output side): is found by open-
circuiting port 1 (because it is connected in series) and looking into port 2, 𝑅22 = 𝑅1 + 𝑅2
5. Draw the A circuit (as shown in Fig. (b).
6. Find A (open loop gain) from circuit A
For the values given, we find that A ≈ 6000 V/V.

7. Find β: from (β circuit)

𝑉𝑓
𝛽= |
𝑉𝑜 𝐼=0
𝑉𝑓 𝑅1
𝛽= = ≈ 10−3 𝑉/𝑉
𝑉𝑜 𝑅1 + 𝑅2
8. Find 𝐴𝛽 (the loop gain)
𝐴𝛽 = 6000 × 10−3 = 6
9. Find 𝐴𝑓 : The closed loop gain (voltage gain with feedback)
𝑉𝑜 𝐴 6000
𝐴𝑓 = = = = 857 𝑉/𝑉
𝑉𝑠 1 + 𝐴𝛽 7
10. Find 𝑅𝑖𝑓 and Rin

1. 𝑅𝑖𝑓 is found using Figure (a)

𝑉𝑠
𝑅𝑖𝑓 = = (1 + 𝐴𝛽)𝑅𝑖
𝐼𝑖
Ri is the input resistance of the A circuit in Fig. (b):
𝑅𝑖 = 𝑅𝑠 + 𝑅𝑖𝑑 + (𝑅1 ∥ 𝑅2 )
For the values given, Ri = 111 kΩ, resulting in
𝑅𝑖𝑓 = 111𝑘 × 7 = 777 𝑘Ω

2. Rin, indicated in Fig. (a):

𝑅𝑖𝑛 = 𝑅𝑖𝑓 − 𝑅𝑠 = 777 𝑘 − 10 𝑘 = 767 𝑘Ω

10. Find 𝑅𝑜𝑓 is the output resistance of the feedback amplifier, including the load resistance RL,
as indicated in Fig. (a). Rof is given by
𝑅𝑜
𝑅𝑜𝑓 =
(1 + 𝐴𝛽)
Ro is the output resistance of the A circuit. Ro is obtained from Fig. (b) with Vi set to zero,
𝑅𝑜 = 𝑟𝑜 ∥ 𝑅𝐿 ∥ (𝑅1 + 𝑅2 )
For the values given, Ro = 666Ω, and
666
𝑅𝑜𝑓 = = 95.5 Ω
7
Rout, is the output resistance of the feedback amplifier excluding RL. From Fig. (a) we see that
𝑅𝑜𝑓 = 𝑅𝑜𝑢𝑡 ∥ 𝑅𝐿

𝑅𝑜𝑢𝑡 = 100 Ω
Example 11.5
Using the feedback method to
analyze the circuit.
a. Determine the feedback
network?
b. what is the type of this
feedback amp.
c. Find expressions for A, β, the
closed-loop gain Vo/Vs, Rin, and
Rout in Fig. a .
d. 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
1. the feedback network consists of R2 and R1.
2. Series-shunt (current-voltage) feedback amp.
1 2
3. Find R11 (The loading effect of the feedback
network at the input side): by short-circuiting port 2
of the feedback network (because it is connected in
shunt) and looking into port 1. 𝑅11 = 𝑅1 ∥ 𝑅2
4. Find R22 (The loading effect of the feedback network at the output side): is found by open-
circuiting port 1 (because it is connected in series) and looking into port 2, 𝑅22 = 𝑅1 + 𝑅2
5. Draw the A circuit (as shown in Fig.
• The A circuit is shown in Fig. (b).
• The gain A is determined as the
product of the gain of Q1 and the
gain of Q2 as follows:
6. Find β from the β circuit in Fig. (c),

