Chapter 12 Feedback
¾ 12.1 General Considerations
¾ 12.2 Properties of Native Feedback
¾ 12.3 Types of Amplifiers
¾ 12.4 Sense and Return Techniques
¾ 12.5 Polarity of Feedback
¾ 12.6 Feedback Topologies
¾ 12.7 Effect of Finite I/O Impedances
Negative Feedback System
¾ A negative feedback system consists of four components:
¾ 1) feedforward system
¾ 2) sense mechanism
¾ 3) feedback network
¾ 4) comparison mechanism
Close-loop Transfer Function
X
F= KY
11
( )
( )
Y A X X
FA X KY
= −
= −
Close-loop Transfer Function
1 1
1 KA A X
Y
= +
1
1 1
( )
Y A X KY Y A KY A X
= −
+ =
Example 12.1: Feedback
¾ A1 is the feedforward network,
R1 and R2 provide the sensing and feedback capabilities, 1 2
1
2 1
1 A
R R
R A X
Y
+ +
=
2
1 2
K R
R R
= +
K
Comparison Error
¾ As A1K increases, the error between the input and feedback signal decreases. Or the feedback signal approaches a
good replica of the input.
0
E ≈
(∵ A K1 1)X ≈ X
F (∵E = X − XF)1 1
E X X F
X
A K
= −
= +
X
F= KY
1
1 1
Y A X
= KA
∵ +
1
1
1 FX KA X
= KA
+
Comparison Error
2
1 1
R R X
Y ≈ +
2
1 2
F
Y R X
R R =
+ ∵ X ≈ X
F= K
1
= K
Loop Gain
¾ When the input is grounded, and the loop is broken at an arbitrary location, the loop gain is measured to be KA1.
test N
V KA
1= − V
= 0
X
Example 12.3: Alternative Loop Gain Measurement
test
N KA V
V = − 1
= 0
X
Incorrect Calculation of Loop Gain
¾ Signal naturally flows from the input to the output of a feedforward/feedback system. If we apply the input the
other way around, the “output” signal we get is not a result of the loop gain, but due to poor isolation.
Gain Desensitization
¾ A large loop gain is needed to create a precise gain, one that does not depend on A1, which can vary by ±20%.
1
K >> 1
A X K
Y 1
≈
F
Y Y
X X
∴ ≈
0 E ≈
if 1 A K
1X X
F∴ ≈ 1
1E X X
FX A K
= −
= +
X
F= KY
ex: gmRD m ox ( GS TH)
g C W V V
μ L
= −
Ratio of Resistors
¾ When two resistors are composed of the same unit resistor, their ratio is very accurate. Since when they vary, they will vary together and maintain a constant ratio.
2
1
1R R X
Y ≈ +
Example 12.4: Gain Desensitization
1 4 K =
Nominal gain
1
1
1A Y
X = A K +
3.984 Y
X = Y 3.968
X =
( A
1= 1000) ( A
1= 500)
¾ Determine the actual gain if A1=1000. Determine the percentage change in the gain if A1 drops to 500.
Merits of Negative Feedback
¾ 1) Bandwidth enhancement
¾ 2) Modification of I/O Impedances
¾ 3) Linearization
1 1
1 KA A X
Y
= +
Bandwidth Enhancement
¾ Although negative feedback lowers the gain by (1+KA0), it also extends the bandwidth by the same amount.
( )
01
0
1 A s A
s
ω
=
+ ( )
(
0)
00 0
1 1
1
ω KA
s KA A X s
Y
+ +
= +
Open Loop Closed Loop
Negative Feedback
1
1 1
A Y
X = KA
∵ +
0
: -3 dB BW
ω
00
0 0 0
0 0
0 0 0
0
1 1
1 1
1 (1 )
1 A
s A
A KA
Y
A s s
X K KA KA
s ω
ω ω
ω
+ +
= = =
+ + +
+ +
+
∵
gain
bandwidth
Bandwidth Extension Example
¾ As the loop gain increases, we can see the decrease of the overall gain and the extension of the bandwidth.
