Sei sulla pagina 1di 24

CHAPTER 8 FREQUENCY RESPONSE

Chapter Outline

8.1 Low-Frequency Response of the CS and CE Amplifiers

8.2 Internal Capacitive Effects and the High-Frequency Model

8.3 High-Frequency Response of the CS and CE Amplifiers

8.4 Tools for the Analysis of the High-Frequency Response of Amplifiers

8.5 A Closer Look at the High-Frequency Response*

8.6 High-Frequency Response of the CG and Cascode Amplifiers*

NTUEE

Electronics

L.H. Lu

8-1

Frequency response of amplifiers

Midband:

The frequency range of interest for amplifiers

Large capacitors can be treated as short circuit and small capacitors can be treated as open circuit

Gain is constant and can be obtained by small-signal analysis

Low-frequency band:

Gain drops at frequencies lower than f L

Large capacitors can no longer be treated as short circuit

The gain roll-off is mainly due to coupling and by-pass capacitors

Hi h-fre uenc band:

g

q

y

Gain drops at frequencies higher than f H

Small capacitors can no longer treated as open circuit

The gain roll-off is mainly due to parasitic capacitances of the MOSFETs and BJTs

 The gain roll-off is mainly due to para sitic capacitances of the MOSFETs and BJTs

NTUEE

Electronics

L.H. Lu

8-2

8.1 Low-Frequency Response of the CS and CE Amplifiers

The CS amplifier

Small-signal analysis:

V

g

 

V

 

R

G

 

V

sig

 

R

G

s

 

sig

R

1

R

   

R

G

R

sig

s

 

1

 
 

G

1

sC

C 1

sig

C

C

1

(

R

G

R

sig

)

 

 

P

1

C

C

1

(

R

G

R

sig

)

I

V

g

 

V

s

d

1

 

1

g

m

g

s

g

m

 

g

m

sC

C

C

S

 

P

2

 

g

C

m

S

V

o

 

I

R



L

I

d

 

R

D

 

R

L



I

d

 

R

D

R

L

 

s

 
 

o

 

R

1

R

   

R

D

R

L

   

1

 

 

1

D

sC

C 2

L

 

s

C

C

2

(

R

D

R

L

)

P

V

o

3

C



C

2

(

g

m

R

R

D

G

(

R

R

D

L

||

)

R

L

)

 

s

   

s

   

s

V

sig

A



g

m

R

G

R

G

(

R

R

sig

D

||

R

L

)

s

P

1

s

P

2

s

P

3

M

   
 

R

G

R

sig

 

NTUEE

Electronics

P 3   M       R G  R sig   NTUEE Electronics
P 3   M       R G  R sig   NTUEE Electronics

L.H. Lu

8-3

Determining the lower 3-dB frequency

Coupling and by-pass capacitors result in a high-pass frequency response with three poles

If the poles are sufficiently separated Bode plot can be used to evaluate the response for simplicity The lower 3-dB frequency is the highest-frequency pole

P2 is typically the highest-frequency pole due to small resistance of 1/g m

If the poles are located closely

The lower 3-dB frequency has to be evaluated by the transfer function which is more complicated Determining the pole frequency by inspection

Reduce V

sig

to zero

Consider each capacitor separately (treat the other capacitors as short circuit)

Find the total resistance between the terminals

Selecting values for the coupling and by-pass capacitors

These capacitors are typically required for discrete amplifier designs

C S is first determined to satisfy needed f L

C C1 and C C2 are chosen such that poles are 5 to 10 times lower than f L

C 1 and C C 2 are chosen such that poles are 5 to 10 times

NTUEE

Electronics

L.H. Lu

8-4

The CE amplifier

The CE amplifier  Small-signal analysis  Considering the effect of each capacitor separately  Considering

Small-signal analysis

The CE amplifier  Small-signal analysis  Considering the effect of each capacitor separately  Considering

Considering the effect of each capacitor separately Considering only C C1 :

V

V

 