7. Find 𝐴𝛽 (the loop gain)

𝐴𝛽 = 173.9 × 0.1 = 17.39
8. Find the closed-loop gain Af = Vo/Vs

9. Rif and Ri are both infinite because of the infinite input resistance of the MOSFET.
10. Rof and Ro and Rout
Rof = Rout because there is no RL in the cct. RL = ∞

where Ro is the output resistance of the A circuit. From Fig. (b),

The amount of feedback is

which is relatively low given that the open-loop amplifier has Ro = 5000 Ω.
11.5 Other Feedback-Amplifier Types: there are four types of feedback amplifier:
1. the feedback voltage amplifier,
2. the feedback transconductance amplifier,
3. the feedback current amplifier.
4. the feedback transresistance amplifier.
11.5.1 Basic Principles
1. Sensing at the OUTPUT:
i. the feedback voltage and
transresistance amplifiers (Vo is
the output signal of interest): The
feedback network is connected in parallel (or shunt) with the amplifier output node, just as a
voltmeter is connected to measure a voltage.
ii. the feedback transconductance and current amplifiers (Io is the output signal of interest): the
feedback network is connected in series with the output loop, just as a current meter is connected
to measure a current.
2. Mixing at the INPUT:
i. If the input signal to be amplified is a voltage, as in the case of voltage and transconductance
amplifiers, the signal source is represented by its Thevenin equivalent and the feedback voltage
signal Vf is connected in series with the input signal source Vs.
ii. if the signal to be amplified is a current, as in the case of current and transresistance amplifiers,
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.
3. Feedback topology: for each of the four amplifier types there is a uniquely appropriate feedback
topology: The feedback topology:
i. stabilizes the gain of interest
ii. makes the input and output resistances more ideal.
4. Input and output resistance:
• Series connection always increases the resistance by the amount of feedback, (1 + Aβ).
• parallel (shunt) connection always decreases the resistance 1/(1 + Aβ).
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. Analysis using the loop gain: using systematic method
11. 5.2. Feedback Transconductance Amplifier (Series–Series)
𝐼𝑜 𝐴
𝐴𝑓 = =
𝑉𝑠 1 + 𝐴𝛽
𝑅𝑖𝑓 = (1 + 𝐴𝛽)𝑅𝑖 𝑅𝑜𝑓 = (1 + 𝐴𝛽)𝑅𝑜
𝐼𝑜
𝐴=
𝑉𝑖
𝑅𝑖𝑛 = 𝑅𝑖𝑓 − 𝑅𝑠 𝑅𝑜𝑢𝑡 = 𝑅𝑜𝑓 − 𝑅𝐿
𝑉𝑓
𝛽= ⌋
𝐼𝑜 𝐼
1 =0

Example 11.7
Figure (a) shows a feedback amplifier
composed of
1. A differential amplifier A1 with an
• Rid,
• open-circuit voltage gain A1,
• Ro1,
connected in cascade with a common
source MOSFET Q2 having a
• transconductance gm2 and
• ro2.
Use the feedback-analysis method to
1. Determine the type of the feedback amp. And identify the basic amplifier
2. Determine the feedback circuit and verify that the feedback is negative.
3. find an approximate value for 𝐴𝑓

4. determine the closed-loop Af.

5. Rin, and Rout. (Rout is seen between the terminals of RL,
looking back into the output loop).
Solution (follow these steps)
1. It is a feedback transconductance amplifier (series -series)
The basic amplifier consists of
the differential amplifier A1 cascaded with the CS PMOS transistor Q2. The output current Io is
sensed by the series resistance RF.
2. RF is the feedback cct. (Fig. b). It develops a voltage Vf that is mixed in series with the input
loop.
verify that the feedback is negative.
𝐴𝑠 𝑉𝑠 ↑ → 𝑉𝐴1(−) ↑→ 𝑉𝐺2 ↓→ 𝑉𝑆𝐺 ↑→ 𝐼𝑜 ↑→ 𝑉𝑅𝐹 ↑→ 𝑉𝑅𝐹 𝑎𝑛𝑑 𝑉𝑠 𝑜𝑓 𝑠𝑎𝑚𝑒 𝑝𝑜𝑙𝑎𝑟𝑖𝑡𝑦

→ 𝑓𝑒𝑒𝑑𝑏𝑎𝑐𝑘 𝑖𝑠 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒
3. the approximate value of 𝐴𝑓 ≡ Io/Vs for Aβ >> 1. This value is determined first to help us at
the end whether our analysis is correct:
If 𝐴𝛽 ≫ 1 then the final 𝐴𝑓 should be close to the value
initially determined.
From the circuit of Fig. 11.20(d),
𝛽 = 𝑅𝐹
𝐼𝑜 1 1
𝑊ℎ𝑒𝑛 𝐴𝛽 ≫ 1 → 𝐴𝑓 = ≈ =
𝑉𝑠 𝛽 𝑅𝐹
4. determine the A circuit.

