Miller’s

Theorem

Z R
Z 1= → R1 = , C =C(1−A)
1− A 1− A 1

Bode Plot We evaluate the transfer function T ( s ) along the positive jω axis (
s= jω
) for positive ω , and poles and zeros located on negative σ axis.
( jω+ω z 1 ) ( jω+ω z 2 ) …( jω+ω zn)
T ( jω ) =K
( jω+ ω p 1 ) ( jω+ω p 2 ) …( jω+ω pN )

T jω =M ( ω ) e jϕ (ω)

ln ( T ( jω ) )=ln ( M ( ω ) ) + jϕ ( ω )

Common representation of Magnitude:

20 log |T ( jω )|=20 log [ M ( ω ) ]

20 log |T ( jω )|=20 log |K|+20 log √ ω2 +ω 2z 1 +20 log √ ω2 +ω 2z 2 +…+ 20 log √ω 2+ ω2zn log |K|−20 log √ ω

ω ω ω ω ω ω
ϕ ( ω ) =tan −1 +tan −1 +…+ tan −1 −tan−1 −tan−1 −…−tan −1
ωz 1 ωz 2 ω zn ωp 1 ωp 2 ω pN
 ω
tan −1 =45 ° ω=ω z 1
Key: ωz 1 when , and 90 ° and 0 ° above or below
ω=ω z 1
one dB with respect to
Bode Diagram Effects of poles and zeros is additive in bode diagrams
G ( jω )=G ( s ) |s= jω =R ( ω ) + jX (ω)

¿ G ( jω )∨¿
 Logarithmic gain: 20 log10 ¿

X (ω )
ϕ ( ω ) =tan −1
 Phase angle: R(ω)
 0.1 10 1
ω= →θ=0, ω= →θ=−90,ω= →θ=45
τ τ τ

2 2

(( ( ) ) ( ( ) ))
2
ω ω
20 log |G ( jω )|=−10 log 1− +4 ζ2
Typo in Log Gain. It should be ωn ωn
3dB Frequency The frequency at which the power is reduced by one half.
 AM
20 log10 ( A M )−3 dB=20 log 10 ( A M )+ 20 log 10 ( √12 )=20 log ( √2 ) 10

Finding s s

(
) ( ) ( )
)
ω L3 dB 1+ … 1+
and ( s +ω z 1 ) … ( s +ω z N ωz 1 ωz m
ω H 3 dB
(If the
T ( s )= A M F L ( s ) F H ( s ) =A M
( 1+
ωp 1
L

( s+ ωp 1 ) … ( s +ω p N
s
… 1+
L
s
ω pM
L

L
))
(
H

) (
H

)
locations of
H H

poles and zeros ω L3 dB ω H 3 dB

are known) Use the known locations of poles and zeros to estimate and .
Conditions:
ω L3 dB
-At low frequency range, is greater than any of the pole or zero
frequency
ω H 3 dB
-At high frequency range, is smaller than any of the pole or zero
frequency

ω L3 dB ≈ √ ω2p 1 + …+ω 2p N −2 ω 2z L −…−2 ω2Z n

L L L L

2 2 2 2 1
τ H 3 db ≈ √ τ p 1 H +…+ τ pMh−2 τ z 1 H −…−2 τ zmH , ω H 3 db=
τ H 3 db

Finding The Method of Open-Circuit and Short-Circuit Time Constants

ω L3 dB Conditions:
and
ω H 3 dB
For a pole: the next nearest poles, on either side, are at least two octaves (two
(If the octaves are a factor of four) away.
locations of
poles and zeros  FL (s )
are not known)
o Find Ris by
 Opening all the high frequency (small) capacitors
 Shorting all the other low frequency (large) capacitors
1
Ris C i i th “short-circuit time constant”,
o is the R is Ci is

the “short-circuit frequency”

o If one of the pole frequencies is greater than any of the other
pole or zero frequencies by about two octaves or above
o Then at frequencies above those of the other poles and zeros
N
s 1 1
FL (s ) ≈ b1 = ∑ ∵ s=b1 → F L ( s )=
s+b 1 , where i=i C i Ris . 2
N
1
o ∴ ω L3 dB ≈ b1=∑
i=i Ci Ris
  FH ( s )