R

B ||

r

sig

R

B

||

r

R

sig

1

sC

C

1

V o

V

o



g



m

V

g

m

(

R

(

R

B

C

||

||

R

r

L

)

)(

R

C

||

R

L

)

s

V

sig

   

R

B

||

r

R

sig

s

 

1

 

C

C

1

(

R

B

||

r

R

sig

)

1

 

C

C

1

(

R

B

||

r

R

sig

)

 
 

A



g

m

(

R

B

||

r

)(

R

C

||

R

L

)

M

 

R

B

||

r

R

sig

)( R C || R L ) M   R B || r   R

NTUEE

Electronics

L.H. Lu

8-5

Considering only C E :

 

I

b

V

 

R

B

 

1

sig

 

V

o

V

o

R

B

R



I

b



(

R

C

R

B

||

sig

R

L

R

)

B

||

R

(

sig

R

C

||

(

R

L

)

1)

r

e

1

sC

E

s

V

sig

R

B

R

sig

R

B

||

R

sig

(

1)

r

e

s

 

1

         

r

e

R

B

||

R

sig

 

1

C E

 

1

P

2

C

r

R

B

||

R

sig

 

E

e

1

Considering only C C2 :

 
   

R

B

||

r

V

V

sig

||

r

 
 

R

B

R

sig

 

V



V

R

C

 

R

 

o

V

o

g

m

g

m



(

R

C

R

B

||

1

sC

C 2

r

)(

R

R

L

C

||

R

L

)

L

 

s

 

V

sig

 

R

B

||

r

R

sig

 

s

 

1

   

1

 

C

C

2

(

R

C

R

L

)

 

P

3

C

C

2

(

R

C

R

L

)

NTUEE

Electronics

R L )    P  3 C C 2 ( R C  R

L.H. Lu

8-6

Determining the lower 3-dB frequency

Coupling and by-pass capacitors result in a high-pass frequency response with three poles

o

V

V

sig

M

 A

s

s

P1

  

 

  

s

s

P2

  

 

 

s

s

P2

 

The lower 3-dB frequency is simply the highest-frequency pole if the poles are sufficiently separated

The highest-frequency pole is typically P2 due to the small resistance of R E

An approximation of the lower 3-dB frequency is given by

f

L

 

1

1

 

1

 

1

 

2

C

C

1

R

C

1

C

E

R

E

C

C

2

R

C

2

 

Selecting values for the coupling and by-pass capacitors

These capacitors are typically required for discrete amplifier designs

C E is first determined to satisfy needed f L

C C1 and C C2 are chosen such that poles are 5 to 10 times lower than f L

NTUEE

Electronics

L.H. Lu

8-7

8.2 Internal Capacitive Effects and the High-Frequency Model

The MOSFET device

There are basically two types of internal capacitance in the MOSFET

Gate capacitance effect: the gate electrode forms a parallel-plate capacitor with gate oxide in the middle

Junction capacitance effect: the source/body and drain/body are pn-junctions at reverse bias

The gate capacitive effect

MOSFET in triode region:

C

gs

C

gd

1 WLC

2

ox

C

ov

MOSFET in saturation region:

C

gs

2

3

WLC

ox

C

ov

C

gd

C

ov

MOSFET in cutoff region:

C

gs

C C

gd

ov

C

gb

WLC

ox

Overlap capacitance:

C

ov

WL

ov

C

ox

The junction capacitance

C ov  WL ov C ox  The junction capacitance  Junction capacitance includes components

Junction capacitance includes components from the bottom side and from the side walls

The simplified expression are given by

C

sb

C sb

0

1

V

SB

/V

0

are given by C sb  C sb 0 1  V SB / V 0

C

db

C db 0 1  V /V DB 0
C db
0
1
 V
/V
DB
0

NTUEE

Electronics

L.H. Lu

8-8

The high-frequency MOSFET model

 The high-frequency MOSFET model   C ( W / L ) | V |
  C ( W / L ) | V |  g m n
C
(
W
/
L
)
|
V
|
g m
n
ox
ov
 g
g mb
m
r
| V
| /
I
o
A
D
2 WLC
WL C
C gs
3 ox
ov
ox
 WL C
C gd
ov
ox
C sb
0
C sb
1
 |
V
| /
V
SB
o