Example 12.6: Unity-gain bandwidth
¾ We can see the unity-gain bandwidth remains independent of K, if KA0 >>1 and K2<<1
0 dB
0
0 2
2 2
0 0
( ) 1
1 (1 )
A
KA Y j
X
KA
ω ω
ω
= +
+ +
0
0 2
2 2
0 0
1 1
1 (1 )
u
A
KA KA
ω ω
= +
+ +
2 2
0 0 0
2 2 2
0 0 0
0 0
(1 )
u A KA
A K A
A
ω ω
ω ω
= − +
≈ −
≈
1<< KA0
2 1
K <<
Example12.7: Open Loop Parameters
D out
m in
D m
R R
R g
R g
A
=
=
≈
1
0
0
?
?
?
in out
A R R
=
=
=
Example12.7: Closed Loop Voltage Gain
1 2
2
Assuming ,
1
D
out m D
in
R R R
v g R
v R
g R + >>
= +
( )
D m G S
i = g v − v
i
D 01 2
because of
m D
D
A g R
R R R
≈
+ >>
0 0
1
out in
v A
v = KA
+
Example12.7: Closed Loop I/O Impedance
⎟⎟ ⎠
⎜⎜ ⎞
⎝
⎛ + +
=
m Dm
in
g R
R R
R R g
2 1
1
21
D m D
out
R R g
R R R R
2 1
1
2+ +
=
i
D2
1 2
( i R
D D) R R R
−
+
D D
≈ i R
+
−
2
1 2
X D
D m GS m X
i R R
i g v g v
R R
⎛ ⎞
= = ⎜ ⎝ + − ⎟ ⎠
D X
i = − i
2
1 2
v R
XR + R i
DD m GS
i = g v
X
D
v
R
X D X
/
Di ≈ + i v R
Example: Load Desensitization
3
D
/
m D
m
R g R
g →
D m D m
D m D m
R R g
R R
R g
R R g
R R
R g
2
2
3
1 + +
→ + +
W/O Feedback Large Difference
With Feedback Small Difference
2
1 2
( / 3)
1 ( / 3)
m D
m D
g R
R g R R R
+ +
Before feedback
Linearization
Before feedback
After feedback
1 1
1
Gain at
1 x A
= KA +
( )
1
1 1 1 1
1 / 1 1 1 1 1 1
1 1
1 / 1 1 1 /
K
KA K KA K KA K KA
⎛ ⎞− ⎛ ⎞
= = ⎜ + ⎟ ≈ ⎜ − ⎟
+ + ⎝ ⎠ ⎝ ⎠
2 2
1 1
Gain at 1
1 x A
KA K KA
⎛ ⎞
= + ≈ ⎜⎝ − ⎟⎠
Four Types of Amplifiers
Current meter Voltage meter
Voltage source
Ideal Models of the Four Amplifier Types
(a) Voltage amplifier
(c) Transconductance amplifier (d) Current amplifier (b) Transresistance amplifier infinite
input impedance
zero output
impedance
R
in= 0 R
out= 0
R
out= ∞ R
out= ∞
R
in= ∞
in
0
R =
Realistic Models of the Four Amplifier Types
Examples of the Four Amplifier Types
Sensing a Voltage
¾ In order to sense a voltage across two terminals, a voltmeter with ideally infinite impedance is used.
Sensing and Returning a Voltage
¾ Similarly, for a feedback network to correctly sense the output voltage, its input impedance needs to be large.
¾ R1 and R2 also provide a means to return the voltage.
∞
≈ +
21
R
R
Feedback Network
Sensing a Current
¾ A current is measured by inserting a current meter with ideally zero impedance in series with the conduction path.
¾ The current meter is composed of a small resistance r in parallel with a voltmeter.
Sensing and Returning a Current
¾ Similarly for a feedback network to correctly sense the current, its input impedance has to be small.
¾ RS has to be small so that its voltage drop will not change Iout.