o Find Rio by
 Opening all the other high frequency (small) capacitors
 Shorting all the low frequency (large) capacitors
1
Rio C i
o is the i th “open-circuit time constant”, R io Ci is

the “open- circuit frequency”

o If one of the pole frequencies is smaller than any of the other
pole or zero frequencies by about two octaves or above
o Then at frequencies below those of the other poles and zeros
N
1 1 1
FL (s ) ≈ d 1=∑ C i Rio . ∵ s= → F H ( s )=
1+d 1 s , where i=i
d1 2

1 1
∴ ω H 3dB ≈ =
d1 N
o ∑ C i Rio
i=i

BJT (npn)
Current Flow
V C >V B >V E

BE Conjunction:
Electrons from emitter to base > holes from base to emitter

Method:
Density of electrons in E > density of holes in C

Electron concentration in B at the E side:

v BE
VT
n p ( 0 )=n p 0 e

Straight-line concentration profile in B

d n p( x )
I n=A E q D n ( dx )
−n p ( 0 )
¿ A E q Dn ( W )
The Collector V BE
A q Dn n p 0 V
V BE

Current
i c =I s e VT
(
= E
W
e ) T

A E q Dn n p 0 A q Dn n2i
I s= ( W
= E
N AW ) (saturation current, strong function of
2
temperature because of ni (intrinsic carrier density).

iC v CB
Important: is independent of (as long as the collector is positive
with respect to base, the electrons that reach the collector side of the base
region will be swept into the collector and register as collector current)

The Base iC I S V
v BE

Current
i B= =
β β
e ( ) T

β : common-emitter current gain

The Emitter β+1

v BC

i E =iC +i B= I se V T

Current β

Recapitulation v BE
V BE

 The forward-bias voltage causes ic =I s e VT

iC
 is independent of v CB (as long as v CB ≥ 0 ¿ → the collector

terminal behaves as an ideal constant-current source determined by

v BE

1
i B= i C
 β , (β ≫ 1)

 i E =i B +i C
, ( iE ≅ iC )
i C =α i E
 , ( α≅ 1 )
 V BE =0.7(V )
Equivalent
Circuit Models

Small Signal IC ∂ iC
Approximation
i C =I C +
VT ( )
v be =I C + g m v be=I C +
∂ v BE i C= I C
v be

Input gm
i b= v
Resistance at β be
the Base. Base ∴ The small-signal input resistance between base and emitter, looking
Current
into the base, is
v be β V T V
rπ ≡ = = =β T
ib gm I B IC ( )
Input i I i
iE = C = C + c
Resistance at α α α
the Emitter. i I I
Emitter i e = c = C v be= E v be
α VT VT
Current
ie ic ib v be
Note: , and are proportional to . (Voltage-controlled
current source).

v be V T
re ≡ =
ie IE

v be =i b r π =i e r e

ie
rπ = () r
ib e

r π =( β +1 ) r e

Voltage Gain v C =V C −i R C

v c =−i R C =−g m v be RC =( −gm RC ) v be

 vc
≡ =−gm RC
Voltage gain v be

The voltage- −1

controlled
current
r 0=
[ | ]
∂ iC
∂ v CE v =constat
BE

source’s
output
impedance
Biasing the BJT The two operating points
−V BE V BB
I B= +
( 1+ β ) R E + R BB ( 1+ β ) R E + R BB

V BE
IC I S V
We also have I B= = e T

β β

Two Versions
of the 1/3rd
Rule

1
V B = V CC
 3
2
V C = V CC
 3
IE
 I1 =
√β
 2
V C = V CC
 3
1
V E = V CC
 3
IE
 I1 =
√β
The Common-
Emitter
Amplifier

Small Signal Model

Midband
 RBB ∨¿ r π
−g m R c ∨|RL ) ×( )
R S + RBB ∨¿ r π
v v v
A M = 0 = 0 × s =¿
v s vπ v0

v0
=−g m R c ∨¿ R L ≫ 1
Miller Gain: vπ

High Frquency

Z
Z 1= =C μ (1+ g m v π R C ∨¿ R L )
1− A
ZA
Z 2= ≈ Z=C μ
1− A
ω Z 1 H =ω Z 2 H =∞