Since the feedback network in (Fig. b) is connected in series with both the input and output loops,
Including RF in each of these loops.
Rs and RL in the A circuit, and replacing A1 and Q2 with their small-signal models as in Fig. (c).
𝑅𝑖𝑑
𝑉𝑖𝑑 = −𝑉𝑖
𝑅𝑖𝑑 + 𝑅𝑠 + 𝑅𝐹
𝑉𝑔𝑠2 = 𝐴𝑉𝑖𝑑
𝑟𝑜2
𝐼𝑜 = −𝑔𝑚2 𝑉𝑔𝑠2
𝑟𝑜2 + 𝑅𝐿 + 𝑅𝐹
𝐼𝑜 𝑅𝑖𝑑 𝑟𝑜2
𝐴= = (𝐴1 𝑔𝑚2 ) ( )( )
𝑉𝑖 𝑅𝑖𝑑 + 𝑅𝑠 + 𝑅𝐹 𝑟𝑜2 + 𝑅𝐿 + 𝑅𝐹
Usually Rid >> (Rs + RF), ro2 >> (RL + RF), resulting in the approximate expression for A:
𝐴 ≈ (𝐴1 𝑔𝑚2 )
𝑅𝑖 = 𝑅𝑠 + 𝑅𝑖𝑑 + 𝑅𝐹
The output resistance Ro is found by setting Vi = 0, and breaking the output loop at any location,
say between D2 and 𝐷2′ . Thus,
𝑅𝑜 = 𝑟𝑜2 + 𝑅𝐿 + 𝑅𝐹
The loop gain
𝑅𝑖𝑑 𝑟𝑜2
𝐴𝛽 = (𝐴1 𝑔𝑚2 𝑅𝐹 ) ( )( )
𝑅𝑖𝑑 + 𝑅𝑠 + 𝑅𝐹 𝑟𝑜2 + 𝑅𝐿 + 𝑅𝐹
𝐴𝛽 ≈ 𝐴1 𝑔𝑚2 𝑅𝐹
We next determine the closed-loop gain
𝐴
𝐴𝑓 =
1 + 𝐴𝛽
𝐴1 𝑔𝑚2
𝐴𝑓 ≈
1 + 𝐴1 𝑔𝑚2 𝑅𝐹
For A1gm2RF >> 1,
1
𝐴𝑓 =
𝑅𝐹
The series mixing raises the input resistance with feedback,
𝑅𝑖𝑓 = 𝑅𝑖 (1 + 𝐴𝛽)

𝑅𝑖𝑛 = 𝑅𝑠 − 𝑅𝑖 (1 + 𝐴𝛽)
To obtain 𝑅𝑜𝑓 , we note that the series connection at the output raises the output resistance, thus,

𝑅𝑜𝑓 = 𝑅𝑜 (1 + 𝐴𝛽)