≈ 0 R
SFeedback Network
Addition of Two Voltage Sources
¾ In order to add or substrate two voltage sources, we place them in series. So the feedback network is placed in series with the input source.
Feedback Network
e in F
v = v − v
Practical Circuits to Subtract Two Voltage Sources
¾ Although not directly in series, Vin and VF are being
subtracted since the resultant currents, differential and
single-ended, are proportional to the difference of Vin and VF.
Addition of Two Current Sources
¾ In order to add two current sources, we place them in
parallel. So the feedback network is placed in parallel with Feedback
Network
Practical Circuits to Subtract Two Current Sources
¾ Since M1 and RF are in parallel with the input current source, their respective currents are being subtracted. Note, RF has to be large enough to approximate a current source.
e in F
i = − i i
Example 12.10: Sense and Return
¾ R1 and R2 sense and serve as the feedback network.
¾ M1 and M2 are part of the op-amp and also act as a voltage subtractor.
1
I
D↑
2
I
D↓
v
out↑
v
F↑
2
I
D⇑
Voltage Return Voltage Sense
Example 12.11: Feedback Factor
mF
F g
v
K = i =
Voltage Sense
Current Return
v
X↑ X
1
v
out↓
2
v
out↑
F X
i ↑⇒ v ⇓
Topics in Last and Today’s Lectures
¾ 12.3 Types of Amplifiers – Simple Amplifier Models
– Examples of Amplifier Types
¾ 12.4 Sense and Return Techniques
¾ 12.5 Polarity of Feedback
¾ 12.5 Feedback Topologies – Voltage-Voltage Feedback – Voltage-Current Feedback – Current-Voltage Feedback – Current-Current Feedback
Input Impedance of an Ideal Feedback Network
¾ To sense a voltage, the input impedance of an ideal feedback network must be infinite.
¾ To sense a current, the input impedance of an ideal feedback network must be zero.
Voltmeter Amperemeter
Output Impedance of an Ideal Feedback Network
¾ To return a voltage, the output impedance of an ideal feedback network must be zero.
¾ To return a current, the output impedance of an ideal feedback network must be infinite.
Voltage source Current source
¾ 1) Assume the input goes either up or down.
¾ 2) Follow the signal through the loop.
¾ 3) Determine whether the returned quantity enhances or opposes the original change.
Determining the Polarity of Feedback
Negative feedback Positive feedback
Example 12.12: Polarity of Feedback
in
↑
V I
D1↑ , I
D2↓ V
out↑ , V
x↑ I
D2↑ , I
D1↓
Example 12.13: Polarity of Feedback
in
↑
V I
D1↑ , V
A↓ V
out↑ , V
x↑ I
D1↓ , V
A↑
Example 12.14: Polarity of Feedback
in
↑
I I
D1↑ , V
X↑ V
out↓ , I
D2↓ I
D1↑ , V
X↑
Voltage-Voltage Feedback
0 0
1 KA A V
V
in out
= +
1 in F
V = V − V
A V
0 1=
F out
V = KV
0
( )
out in out
V A V KV
∴ = −
Example 12.15: Voltage-Voltage Feedback
1 2
2
A ssu m ing ( ),
( )
1 ( )
O N O P
out m N O N O P
in
R R r r
V g r r
V R
g r r
+ >>
=
+
( )
1 2
out
mN ON OP
in in
v g r r
v v ≈
∵ −
A
0Input Impedance of a V-V Feedback
) 1
( A