C π +C μ ( 1+ gm v π RC||R L ) ] (RS||r π||R BB )

τ P 1 H =¿

τ P 2 H =C μ ( R L ∨¿ RC )
 C π +C μ ( 1+ gm v π RC||R L ) ] (RS||r π||R BB )
¿
C π +C μ ( 1+ gm v π RC||R L ) ] (RS||r π||R BB )
¿

{ }
1
C μ (R L ∨¿ RC )
¿×
1
s+
C μ ( R L ∨¿ RC )
1
s+
¿
¿
1
¿
¿
ω PL1 ω PL2
∴ F H ( s )= × =¿
s+ω PL1 s+ ω PL2

Low

ω Z 1 L =ω Z 2 L =0

ωZ 3 L
comes from Z E=∞ , which is a third condition to make iB=0
1
∴ ω Z3 L=
RE CE

ωP 1 L
comes from Z output =∞ (which gives a very large V 0 ¿
1
∴ ωP 1 L=
C C 2 ( R C+ R L )

ω P 2 L =C C 1( R S +r π ∨¿ R BB )

r π + R S∨¿ R BB
ω P 3 L =C E (R E ∨¿ ( 1+ β
) )
1 1 1
ωP 1 L+ ωP 2 L+ ωP 3 L= + +
τSC 1
τSC 2
τSC 3
 1
∵ ωP 1 L=
τ SC 1

1 1
∴ ωP 2 L+ ωP 3 L= +
τSC 2
τSC 3

CE
In order to minimize the component cost, associates with the smallest
short-circuit time constant
τ P3L ≪ τP2 L
, or ω P 2 L + ω P 3 L ≈ ω P 3 L
Then we can say that
1 1
ωP 3 L ≈ =
τ SC3 r + R ∨¿ RBB
C E ( R E ∨¿ π S
1+ β ( ) )
1
ωP 2 L ≠ τ SC2 CE
But τ SC 2 , because the computation of implies that is
ωP 2 L ≪ ωP 3 L CE
short circuit, while implies that should be an open
circuit.
ω C
Re-calculate P 2 L by opening E
RS + R BB∨|( r π + ( 1+ β ) R E ) ] C C 1
¿
1
∴ ω P 2 L ≈ τ OC 2=
¿

RS + R BB∨|( r π + ( 1+ β ) R E ) ] C C 1
¿
1

( )
s+
RE CE
¿
1
s+
r + R ∨¿ R BB
C E ( R E∨¿ π S
1+ β ( ) )
1
s+ ¿
s
¿
s +ω Z 1 L s +ω Z 2 L s +ω Z 3 L
∴ F L (s )≈
( s+ ω P 1 L )( )(
s+ ω P 2 L s+ ω P 3 L
=
) ( s+
s
1
) ¿

C C 2 ( RC + R L )

General design requirements to save cost:

 CE ∴ ωP 3 L
sees the smallest resistance has the highest frequency and
to be the dominant pole
CC 2
sees an intermediate resistance, ∴ ω P 1 L is in between ω P 3 L and
ωP 2 L
, subdominant pole
CC 1
sees the largest resistance, ∴ ω P 2 L is smallest, second sub-dominant
pole
The Common
Base Amplifier

Midband

rπ rπ
vπ
=
−R E∨¿ ( )
1+ β
≈−
( )
1+ β
vs rπ r
R S + R E∨¿ ( )
1+ β ( )
RS + π
1+ β

v0
=−g m ( RC ∨¿ R L )
vπ
 r
v 0
∴ =g (R ∨¿ R )
( 1+ β )
π

m C L
v r
R +(
1+ β )
s π
S

Low Frequency
ω Z 1 L =ω Z 2 L =0

1
ωZ 3 L=
C B R BB

1 1
ωP 1 L= =
τ P1L ( RC + R L ) C C2
r
¿
RBB ∨¿(r π + ( RE||RS ) ( β+1) ) ) (¿ π + RE ( 1+ β ) ¿ )
τ P 2 L, short =C B ¿ , ¿
RBB ∨¿
τ P 2 L ,open=C B ¿