𝑅𝑜𝑢𝑡 = 𝑅𝐿 − 𝑅𝑜 (1 + 𝐴𝛽)
Example 11.8
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. (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.
- the feedback network
samples the emitter
current Io of Q3, and
- Io is the output 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 in Fig. (a). with
- Ic as the output current,
- the output resistance is Rout
- 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 (not shown), establishes IC1 = 0.6 mA, IC2 = 1 mA, and IC3 = 4 mA.
Also assume that for all three transistors, hfe = 100 and ro = ∞.
To avoid confusion of the BJT current gain β and the feedback factor β, we use hfe = β.
(a) Anticipating that the loop gain will be large, find an approximate expression and value for the
closed-loop gain Af ≡ Io/Vs and hence for Ic/Vs. Also find Vo/Vs.
(b) Use feedback analysis to find A, β, Af , Vo/Vs, Rin, and Rout . For the calculation of Rout,
assume that ro of Q3 is 25 kΩ. given gm2 = 40 mA/V, RC2 = 5 kΩ, hfe = 100, re3 = 25/4 = 6.25
Ω, RE2 = 100 Ω, RF = 640 Ω, and RE1 = 100 Ω
Solution
(a) When Aβ >>1,
𝐼𝑜 1
𝐴𝑓 = ≈
𝑉𝑠 𝛽
where the feedback factor β can be found from
the feedback network. The feedback network is
highlighted in Fig. (a), and the determination of
the value of β is illustrated in Fig. (b),
𝑉𝑓 𝑅𝐸2
𝛽= = × 𝑅𝐸1
𝐼𝑜 𝑅𝐸2 + 𝑅𝐹 + 𝑅𝐸1
100
𝛽= × 100 = 11.9 Ω
100 + 640 + 100
1 1
𝐴𝑓 ≈ = = 84 𝑚𝐴/𝑉
𝛽 11.9
𝐼𝑐 𝐼𝑜
≈ = 84 𝑚𝐴/𝑉
𝑉𝑠 𝑉𝑠
𝑉𝑜 −𝐼𝑐 𝑅𝑐3
= = −84 × 0.6
𝑉𝑠 𝑉𝑠
= −50.4 𝑉/𝑉
(b) Using the rules to get the A
circuit shown in Fig. (c).
To find A ≡ Io/Vi we
First: determine the gain of
the first stage.
𝑉𝑐1 −𝛼1 (𝑅𝑐1 ||𝑟𝜋2 )
= = −14.92
𝑉𝑖 𝑟𝑒1 + [𝑅𝐸1 ||(𝑅𝐹 + 𝑅𝐸2 )]
Second: determine the gain of the second stage, (noting that 𝑉𝑏2 = 𝑉𝑐1 )
𝑉𝑐2
= −𝑔𝑚2 {𝑅𝑐2 ||(ℎ𝑓𝑒 + 1)[𝑟𝑒3 + (𝑅𝐸2 ||(𝑅𝐹 + 𝑅𝐸1 ))]} = −131.2
𝑉𝑐1
Third: for the third stage we can write by inspection
𝐼𝑜 𝐼𝑒3 1
= = = 10.6 𝑚𝐴/𝑉
𝑉𝑐2 𝑉𝑏3 𝑟𝑒3 + [𝑅𝐸2 ||(𝑅𝐹 + 𝑅𝐸1 )]
Combining the gains of the three stages results in
𝐼𝑜
𝐴= = −14.92 × −131.2 × 10.6 × 10−3 = 20.7 𝐴/𝑉
𝑉𝑖
𝐼𝑜 𝐴 20.7
𝐴𝑓 = = = = 83.7 𝑚𝐴/𝑉
𝑉𝑠 1 + 𝐴𝛽 1 + 20.7 + 11.9
which we note is very close to the ideal value found in (a) above. This is not surprising, since the
loop gain Aβ = 20.7×11.9 = 246.3 is large (>> 1). The voltage gain is found from
𝑉𝑜 −𝐼𝑐 𝑅𝑐3 −𝐼𝑜 𝑅𝑐3 𝑉𝑜
= ≈ = = −𝐴𝑓 𝑅𝑐3 = 83.7 × 10−3 × 600 = −50.2
𝑉𝑠 𝑉𝑠 𝑉𝑠 𝑉𝑠
which is also very close to the approximate value found in (a) above.
Fourth: Find 𝑅𝑖𝑛 , 𝑅𝑖𝑓 , 𝑅𝑜𝑓 , 𝑅𝑜𝑢𝑡

𝑅𝑖𝑛 = 𝑅𝑖𝑓 = 𝑅𝑖 (1 + 𝐴𝛽)

𝑅𝑖 is the input resistance of the A circuit. The value of Ri can be found from Fig. (c) as follows:

