Chapter 12 Feedback

 12.1 General Considerations

 12.2 Types of Amplifiers

 12.3 Sense and Return Techniques

 12.4 Polarity of Feedback

 12.5 Feedback Topologies

 12.6 Effect of Finite I/O Impedances

 12.7 Stability in Feedback Systems

188
 Negative Feedback System

 A negative feedback system consists of four components:

1) feedforward system, 2) sense mechanism, 3) feedback
network, and 4) comparison mechanism.
CH 12 Feedback 189
 Close-loop Transfer Function

Y A1

X 1  KA1
CH 12 Feedback 190
 Feedback Example

Y A1

X R2
1 A1
R1  R2

 A1 is the feedforward network, R 1 and R2 provide the

sensing and feedback capabilities, and comparison is
provided by differential input of A 1.
CH 12 Feedback 191
 Comparison Error

X
E
1 A1 K

 As A1K increases, the error between the input and fed back
signal decreases. Or the fed back signal approaches a
good replica of the input.
CH 12 Feedback 192
 Comparison Error

Y R1
1 
X R2

CH 12 Feedback 193
 Loop Gain

VN
X 0
KA1 
Vtest

 When the input is grounded, and the loop is broken at an

arbitrary location, the loop gain is measured to be KA 1.
CH 12 Feedback 194
 Example: Alternative Loop Gain Measurement

VN  KA1Vtest
CH 12 Feedback 195
 Incorrect Calculation of Loop Gain

 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.
CH 12 Feedback 196
 Gain Desensitization

Y 1
A1 K  1 
X K
 A large loop gain is needed to create a precise gain, one
that does not depend on A 1, which can vary by ±20%.

CH 12 Feedback 197
 Ratio of Resistors

 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.

CH 12 Feedback 198
 Merits of Negative Feedback

 1) Bandwidth
enhancement

 2) Modification of I/O
Impedances

 3) Linearization

CH 12 Feedback 199
 Bandwidth Enhancement

Closed Loop
Open Loop
Negative A0
A0
A s   Feedback
Y 1  KA0
s  s 
1 X s
0 1
1  KA0 0

 Although negative feedback lowers the gain by (1+KA 0), it

also extends the bandwidth by the same amount.

CH 12 Feedback 200
 Bandwidth Extension Example

 As the loop gain increases, we can see the decrease of the

overall gain and the extension of the bandwidth.

CH 12 Feedback 201
 Example: Open Loop Parameters

A0  g m RD
1
Rin 
gm
Rout  RD
CH 12 Feedback 202
 Example: Closed Loop Voltage Gain

vout g m RD

vin R2
1 g m RD
R1  R2
CH 12 Feedback 203
 Example: Closed Loop I/O Impedance

RD
1  R2  Rout 
Rin   1  g m RD  R2
g m  R1  R2  1 g m RD
R1  R2
CH 12 Feedback 204
 Example: Load Desensitization

W/O Feedback With Feedback

Large Difference g m RDSmall Difference g m RD

g m R D  g m RD / 3 
R2 R2
1 g m RD 3 g m RD
R1  R2 R1  R2
CH 12 Feedback 205
 Linearization

Before feedback

After feedback

CH 12 Feedback 206
 Four Types of Amplifiers

CH 12 Feedback 207
 Ideal Models of the Four Amplifier Types

CH 12 Feedback 208
 Realistic Models of the Four Amplifier Types

CH 12 Feedback 209
 Examples of the Four Amplifier Types

CH 12 Feedback 210
 Sensing a Voltage

 In order to sense a voltage across two terminals, a

voltmeter with ideally infinite impedance is used.

CH 12 Feedback 211
 Sensing and Returning a Voltage

Feedback

Network

R1  R2 
 Similarly, for a feedback network to correctly sense the
output voltage, its input impedance needs to be large.
 R1 and R2 also provide a mean to return the voltage.
CH 12 Feedback 212
 Sensing a Current

 A current is measured by inserting a current meter with

ideally zero impedance in series with the conduction path.
 The current meter is composed of a small resistance r in
parallel with a voltmeter.

CH 12 Feedback 213
 Sensing and Returning a Current

Feedback

Network

RS 0
 Similarly for a feedback network to correctly sense the
current, its input impedance has to be small.
 RS has to be small so that its voltage drop will not change
Iout.
CH 12 Feedback 214
 Addition of Two Voltage Sources

Feedback

Network

 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.