τ P 3 L ,short =C C 1( R S + R E∨¿ ( 1+R β ))

π
( r +R )
π BB
, τ P 3 L ,open=C C 1 (R S + R E ∨¿ 1+ β )
 High Frequency
Note:
C
- no Miller multiplication of μ
Cπ
- is unaffected when looking into the emitter

1
rπ

F H ( s )=
1
C μ ( R L ∨¿ R C )
×
( )
Cπ
1+ β
||RS||R E
1 1
+s +s
C μ ( R L ∨¿ R C ) rπ
( )
Cπ
1+ β | S | E
|R | R

Common Features:
Collector  Power gain
Amplifier  High input impedance
 Low output impedance
 Wide bandwidth
 D.c. coupled at input
 V CC −V C 1 V CC −V B 2
I C 1= −I B 2= −I B 2
RC 1 RC 1

V B 2−0.7 V B2 V B2
I B 2= ∴ R ¿2= = (( 1+ β2 ) R E 2 ) ≈ 10 sK Ω
( 1+ β 2) R E 2 I B 2 V B 2−0.7

V CC −V C 1 V −V C 1
∴ I B2≪ , I C 1 ≈ CC
RC 1 RC 1

The CC stage does not significantly load the preceding stage at D.C.
Midband

R¿2=r π 2+(1+ β 2)(R E 2∨¿ R L )

v π 2 v 01
i b 2= =
r π 2 R¿ 2
 v 01 v v
i e2=( 1+ β2 ) i b 2 i i= +i b 2= 01 + 01
, R C1 R C 1 R ¿2

ie 2 ( 1+ β 2 ) R C1 RC1
∴ = ≈
i i RC 1 +r π 2+(1+ β 2)(R E 2 ∨¿ R L ) R E 2∨¿ R L

The CC amplifier gives a small-signal current gain

v 02 (1+ β 2 )( RE 2∨¿ R L )
= ≈1
v 01 r π 2 +(1+β 2 )(RE 2∨¿ RL )

The small-signal voltage gain is approximately +1 making the power gain

approximately equal to the current gain.

r π 2 + RC 1 r π 2 + RC 1
Routput =RE 2∨¿ ( 1+ β 2 )
≈
1+ β 2

The CC stage has a low out impedanc and is not easily loaded

High Frequency
 v x r π2 gm2 R∥
Z EQ= = +R∥
i x 1+ s r π 2 c π 2

r π 2 +r π 2 gm 2 R ∥ r π 2 (1+ gm 2 R ∥ ) 1
Zb= + R ∥= + R∥ =r π 2 ( 1+g m 2 R∥ ) ∥ +R∥
1+ s r π 2 c π 2 cπ2 s cπ 2
1+ s r π 2 (1+ gm 2 R ∥ )(
1+ gm 2 R∥
)
( )
( 1+ gm 2 R∥)

Open circuit:
Poles -
τ P 1 H = { RC 1 ∥ [ ( 1+ gm 2 R∥ ) r π 2 + R∥ ] } ( c μ 2+ c μ 1 ) ≈ R C1 ( c μ1 + c μ 2 )

cπ 2 cπ 2
τ P 2 H =[ ( 1+ g m 2 R∥ ) r π 2 ∥ ( RC 1 + R∥ ) ] ( 1+ g m 2 R∥ )
=( R C1 + R∥ ) (
1+ gm 2 R ∥ )
∵ τP 1H ≫ τP 2H

1
∴ ωP 1 H ≈
RC 1 ( c μ 1 +c μ 2 )

ωP 2 H
Recalculate by shorting (c μ 1 +c μ 2)
1 1+ β 2
ωp 2 H = ≈
cπ 2 r π 2 cπ 2
( )
R ∥ r ( 1+ g m 2 R∥ ) )
1+g m 2 R∥ ( ∥ π 2

Zeros (The frequency at which the output is 0) –

v o 2 =( y π 2 v π 2 +gm2 v π 2 ) R∥ =0

1 β 2 1+ β 2 1+ β 2
0= y π 2 + g m2 = +s c π 2 + = + s c π 2 → ωZ 1 H =
r π2 rπ 2 rπ 2 r π 2 cπ 2

Pole-zero Cancellation

The high frequency response of the common-collector amplifier is very

similar to a single-time-constant low pass filter with a frequency response of
 ω pH 1
FH ( s )≈ ω pH ≈
s +ω pH , where 2 RC 1 c μ 1

Differential Common-emitter, common-base, and cascode amplifier s use bypass

Amplifier capacitors to establish dc bias conditions, which have a detrimental effect of
introducing low-frequency zeros.