𝑅𝑖 = (ℎ𝑓𝑒 + 1)[𝑟𝑒1 + (𝑅𝐸1 ||(𝑅𝐹 + 𝑅𝐸2 ))] = 13.11 𝑘Ω

𝑅𝑖𝑓 = 𝑅𝑖 (1 + 𝐴𝛽) = 3.24 𝑀Ω

To determine Rout, which is the resistance looking into the collector of Q3, since Id = Is.
Since the feedback senses Io, 𝑅𝑜𝑓 is the resistance seen in the emitter circuit between Y and Y’,

𝑅𝑜𝑓 = 𝑅𝑜 (1 + 𝐴𝛽)

Ro can be determined from the A circuit in Fig. (c) by breaking the circuit between Y and Y’.
𝑅𝑐2
𝑅𝑜 = [𝑅𝐸2 ||(𝑅𝐹 + 𝑅𝐸1 )] + 𝑟𝑒3 + = 143.9 Ω
ℎ𝑓𝑒 + 1

The output resistance Rof of the feedback amplifier is:

𝑅𝑜𝑓 = 𝑅𝑜 (1 + 𝐴𝛽) = 35.6 𝑘Ω

Fifth: Find 𝑅𝑜𝑢𝑡 from the equivalent circuit in Fig. (d) by assuming that
the effect of the feedback is to place 𝑅𝑜𝑓 in the emitter of Q3.
𝑟𝜋3
𝑅𝑜𝑢𝑡 = 𝑟𝑒3 + [𝑅𝑜𝑓 ||(𝑟𝜋3 + 𝑅𝐶2 )] [1 + 𝑔𝑚3 𝑟𝑜3 ] = 2.19 𝑀Ω
𝑟𝜋3 + 𝑅𝐶2
Thus Rout is increased (from ro3) but not by (1+Aβ).
11.5.3 The Feedback Transresistance Amplifier (Shunt–Shunt)
𝑉𝑜 𝐴
𝐴𝑓 = =
𝐼𝑠 1 + 𝐴𝛽
𝑅𝑖𝑓 = 𝑅𝑖 /(1 + 𝐴𝛽)

𝑅𝑜𝑓 = 𝑅𝑜 /(1 + 𝐴𝛽)