0K I R
V
in in
in
= +
¾ The impedance modification brings the circuit closer to an ideal voltage amplifier
( )
0in in in F
in in in
I R V V
V I R A K
= −
= −
Example12.16: V-V Feedback Input Impedance
1 2
2
Assuming , 1 1
D in
m D
R R R
V R
I g R R g R + >>
⎛ ⎞
= ⎜ ⎝ + + ⎟ ⎠
) 1
( A0K I R
V
in in
in = +
m
1
= g
Output Impedance of a V-V Feedback
( 1 R KA
0)
I
V
outX X
= +
¾ A better voltage source
( )
X 0 X
X
out
V KA V
I R
= − −
KVX
KVX
−
0 X
−KA V
×
Example 12.17: V-V Feedback Output Impedance
1 2
,
2 1 2
1
A ssum ing ( ),
1 / ( ) ( )
1 1
O N O P
O N O P
out closed
m N O N O P
R R r r
r r
R R R R g r r
R
+ >>
= + + ⋅ ⋅
⎛ ⎞
≈ ⎜ + ⎟ (
out)
X
X 0
R V
I = 1 KA
∵ +
Voltage-Current Feedback
1
out O
V R
I = KR +
in F
I I
= −
0
I R
e=
( )
0out in out
V I KV R
∴ = −
KV
out=
Example 12.18: Voltage-Current Feedback
out in
Assuming is very large, open loop gain (V /I ):
( )
R
FR = R − g ⋅ R
1 in D
i R
( )
0 1 2 2
out
D m D
in
V R R g R
i = = −
Example 12.18: Voltage-Current Feedback
1
Assuming is very large,
1 / 1 /
F
out out
RF
F m F
F
R
V V
I R g R
K R
= ≈
+
= −
∞
1 / g
mI
FF out
I = KV
∵
Example 12.18: Voltage-Current Feedback
2 1 2
2 1 2
1
out m D D
m D D
in closed
V g R R
g R R I
R
= − +
R
0KR
0if
out
F in closed
V R
I
g R R R
≈
>>
Input Impedance of a V-I Feedback
1
0in X
X
R V
I = R K
+
¾ A better current sensor.
Current sensor or ammeter
X in
V
≈ R
X 0in
V R
= R
0 X in
K V R
= R
0
X X
X
in in
V V
I K R
R R
− =
By KCL
Example 12.19: V-I Feedback Input Impedance
D D
m m closed
in
g g R R
R
2 1
1 2 ,
1 . 1 1
+
=
1
0in X
X
R V
I = R K
∵ +
Output Impedance of a V-I Feedback
¾ A better voltage source.
K R
R I
V
outX X
1 +
0=
KV
X= −
Input current source
KV
X=
0
KV R
X= −
×
Large
Input impedance
0
X X
out
V KV R R
= +
Example12.20: V-I Feedback Output Impedance
2 ,
2 1 2
1
D out closed
m D D
R R
g R R
=
+
1
0o u t X
X
V R
I = R K
∵ +
2
if is very larg e
o u t D
F
R R
R
≈
1 in D
i R
( )
0
1 2 2
out in
D m D
V R
i
R g R
=
= −
∵
Current-Voltage Feedback
m m in
out
KG G V
I
= + 1
F
/
outV I
=
( )
m in F
G V V
= −
KI
out=
( )
out m in out
I = G V − KI
Example12.21: Current-Voltage Feedback
( )
3 3 5 1
| ||
out
m open m O O m
G = I = g ⋅ r r ⋅ g
Laser
3
(
3 5)
m o o in
g r r V
−
+
−
[ ]
1 1 3
(
3 5)
D m m o o in
i = g − g r r V
1
out D
I = − i
Example12.21: Current-Voltage Feedback
F
M out
K V R
= I =
Laser
out M
I R
=
Example12.21: Current-Voltage Feedback
( )
( )
1 3 3 5
1 3 3 5
|| 1
| 1 1 ||
m m O O
out m
closed
in m m m O O M M
g g r r
I G
V = K G = g g r r R ≈ R
+ ⋅ +
Laser
Topics in Last and Today’s Lectures
¾ 12.5 Polarity of Feedback
¾ 12.5 Feedback Topologies – Voltage-Voltage Feedback – Voltage-Current Feedback – Current-Voltage Feedback – Current-Current Feedback
¾ 12.6 Effect of Finite I/O Impedances – Inclusion of I/O Effects
¾ 12.7 Stability in Feedback Systems
Input Impedance of a I-V Feedback
¾ A better voltage sensor.