CH 12 Feedback 215
 Practical Circuits to Subtract Two Voltage Sources

 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.

CH 12 Feedback 216
 Addition of Two Current Sources

Feedback

Network

 In order to add two current sources, we place them in

parallel. So the feedback network is placed in parallel with
the input signal.
CH 12 Feedback 217
 Practical Circuits to Subtract Two Current Sources

 Since M1 and RF are in parallel with the input current source,

their respective currents are being subtracted. Note, R F has
to be large enough to approximate a current source.

CH 12 Feedback 218
 Example: Sense and Return

 R1 and R2 sense and return the output voltage to

feedforward network consisting of M 1- M4.
 M1 and M2 also act as a voltage subtractor.

CH 12 Feedback 219
 Example: Feedback Factor

iF
K  g mF
CH 12 Feedback
vout 220
 Input Impedance of an Ideal Feedback Network

 To sense a voltage, the input impedance of an ideal

feedback network must be infinite.
 To sense a current, the input impedance of an ideal
feedback network must be zero.

CH 12 Feedback 221
 Output Impedance of an Ideal Feedback Network

 To return a voltage, the output impedance of an ideal

feedback network must be zero.
 To return a current, the output impedance of an ideal
feedback network must be infinite.

CH 12 Feedback 222
 Determining the Polarity of Feedback

 1) Assume the input goes

either up or down.

 2) Follow the signal through

the loop.

 3) Determine whether the

returned quantity enhances or
opposes the original change.

CH 12 Feedback 223
 Polarity of Feedback Example I

Vin  I D1  , I D 2  Vout  ,V x  I D 2  , I D1 
CH 12 Feedback Negative Feedback 224
 Polarity of Feedback Example II

Vin  I D1  ,V A  Vout  ,V x  I D1  , V A 
CH 12 Feedback Negative Feedback 225
 Polarity of Feedback Example III

I in  I D1  , V X  Vout  , I D 2  I D1  , V X 

CH 12 Feedback Positive Feedback 226

 Voltage-Voltage Feedback

Vout A0

Vin 1  KA0
CH 12 Feedback 227
 Example: Voltage-Voltage Feedback

Vout g mN (rON || rOP )


Vin R2
1 g mN ( rON || rOP )
R1  R2
CH 12 Feedback 228
 Input Impedance of a V-V Feedback

Vin
 Rin (1  A 0 K )
I in

 A better voltage sensor

CH 12 Feedback 229
 Example: V-V Feedback Input Impedance

Vin 1  R2 
  1  g m RD 
I in g m  R1  R2 
CH 12 Feedback 230
 Output Impedance of a V-V Feedback

VX Rout

I X 1  KA0 
 A better voltage source
CH 12 Feedback 231
 Example: V-V Feedback Output Impedance

 R1  1
Rout , closed  1  
 R2  g mN
CH 12 Feedback 232
 Voltage-Current Feedback

V out RO

I in 1  KRO
CH 12 Feedback 233
 Example: Voltage-Current Feedback

Vout  g m 2 RD1 RD 2

I in g m 2 RD1 RD 2
1
RF
CH 12 Feedback 234
 Input Impedance of a V-C Feedback

VX Rin

IX 1  R0 K

 A better current sensor.

CH 12 Feedback 235
 Example: V-C Feedback Input Impedance

1 1
Rin , closed  .
g m1 g m 2 RD1 RD 2
1
RF
CH 12 Feedback 236
 Output Impedance of a V-C Feedback

VX Rout

IX 1 R0 K

 A better voltage source.

CH 12 Feedback 237
 Example: V-C Feedback Output Impedance

RD 2
Rout , closed 
g m 2 RD1 RD 2
1
RF
CH 12 Feedback 238
 Current-Voltage Feedback

I out Gm

Vin 1  KGm
CH 12 Feedback 239
 Example: Current-Voltage Feedback

Laser

I out g m1 g m 3  rO 3 || rO 5 
|closed 
Vin 1  g m1 g m3  rO 3 || rO 5  RM
CH 12 Feedback 240
 Input Impedance of a C-V Feedback

V in
 Rin (1  KGm )
I in

 A better voltage sensor.

CH 12 Feedback 241
 Output Impedance of a C-V Feedback

VX
 Rout (1  KGm )
IX

 A better current source.