1 1
i E 1= v B2− v B1
i E 2= v B1− v B2
VT , VT
1+e 1+e

Small signal linear operation requires that the differential voltage

|V d|=|V B 1−V B 2|≪ 2 V T
v
i C1 =
αI
2 (
1+ d
2VT )
 v
i C 2=
2(
αI
1− d
2VT )
αI αI v d v αI
i C =I C +i c = ±
2 2V T 2( ) ( ) , giving
=I C ± g m d
2
gm=
2VT

Differential Voltage Gain

v0
=−gm RC
vd

Small Signal Model

v π 1=−v π 2
Miller Gain
−gm v π
2 R C RL
v0 R L +2 RC −RC RL g m
k= = =
2vπ 2vπ R L +2 R C

Output stage:

V1 1
Z 1= =
I 1 s cμ
( 1−k )
2
V2 1
Z 2= =
I 2 s cμ 1
2 ( )
1−
k
High frequency Poles:
1
ω HP 1=
c μ ( 1−k ) c π
( 2
+
2 )
( 2 r π ∥ Rs )

1
ω HP 2=
cμ
( ( ))
2
1−
1
k
( 2 RC ∥R L )

Midband

−gm v π
2 R C RL
v0 R L +2 RC −RC RL g m
k= = =
2vπ 2vπ R L +2 R C

2 vπ 2 rπ
=
v s Rs +2 r π

v 0 v π −RC R L gm 2r π RC RL rπ
∴ A M =2
2 vπ vs
= (
R L +2 RC )( )
R s +2 r π (
=(−2 g m ) )(
R L +2 R C Rs +2 r π
)
Current Mirror

I REF
I0 =
(1+ 2β )
Basic Op-Amps

15 V −0.7V
I C 3=0.5 mA
I Ref =
28.6 k Ω
=0.5 mA , , I E 1=I E 2 ≈ I C 1=I C 2=0.25 mA

0.25 mA
gm 1=gm 2= =0.01 ℧ , I C 6 ≈ 0.5 mA ,
25 mV
 1mA
I C 4=I C 5=1 mA gm 4 =gm 5= =0.04 ℧
, 25 mV ,
β
r π 4 =r π 5= =100 ×25 ℧ =2.5 k Ω
g m 4,5

∵ I C 5=1 mA
, V B 7 =V C 5=15 V −3 k Ω ×1 mA=12 V
∴ V E 7=12 V +0.7 V =12.7V

15V −12.7 V 2.3 V

∴ I E 7= = =1 mA
2.3 k Ω 2.3 k Ω
1 mA β
∴ g m 7= =0.04 ℧ ,r π 7 = =2.5 k Ω
25mV gm 7

∴ R input 7=r π 7 + β RE 7 ≈ 230 k Ω

Midband

Stage 1
v o 1=v i 2=−g m v π 1 ( 2 RC ∥ ( r π 4 +r π 5 ) )=−gm 1 ( v2 ) 2 R ( r
d
C
r π 4 +r π 5
π 4 +r π 5 +2 R C
)
vi 2
∴ A M 1= =−22
vd

Stage 2
 v
v o 2 =v i 3=gm 4 v π 4 R C 5 ∥ Ri 7 i2
, v π4= 2
vo 2
A M 2= =60
vi 2

Stage 3
Ri 8 ≈ 300 k Ω

kΩ∥R
A M 3 =−β ( 15.7
β ×2.3 k Ω )
=−6.42 i8

Stage 4
A M 4 ≈1

∴ A M = A M 1 A M 2 A M 3 A M 4 =−22 ×60 × (−6.42 ) ×1=8574 ≈ 8600

Slew Rate Maximum rate of change of the output waveform

d v0
∨max
dt

In the case of unity-gain butter with sinusoidal input signal:

The input has its largest slope when sin ( ωt )=0
If the buffer is to follow this signal
The output must be able to change at a rate equal to the great slope of the
input
d vi
dt |
sin ( ωt ) =0
=
d V i sin ⁡( ωt)
dt |
sin ( ωt ) =0
=ω V i cos ⁡( ωt )|sin (ωt )=0=ω V i

ωt V i ωt
Slew Rate = , where is the maximum radian frequency that will
Vi
not result in distortion of the output signal for a given input voltage
Full-Power Bandwidth at which the amplifier can deliver maximum power to the load
Bandwidth without distorting the waveform of a test sinusoid.
 This happens when the maximum voltage is generated at the output
ω m V max =¿ Slew Rate

SlewRate
f m=
2 π V max

Common- Two inputs of an op amp are shorted and a common signal is applied to the
Mode input terminals
Rejection Ratio v0
A =
cm
v icm

−¿
+¿−v ¿
Differential Signal: v =v
id ¿

v −¿
+¿+
2
Average, or common-mode input signal: v¿
vicm =¿

v 0 =A d v id + A cm v icm

Common-Mode Rejection Ratio

|A d| |A d|
CMRR=
|Acm|
∨20 log
( )
|A cm|

v 0 =−gm v π [ R c− ( Rc + Δ Rc ) ]

rπ
v
2 cm
v π=
( r2 )+ R(1+ β)
π
 v cm Δ Rc v cm Δ Rc
v 0 =g m
( )
1+ 2 g π R ( 1+ β )
=
rπ
≈ α v cm ( Δ2RR )≈ v ( Δ2RR )
c
cm
c

1
gm
+2 R( )( )
1+ β
rπ β

Δ Rc
A cm =
2R
∵ A d=−gm Rc
(without the rest stage attached)

CMRR= | ||
Ad
Acm
=
−gm Rc
Δ Rc
|
=gm 2 R
Rc
Δ Rc ( )
( )
2R

Introduction to Feedback
It is used to  De-sensitize Gain
 Reduced Distortion
 Extend Bandwidth
 Control Input and Output Impedence
 Increase Signal-to-Noise Ratio

At the price of a reduction of gain

Basic Feedback
Configuration

Terminologies:
Gain without feedback: A
x0
Af ≡
Gain of the feedback amplifier: xs

Loop gain: Aβ
Amount of feedback: 1+ Aβ
 x 0 A xi A xi Axi A 1
Af ≡ = = = = ≈ ( AB ¿ 1 ¿
x s x i + x f x i+ β x 0 x i + Aβ xi 1+ Aβ β

Significance: The feedback amplifier is entirely determined by the feedback

network when A (gain without feedback, A), is large.

x f = β x 0= Aβ x i=Aβ ( x s−x f )

Aβ x s
xf = =x s
1+ Aβ

Significance: The output tracks the input.

+¿
−¿=v ¿
( +¿=v s , v ¿
v¿
R 1+R 2 1
v +¿ = =
R1 β
v−¿(R +R )
1 2

R1
¿
v0
A f = =¿
vs

R1
β=
R 1+ R 2

Properties of Gain De-sensitivity

Negative
Feedback d Af 1 Aβ 1
= − =
dA 1+ Aβ ( 1+ Aβ ) (1+ Aβ )2
2

“The gain is de-sensitized by the amount of feedback”, therefore, 1+ Aβ is

 also called the “de-sensitivity factor”.
Bandwidth Low Pass, Single-pole amplifier
Extension AM
A ( s )=
s
+1
ωH

AM
1+ A M β
A f ( s )=
s
1+
ω H (1+ A M β )

 Mid-band gain is reduced by 1+ A M β

 3-dB frequency is extended by 1+ A M β (bandwidth is extended)