(1 )
in
in m
in
V R K G
I = +
in F
in in
V V I R
= −
=
m in in
G I R
=
F out
m in in
V KI
KG I R
=
=
in in in F
in m m in
I R V V
V KG I R
= −
= −
Output Impedance of a I-V Feedback
¾ A better current source.
) 1
(
mout X
X
R KG
I
V = +
+
−
m X
− KG I
By KCL
X
X m X
out
I V KG I
= R −
F X
V = KI
F X
K V
= I
Laser
Example: Current-Voltage Feedback
) 1
1 (
|
) 1
1 (
|
| 1
2 1
2
2 1
1
2 1
2 1
M D
m m
m closed
out
M D
m m
m closed
in
M D
m m
D m
m closed
in out
R R
g g g
R
R R
g g g
R
R R
g g
R g
g V
I
+
=
+
=
= +
Very small
1 in m D
V g R = g
m2( V g R
in m1 D) for open loop circuit
1 2
m m D m
G g R g
∴ = ⋅
F
M out
K V R
= I =
∵
1 / g
m11 / g
m2out m
I G
= +
∵ V
in in(1
m)
R K G
= +
∵ V
X o u t(1
m)
R K G
= +
∵
Wrong Technique for Measuring Output Impedance
¾ If we want to measure the output impedance of a I-V closed- loop feedback topology directly, we have to place VX in
series with K and Rout. Otherwise, the feedback will be disturbed.
I
XOutput current sensing
Current-Current Feedback
I
out
A
I =
( )
I in F
A I I
= −
F
/
outI I
=
in F
I I
= −
KI
out= I
out= A I
I(
in− KI
out)
Input Impedance of I-I Feedback
¾ A better current sensor.
I in X
X
KA R I
V
= + 1
I X
/
inA V R
=
F I X
/
inI = KA V R
X
X F
in
X X
I
in in
I V I
R
V V
R KA R
= +
= +
Output Impedance of I-I Feedback
¾ A better current source.
) 1
(
Iout X
X
R KA
I
V = +
KI
X=
I X
− KA I
By KCL
X
X I X
out
I V KA I
= R −
High R
Example 12.24: Test of Negative Feedback
in
↑
I V
D1↑ , I
out↓ V
P↓ , I
F↑ V
D1↓ , I
out↑
Laser
Example 12.24: I-I Negative Feedback
2 2
|
21
out m X m D in
I open m D
in in in
F P M
I g V g R I
A g R
I I I
I V R
K I R I R
− −
= = = = −
= = − ⋅ = −
Laser
X in D X
V = I R
2 2
D m X
I = g V
2 2
out m X
m D in
I g V
g R I
= −
= −
P M
V
= R
Very small
P out M
V = I R
Example: I-I Negative Feedback
)]
/ (
1 [
|
) /
( 1
. 1
| 1
) /
(
| 1
2 1
2
2
F M
D m
m closed
in
F M
D m
D m
closed I
R R
R g
r R
R R
R g
R g
R R
R g
R A g
+
=
= + +
= −
Laser
1
out I
in I
I A
I = KA
∵ +
1
in X
X I
V R
I = KA
∵ +
X out(1
I)
X
V R KA
I = +
∵
Summary
¾ Determine whether the returned quantity enhances or opposes the original change.
Negative feedback Positive feedback
Voltage-Voltage Voltage-Current
Current-Voltage Current-Current
0 0
1 KA A V
V
in out
= + R (1 A0K) I
V
in in
in = +
(
1 RKA0)
I
V out
X X
= +
1
out O
in O
V R
I = KR
+ 1 0
X in X
R V
I = R K +
K R R I
V out
X X
1 + 0
=
m m in
out
KG G V
I
= +
1 (1 )
in
in m
in
V R K G
I = +
I I in
out
KA A I
I
= +
1
Iin X
X
KA R I
V
= +
1
How to Break a Loop
¾ The correct way of breaking a loop is such that the loop does not know it has been broken. Therefore, we need to present the feedback network to both the input and the output of the feedforward amplifier.