CH 12 Feedback 242
 Example: Current-Voltage Feedback

I out g m1 g m 2 RD
|closed 
Vin 1  g m1 g m 2 RD RM
1
Laser Rin |closed  (1  g m1 g m 2 RD RM )
g m1
1
Rout |closed  (1  g m1 g m 2 RD RM )
g m2

CH 12 Feedback 243
Wrong Technique for Measuring Output Impedance

 If we want to measure the output impedance of a C-V closed-

loop feedback topology directly, we have to place VX in series
with K and Rout. Otherwise, the feedback will be disturbed.

CH 12 Feedback 244
 Current-Current Feedback

I out AI

CH 12 Feedback
I in 1  KAI 245
 Input Impedance of C-C Feedback

VX Rin

I X 1  KAI
 A better current sensor.

CH 12 Feedback 246
 Output Impedance of C-C Feedback

VX
 Rout (1  KAI )
IX
 A better current source.

CH 12 Feedback 247
 Example: Test of Negative Feedback

Laser

I in  VD1  , I out  VP  , I F  VD1  , I out 

CH 12 Feedback Negative Feedback 248
 Example: C-C Negative Feedback

Laser

 g m 2 RD
AI |closed 
1  g m 2 R D ( RM / R F )
1 1
Rin |closed  .
g m1 1  g m 2 RD ( RM / RF )
Rout |closed rO 2 [1  g m 2 RD ( R M / RF )]
CH 12 Feedback 249
 How to Break a Loop

 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.

CH 12 Feedback 250
 Rules for Breaking the Loop of Amplifier Types

CH 12 Feedback 251
 Intuitive Understanding of these Rules

Voltage-Voltage Feedback

 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.
 Similar ideas apply to the other types.
CH 12 Feedback 252
 Rules for Calculating Feedback Factor

CH 12 Feedback 253
 Intuitive Understanding of these Rules

Voltage-Voltage Feedback

 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).
 Similar ideas apply to the other types.
CH 12 Feedback 254
 Breaking the Loop Example I

Av , open  g m1  RD ||  R1  R2  
Rin , open 1 / g m1
Rout ,open  RD ||  R1  R2 
CH 12 Feedback 255
 Feedback Factor Example I

K  R2 /( R1  R2 )
Av , closed  Av , open /(1  KAv , open )
Rin , closed  Rin , open (1  KAv , open )
Rout , closed  Rout , closed /(1  KAv , open )
CH 12 Feedback 256
 Breaking the Loop Example II

Av ,open  g mN  rON || rOP ||  R1  R2  

Rin, open 
Rout ,open rON || rOP ||  R1  R2 
CH 12 Feedback 257
 Feedback Factor Example II

K  R2 /( R1  R2 )
Av , closed  Av , open /(1  KAv ,open )
Rin , closed 
Rout ,closed  Rout , open /(1  KAv , open )
CH 12 Feedback 258
 Breaking the Loop Example IV

Vout RF RD1
|open  .  g m 2  RD 2 || RF  
I in 1
RF 
g m1
1
Rin ,open  || RF
g m1
Rout ,open  RD 2 || RF
CH 12 Feedback 259
 Feedback Factor Example IV

K  1 / R F
Vout Vout Vout
|closed  |open /(1  K |open )
I in I in I in
Vout
Rin ,closed  Rin, open /(1  K | open )
I in
Vout
Rout ,closed  Rout ,open /(1  K |open )
CH 12 Feedback I in 260
 Breaking the Loop Example V

I out g m 3  rO 3 || rO 5  g m1rO1
|open 
Vin rO1  RL  RM
Rin ,open 
Rout ,open rO1  RM
CH 12 Feedback 261
 Feedback Factor Example V

K  RM
( I out / Vin |closed ) ( I out / Vin |open ) /[1  K ( I out / Vin ) |open ]
Rin , closed 
Rout , closed  Rout ,open [1  K ( I out / Vin ) |open ]
CH 12 Feedback 262
 Breaking the Loop Example VI

I out g m1 RD
|open 
Vin R L  RM  1 / g m 2
Rin ,open 1 / g m1
CH 12 Feedback
Rout ,open (1 / g m 2 )  RM 263
 Feedback Factor Example VI

K  RM
( I out / Vin |closed ) ( I out / Vin |open ) /[1  K ( I out / Vin ) |open ]
Rin , closed  Rin ,open [1  K ( I out / Vin ) |open ]
Rout ,closed  Rout , open [1  K ( I out / Vin ) |open ]
CH 12 Feedback 264
 Breaking the Loop Example VII