High Pass, Single-pole amplifier

A ( s )= A M ( s +ωs )
L

AM
s
1+ A M β
A f ( s )=
ωL
s+
1+ A M β

 Mid-band gain is reduced by 1+ A M β

 3-dB frequency is also reduced by 1+ A M β (bandwidth is extended)
Nonlinear
Distortion
Reduction
Input and
Output
Impedance
Control Using
Feedback
(Voltage
Amplifier,
Series-Shunt)

V0 A
Af = =
V s 1+ Aβ

Rif =Ri (1+ Aβ)

R0
Rof = → ideal for voltage amplifiers
1+ Aβ
Input and
Output
Impedance
Control Using
Feedback
(Current
Amplifier,
Shunt-Series)

I0 A
Af = =
I s 1+ Aβ

Ri
Rif = Rof =R 0 (1+ Aβ ) → ideal for current amplifier
1+Aβ ,

Shunt-Shunt A Ri R0
Af = Rif = Rof =
Feedback 1+ Aβ , 1+ Aβ , 1+ Aβ
Amplifier A: transresistance
β : transconductance

Series-Series A
Af = Rif =Ri (1+ Aβ) , Rof =R 0 (1+ Aβ )
1+ Aβ

A: transconductance
β : transresistance
Practical Case

h12=
V1
V2 |
I 1=0
( I 1 is the black arrow)
Method At first, the circuit is

 Identify the feedback network so as to find h parameters

 The feedback network is
In order to find the relation of h parameters. At the input
 Series connection at the input -> choose a series arrangement of h for
the input. Choose parallel arrangement for the output.
I1
 and V 2 are the independent parameters -> mark the
I1 V2
dependent source using and

V 1=I 1 h11 + V 2 h12

 , I 2 =I 1 h21+ V 2 h22
Therefore

Therefore
 h11 =R1∨¿ R2
R1
 h12=
R 1+ R 2

1
h22=
 R1+ R 2

V2 V1
(Note and are independent parameters)
h11 h22
At this time, and are in the feedback network.

 Using these relations, find all the h parameter

  Absorb h11 and h22 out of the feedback network and into the
amplifier circuit

Now the New amplifier circuit is

R1
β=h12=
R1 + R2

A'
Af =
1+ A ' β
' ' '
Rif =Ri ( 1+ A β ) =(R s + Ri + R1∨¿ R2 )(1+ A β )

Ri
Be careful about the difference of (the input impedence of the A
Ri '
circuit), (that of A’ circuit (i.e., with h parameters and load
Rif
impedence)), (that of the circuit with feedback (it’s calculated using
the convenient source), and R¿ (the actual input impedence of the circuit
 Rs
with feedback, which excludes )
Stability A f ( jω )=
A ( jω)
1+ A ( jω ) β ( jω)
j ϕ (ω )
Loop gain: L ( jω )= A ( jω ) B ( jω )=| A ( jω ) B ( jω )| e

When ϕ=180 °
1) 1>| A ( jω ) B ( jω )|> 0 , A f ( jω )> A ( jω) , stable
2) | A ( jω ) B ( jω )|=1, A f ( jω ) =∞ , oscillate
3) | A ( jω ) B ( jω )|>1 , in which case the input signal will grow steadily until
some form of non-linearity in the circuit causes the loop-gain to be reduced
to 1 and the circuit will oscillate.
Amplifier with
a Single-Pole
Response

,
0 p A ω
When ω ≫ ω pf , A f ( s )= s ≈ A ( s )

Amplifier with
a two-pole
Response

Stability Using A0 β ≫ 1
For
Bode Plots A0 1
Af = ≈
1+ A 0 β β

A ( j ω1 ) ℶa=1e− jθ

( θ=180 °− phase margin )

 1 − jθ
e
A ( jω 1) β
A f ( j ω1 )= =
1+ A ( j ω 1) β 1+e jθ
− jθ
¿ 1+ e ∨¿
1
| A f ( j ω1 )|= β¿

20 log |A ( jω )|−20 log

| | 1
β ( jω )
=20 log |A ( jω ) β ( jω )|
Closure Rule of If
Thumb  The intercept is on the section with a slope of -20dB/decade

 20 log |1β | closes on 20 log|A ( jω )| with a difference of less than

20dB/decade

The amplifier will not be unstable

Frequency
Compensation