Rules for Breaking the Loop of Amplifier Types
V-V V-I
I-V I-I
Zout= low
Return Duplicate
Zin= high
Zout= low
Zin= low
Zin= low Zin= high
Zout= high Zout= high
Sense Duplicate
Intuitive Understanding of these Rules
¾ Since ideally, the input of the feedback network sees zero impedance (Zout of an ideal voltage source), the return
replicate needs to be grounded. Similarly, the output of the feedback network sees an infinite impedance (Zin of an ideal voltage sensor), the sense replicate needs to be open.
¾ Similar ideas apply to the other types.
Voltage-Voltage Feedback
Rules for Calculating Feedback Factor
V-V V-I
I-V I-I
Intuitive Understanding of these Rules
¾ Since the feedback senses voltage, the input of the
feedback is a voltage source. Moreover, since the return quantity is also voltage, the output of the feedback is left open (a short means the output is always zero).
¾ Similar ideas apply to the other types.
Voltage-Voltage Feedback
Example 12.26: Breaking the Loop
( )
[ ]
( )
1 ,
2 1
1 ,
/ 1
||
g R
R R
R g
A
m open
in
D m
open v
=
+
=
Example 12.26: Feedback Factor
) 1
/(
) 1
(
) 1
/(
) /(
, ,
,
, ,
,
, ,
,
2 1
2
open v
closed out
closed out
open v
open in
closed in
open v
open v
closed v
KA R
R
KA R
R
KA A
A
R R
R K
+
=
+
=
+
=
+
=
Example 12.27: Breaking the Loop
( )
[ ]
( )
,
2 1
,
|| ||
R
R R
r r
g A
open in
OP ON
mN open
v
+
=
∞
=
+
=
Example 12.27: Feedback Factor
) 1
/(
) 1
/(
) /(
,
, ,
,
2 1
2
closed in
open v
open v
closed v
KA R
R R
KA A
A
R R
R K
+
=
∞
=
+
=
+
=
Example 12.29: Breaking the Loop
( )
[ ]
F m
open in
F D
m
m F
D F
open in
out
g R R
R R
g R g
R R I
V
1 ||
||
1 .
|
1 ,
2 2
1 1
=
− +
=
Example 12.29: Feedback Factor
)
| 1
/(
)
| 1
/(
)
| 1
/(
|
|
/ 1
, ,
out open in
out open
in closed
in
open in
out open
in out closed
in out
F
K V R
R
I K V R
R
I K V I
V I
V
R K
+
=
+
=
+
=
−
=
1 2 F
V = − I R
2 1
K I
= V
Example 12.30: Breaking the Loop
( )
open in
M L
O
O m O
O m
open in
out
R
R R
r
r g
r r
g V
I
+
=
∞
=
+
= +
,
1
1 1 5
3
3
||
|
1
ro
RL
RM
( )
3 3 5
X
m o o
in
V g r r
V = −
1 m X
g V
( )
1 1
1
o m X
out
o L M
r g V I = − r R R
+ +
M 2 1
out in closed out in open out in open
in ,closed
out ,closed out ,open out in open
K R V / I
( I / V | ) ( I / V | ) / [ 1 K ( I / V ) | ] R
R R [ 1 K ( I / V ) | ]
= =
= +
= ∞
= +
Example 12.30: Feedback Factor
Example 12.31: Breaking the Loop
m open
in
m M
L
D m
open in
out
g R
g R
R
R g
V I
=
+
= + / 1
/
| 1
1 ,
2 1
1 X
m D
in
V g R V =
A
1 / 2
A L M
X L M m
V R R
V R R g
= +
+ +
A out
I = V
+
out open ?
R =
RM
VX
+
− IX
Example 12.31: Feedback Factor
]
| ) /
( 1
[
]
| ) /
( 1
[
]
| ) /
( 1
/[
)
| /
( )
| /
(
,
,closed in open out in open
in
open in
out open
in out
closed in
out
M