𝐼𝑓
𝛽= ⌋
𝑉𝑜 𝑉 =0
1

1
𝑅𝑖𝑛 =
1 1
−
𝑅𝑖𝑓 𝑅𝑠
1
𝑅𝑜𝑢𝑡 =
1 1
−
𝑅𝑜𝑓 𝑅𝐿

Example 11.9
Fig. (a) shows a feedback
transresistance amplifier.
It is formed by connecting:
RF in the negative-feedback
path of a
voltage amplifier: gain μ, Rid
and ro.
(a) If the loop gain is large,
find an approximate
expression for the closed-loop transresistance Vo/Is of the feedback amplifier.
(b) Find the A circuit and expressions for A, Ri, and Ro.
(c) Find expressions for the loop gain, Af , Ri f , Rin, Rof , and Rout .
(d) Find Ri, Ro, A, β, Aβ, Af , Ri f , Rin, Rof , and Rout for the case μ = 104 V/V, Rid =∞, ro =
100 Ω, RF = 10 k Ω, and Rs = RL = 1 k Ω.
(e) If instead of a current source Is having a source resistance Rs = 1 k Ω, the amplifier is fed
from a voltage source Vs having a source resistance Rs = 1 k Ω, find the voltage gain Vo/Vs.
Solution
(a) If the loop gain 𝐴𝛽 ≫ 1,
𝑉𝑜 1
𝑊ℎ𝑒𝑛 𝐴𝛽 ≫ 1 → 𝐴𝑓 = ≈
𝐼𝑠 𝛽
from the β circuit in Fig. (b), β is found from Fig. (c)
𝐼𝑓 1
𝛽= =−
𝑉𝑜 𝑅𝐹
𝑉𝑜
≈ −𝑅𝐹
𝐼𝑠
Note that in this case the voltage at the input node (the
inverting input terminal of μ) will be very close to
ground and thus very little, if any, current flows into the
input terminal of the amplifier. Nearly all of Is will flow
through RF, resulting in 𝑉𝑜 = 0 − 𝐼𝑠 𝑅𝐹 = −𝐼𝑠 𝑅𝐹 .
(b) From the feedback network in Fig. (b), we see that
the loading effect at the amplifier input and output will simply be RF . This is indicated in the A
circuit shown in Fig. (c), where we have replaced the amplifier μ with its equivalent-circuit model.
The open-loop transresistance A is
𝑉𝑖𝑑 = 𝐼𝑖 𝑅𝑖
𝑅𝑖 = 𝑅𝑖𝑑 ||𝑅𝐹 ||𝑅𝑠
𝑅𝐹 ||𝑅𝐿
𝑉𝑜 = −𝜇𝑉𝑖𝑑
𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
𝑉𝑜 𝑅𝐹 ||𝑅𝐿
𝐴= = −𝜇𝑅𝑖
𝐼𝑖 𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
The open-loop Ro can be
obtained by the A circuit with Ii set to 0. Then Vid = 0, and
𝑅𝑜 = 𝑟𝑜 ||𝑅𝐹 ||𝑅𝐿
(c) The loop gain Aβ can be obtained by combining
𝑅𝑖 𝑅𝐹 ||𝑅𝐿
𝐴𝛽 = 𝜇 ( )
𝑅𝐹 𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
Observe that although
a. both A and β are negative, Aβ is positive, confirming that the feedback is negative.
b. Also note that Aβ is dimensionless, as it must always be.
The closed-loop gain Af can now be found as
𝑉𝑜 𝐴
𝐴𝑓 = =
𝐼𝑠 1 + 𝐴𝛽
𝑅𝐹 ||𝑅𝐿
−𝜇𝑅𝑖
𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
𝐴𝑓 =
𝑅 𝑅𝐹 ||𝑅𝐿
1 + 𝜇( 𝑖)
𝑅𝐹 𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
𝑅𝑖 𝑅𝐹 ||𝑅𝐿
𝜇( ) ≫ 1 → 𝐴𝑓 ≈ −𝑅𝐹
𝑅𝐹 𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
The input resistance with feedback, Rif , is obtained by dividing Ri by (1+Aβ) with the result
𝑅𝑖
𝑅𝑖𝑓 =
1 + 𝐴𝛽
1 1 𝐴𝛽 1 𝜇 𝑅𝐹 ||𝑅𝐿
= + = +
𝑅𝑖𝑓 𝑅𝑖 𝑅𝑖 𝑅𝑖 𝑅𝐹 𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )

Substituting for Ri and let

𝜇(𝑅𝐹 ||𝑅𝐿 )
𝜇′ =
𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
𝑅𝐹
𝑅𝑖𝑓 = 𝑅𝑖𝑑 ||𝑅𝐹 ||𝑅𝑠 ||
𝜇′
𝑅𝐹
𝑅𝑖𝑓 = 𝑅𝑠 ||𝑅𝑖𝑑 ||𝑅𝑠 ||
𝜇′ + 1
Since Rif = Rs || Rin, we see that
𝑅𝐹
𝑅𝑖𝑛 = 𝑅𝑖𝑑 ||
𝜇′ + 1
Rid is usually large
𝑅𝐹 𝑅𝐹
𝑅𝑖𝑛 ≈ ≈
𝜇′ + 1 𝜇′
for large amplifier gain μ, 𝑅𝑖𝑛 will be low. Rof can be found by dividing Ro by (1+Aβ):
𝑅𝑜
𝑅𝑜𝑓 =
1 + 𝐴𝛽
 1 1 𝐴𝛽 1 𝑅𝑖 𝑅𝐹 ||𝑅𝐿 1
= + = +𝜇
𝑅𝑜𝑓 𝑅𝑜 𝑅𝑜 𝑅𝑜 𝑅𝐹 𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 ) 𝑅𝑜
1 1 1 1 𝑅𝑖 1 1 1 1 𝑅𝑖
= + + +𝜇 = + + (1 + 𝜇 )
𝑅𝑜𝑓 𝑅𝐿 𝑅𝐹 𝑟𝑜 𝑅𝐹 𝑟𝑜 𝑅𝐿 𝑅𝐹 𝑟𝑜 𝑅𝐹
𝑟𝑜
𝑅𝑜𝑓 = 𝑅𝐿 ||𝑅𝐹 ||
𝑅𝑖
1+𝜇
𝑅𝐹
𝑅𝑜𝑓 = 𝑅𝐿 ||𝑅𝑜𝑢𝑡
𝑟𝑜
𝑅𝑜𝑢𝑡 = 𝑅𝐹 ||
𝑅𝑖
1+𝜇
𝑅𝐹
𝑟𝑜
𝑢𝑠𝑢𝑎𝑙𝑙𝑦 𝑅𝐹 || ≫1
𝑅
1+𝜇 𝑖
𝑅𝐹
𝑟𝑜 𝑅𝐹 𝑟𝑜
𝑅𝑜𝑢𝑡 ≈ ≈( )( ) 𝑖𝑓 𝑡ℎ𝑒 𝑔𝑎𝑖𝑛 𝜇 𝑖𝑠 𝑙𝑎𝑟𝑔𝑒 𝑅𝑜𝑢𝑡 𝑖𝑠 𝑠𝑚𝑎𝑙𝑙
𝑅 𝑅𝑖 𝜇
1+𝜇 𝑖
𝑅𝐹
(d) For the numerical values given:
𝑅𝑖 = 𝑅𝑖𝑑 ||𝑅𝐹 ||𝑅𝑠 = ∞||10||1 = 0.91 𝑘Ω

𝑅𝑜 = 𝑟𝑜 ||𝑅𝐹 ||𝑅𝐿 = 0.1||10||1 = 90 Ω

𝑅𝐹 ||𝑅𝐿
𝐴 = −𝜇𝑅𝑖 = −9198 𝑘Ω
𝑟𝑜 + (𝑅𝐹 ||𝑅𝐿 )
1 1
𝛽=− =− = −0.1 𝑚𝐴/𝑉
𝑅𝐹 10
𝐴𝛽 = 819.8
1 + 𝐴𝛽 = 820.8
𝐴 8198
𝐴𝑓 = =− = −9.99 𝑘Ω
1 + 𝐴𝛽 820.8
Which is very close to the ideal value -RF = -10 kΩ
𝑅𝑖 910
𝑅𝑖𝑓 = = = 1.11 Ω
1 + 𝐴𝛽 820.8
 1 1
𝑅𝑖𝑛 = = = 1.11 Ω
1 1 1 1
− −
𝑅𝑖𝑓 𝑅𝑖 1.11 1000

Which is very low and highly desirable property

𝑅𝑜 90
𝑅𝑜𝑓 = = = 0.11 Ω
1 + 𝐴𝛽 820.8
1 1
𝑅𝑜𝑢𝑡 = = = 0.11 Ω
1 1 1 1
− −
𝑅𝑜𝑓 𝑅𝐿 0.11 1000

which as well is very low, another highly desirable property.

(e) If the amplifier is fed with a voltage source Vs having a resistance Rs = 1 kΩ, the output
voltage can be found from
𝑉𝑠
𝑉𝑜 = 𝐴𝑓 𝐼𝑠 = 𝐴𝑓
𝑅𝑠
𝑉𝑜 𝐴𝑓 9.99 𝑘
= = = −9.99 𝑉/𝑉
𝑉𝑠 𝑅𝑠 1𝑘

11.5.4 The Feedback Current Amplifier (Shunt–Series)

𝐼𝑜 𝐴
𝐴𝑓 = =
𝐼𝑠 1 + 𝐴𝛽
𝑅𝑖𝑓 = 𝑅𝑖 /(1 + 𝐴𝛽)