( R F  RM ) R D  g m 2 rO 2
AI ,open  .
1 rO 2  RL  RM || RF
RF  RM 
g m1
1
Rin, open  || ( RF  RM )
g m1
Rout , open rO 2  RF || RM
CH 12 Feedback 265
 Feedback Factor Example VII

K  RM /( RF  RM )
AI ,closed  AI ,open /(1  KAI ,open )
Rin,closed  Rin,open /(1  KAI ,open )

CH 12 Feedback
Rout ,closed  Rout ,open (1  KAI , open ) 266
 Breaking the Loop Example VIII

Vout RF RD
|open  [  g m 2 ( RF || RM )]
I in RF  1 / g m1
1
Rin ,open  || RF
g m1
Rout , open  RF || RM
CH 12 Feedback 267
 Feedback Factor Example VIII

K  1 / RF
(Vout / I in ) |closed (Vout / I in ) |open /[1  K (Vout / I in ) |open ]
Rin , closed  Rin ,open /[1  K (Vout / I in ) |open ]
Rout ,closed  Rout , open /[1  K (Vout / I in ) |open ]
CH 12 Feedback 268
 Example: Phase Response

 As it can be seen, the phase of H(jω) starts to drop at 1/10

of the pole, hits -45o at the pole, and approaches -90 o at 10
times the pole.
CH 12 Feedback 269
 Example: Three-Pole System

 For a three-pole system, a finite frequency produces a

phase of -180o, which means an input signal that operates
at this frequency will have its output inverted.

CH 12 Feedback 270
 Instability of a Negative Feedback Loop

Y H ( s)
(s) 
X 1  KH ( s )
 Substitute jω for s. If for a certain ω1, KH(jω1) reaches
-1, the closed loop gain becomes infinite. This implies for a
very small input signal at ω1, the output can be very large.
Thus the system becomes unstable.
CH 12 Feedback 271
 “Barkhausen’s Criteria” for Oscillation

| KH ( j1 ) |1
KH ( j1 )  180

CH 12 Feedback 272
 Time Evolution of Instability

CH 12 Feedback 273
 Oscillation Example

 This system oscillates, since there’s a finite frequency at

which the phase is -180o and the gain is greater than unity. In
fact, this system exceeds the minimum oscillation
requirement.
CH 12 Feedback 274
 Condition for Oscillation

 Although for both systems above, the frequencies at which |

KH|=1 and KH=-180o are different, the system on the left is
still unstable because at KH=-180o, |KH|>1. Whereas the
system on the right is stable because at KH=-180o, |KH|<1.

CH 12 Feedback 275
 Condition for Stability

GX   PX
 ωPX, (“phase crossover”), is the frequency at which
KH=-180o.
 ωGX, (“gain crossover”), is the frequency at which |KH|=1.
CH 12 Feedback 276
 Stability Example I

| H p | 1
K 1

CH 12 Feedback 277
 Stability Example II

0.5 | H p | 1
K 0.5

CH 12 Feedback 278
 Marginally Stable vs. Stable

Marginally Stable Stable

CH 12 Feedback 279
 Phase Margin

 Phase Margin = H(ωGX)

+180

 The larger the phase

margin, the more stable
the negative feedback
becomes

CH 12 Feedback 280
 Phase Margin Example

PM 45 

CH 12 Feedback 281
 Frequency Compensation

 Phase margin can be improved by moving ωGX closer to

origin while maintaining ωPX unchanged.
CH 12 Feedback 282
 Frequency Compensation Example

 Ccomp is added to lower the dominant pole so that ωGX

occurs at a lower frequency than before, which means
phase margin increases.
CH 12 Feedback 283
 Frequency Compensation Procedure

 1) We identify a PM, then -180o+PM gives us the new ωGX, or

ωPM.
 2) On the magnitude plot at ωPM, we extrapolate up with a slope
of +20dB/dec until we hit the low frequency gain then we look
“down” and the frequency we see is our new dominant pole,
ωP’.
CH 12 Feedback 284
 Example: 45o Phase Margin Compensation

 PM  p 2
CH 12 Feedback 285
 Miller Compensation

Ceq [1  g m5 (rO 5 || rO 6 )]Cc

 To save chip area, Miller multiplication of a smaller
capacitance creates an equivalent effect.

CH 12 Feedback 286