𝑅𝑜𝑓 = (1 + 𝐴𝛽)𝑅𝑜
𝐼𝑓
𝛽= ⌋
𝐼𝑜 𝑉 =0
1

1
𝑅𝑖𝑛 =
1 1
−
𝑅𝑖𝑓 𝑅𝑠

𝑅𝑜𝑢𝑡 = 𝑅𝑜𝑓 − 𝑅𝐿
Example 11.10
Fig. (a) 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, senses 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 DC bias
is not shown.
The amplifier μ can be
implemented in:
a. an op amp, a
differential amplifier,
or a single-ended
inverting amplifier.
The simplest approach
is to implement μ with
a CS MOSFET
amplifier.
Assume that the
amplifier μ has an
input resistance Rid , an open-circuit voltage gain μ, and an output resistance ro1.
(a) If the loop gain is large, find an approximate expression for the closed-loop gain Af ≡ Io/Is.
(b) Find the A circuit and derive expressions for A, Ri, and Ro.
(c) Give expressions for Aβ, Af , Ri f , Rin, Rof , and Rout .
(d) Find numerical values for A, β, Aβ, Af , Ri, Ri f , Rin, Ro, Rof , and Rout for the following
case: μ = 1000 V/V, Rs = ∞, Rid = ∞, ro1 = 1 kΩ, R1 = 10 kΩ, R2 = 90 kΩ; for Q: gm = 5 mA/V
and ro = 20 kΩ.
Solution
(a) When the loop gain Aβ >>1, Af ≈ 1/β.
The β circuit in Fig. (b), is used to determine β as in Fig. (c),
𝐼𝑓 𝑅1
𝛽= =−
𝐼𝑜 𝑅1 + 𝑅2
 1 𝑅1
𝐴𝑓 ≈ = −( )
𝛽 𝑅1 + 𝑅2
refer to Fig. (d).
Assumed Aβ >>1, Ii ≈ 0 and thus If ≈ Is.
because 𝐼𝑖 ≈ 0, 𝑉𝑖 ≈ 0 .
the voltage at the source of Q is
𝑉𝑆𝑄 = −𝐼𝑓 𝑅2 ≈ −𝐼𝑠 𝑅2 .
𝑅2
𝐼𝑅1 = 𝐼𝑠
𝑅1
The source current of Q will be
𝐼𝑜 = −(𝐼𝑠 + 𝐼𝑠 𝑅2 /𝑅1 )

𝑅2
𝐼𝑜 = −𝐼𝑠 (1 + )
𝑅1
which confirms the expression for Af
(b) To obtain the A circuit we load the input side of the basic amplifier with Rs and R11.
R11 = R1 + R2 (because port 2 of the feedback network is opened).
We also load the output of the basic amplifier with R22, which in this case is
R22 = R1 || R2 (because port 1 of the feedback network is shorted).
The resulting
A circuit is
shown in Fig.
(e), where we
have replaced
the amplifier
μ with its
equivalent
circuit.
Analysis of the A circuit is straightforward and proceeds as follows:

(c) The loop gain:

The closed-loop gain Af can be obtained by substituting A and Aβ in Af = A/(1+Aβ).

The input resistance Ri f is found as

Since by definition,

we can easily find Rin.

For the output resistance, we have

(d) For the numerical values given,

11.6 Summary of the Feedback-Analysis Method
Table 11.2 provides a summary of the rules and relationships employed in the analysis and design
of the four types of feedback amplifier. In addition to the following important analysis tips.
1. Always begin the analysis by determining an approximate value for the closed-loop gain Af ,
assuming that the loop gain Aβ is large and thus Af ≈ 1/β.
This value should serve as a check on the final value you find for Af . How close the actual Af is
to this ideal value will depend on how large Aβ is compared to unity.
2. The shunt connection at input or output always results in reducing the corresponding resistance
(input or output).
The series connection at input or output always results in increasing the corresponding resistance
(input or output).
3. In utilizing negative feedback to improve the properties of an amplifier under design, the
starting point in the design is
a. the selection of the feedback topology appropriate for the application at hand.
b. Then the required amount of negative feedback (1+Aβ) can be ascertained utilizing the fact that
it is this quantity that determines the magnitude of improvement in the various amplifier
parameters.
c. Also, the feedback factor β can be determined from the closed-loop gain Af , β ≈ 1/ Af