C db

C

 db 0 1  | V | / V DB o
 db
0
1
 |
V
| /
V
DB
o
2  C ( W / L I ) n ox D
2
C
(
W
/
L I
)
n
ox
D

Simplified high-frequency MOSFET model

Source and body terminals are shorted

C gd plays an important role in the amplifier frequency response

C db is neglected to simplify the analysis

the amplifier frequency response  C d b is neglected to simplify the analysis NTUEE Electronics

NTUEE

Electronics

L.H. Lu

8-9

The unity-gain frequency (f T )

The frequency at which the current gain becomes unity

Is typically used as an indicator to evaluate the high-frequency capability

Smaller parasitic capacitances C gs and C gd are desirable for higher unity-gain frequenc

I

o

g

m

V

gs

sC

gd

V

gs

V

I

i

gs

(

s C

gs

C

gd

)

I

o

g

m

I

i

(

s C

gs

C

gd

)

 

1

g

m

f

T

2

C

gs

 

C

gd

 

g

m

V

gs

C gs    C gd   g m V gs   The unity-gain frequency

The unity-gain frequency can also be expressed as

I

o

f

T

f

T

g V  sC V  g V m gs gd gs m gs 1
g
V
sC
V
g
V
m
gs
gd
gs
m
gs
1
2
(
W
/
L I
)
 n
C ox
D
2
 C
 C
gs
gd
1
C
(
W
/
L V
)
n
ox
OV
2
C
C
gs
gd
3  I n D  2  L 2 C WL ox 3 
3
I
n
D
2
L
2 C WL
ox
3
n
V OV
2
4
L

The unity-gain frequency is strongly influenced by the channel length Higher unity-gain frequency can be achieved for a given MOSFET by increasing the bias current or the overdrive voltage

NTUEE

Electronics

L.H. Lu

8-10

The BJT Device

High-frequency hybrid-model:

The base-charging or diffusion capacitance C de :

C

de

F

g

m

I

C

F V

T

The base-emitter junction capacitance C je :

C

je

C

2

j

0

The collector-base junction capacitance C :

C

C

0

1

V

CB

V

0c

m

C  0     1  V CB V 0 c  

The cutoff (unity-gain) frequency:

I

c

V

h

h

fe

fe

f

T

(

g

m

sC

)

V

 

I

b

(

I

c

||

C

||

C

)

 

I

b

r

g

m

sC

1/ r

sC

sC

I

b

1/

g

r

m

r

s ( C

C

)

0

 

1

1

s ( C

g

m

)

C

r

 

1

1

s ( C

I C

C

)

r

 

 

2

C

C

2

(

C

C

)

V

T

 C   C  2  ( C   C  ) V

NTUEE

Electronics

L.H. Lu

8-11

8.3 High-Frequency Response of the CS and CE Amplifiers

The common-source amplifier

Midband gain:

A

M



R

G

R

G

R

sig

(

g

m

R

L

)

Frequency response:

V sig

V

gs

V

gs

R sig

R

G

sC

gs V

gs

sC

gd

(

V

gs

V

o

)

g

m

V

gs

sC

V

o

R

L

gd

(

V

gs

V

o

)

V o V sig
V
o
V
sig



R R   1 s

C

 

g

 

L

G

 

gd

 

m

R

 

R

   

g

 

sig

L

m

1

R

sig

R

G

  

 

 

R

L

R

L

1 g

m

R

C

C   s

2

R R

L

sig

R

G

C

C

s

R

sig

R

G

R

sig

R

G

L

gd

gs

R

sig

R

G

gd

gs

The common-source amplifier has one zero and two poles at higher frequencies

The amplifier gain falls off at frequencies beyond midband

The amplifier bandwidth is defined by the 3-dB frequency which is typically evaluated by the dominant pole (the lowest-frequency pole) in the transfer function

evaluated by the dominant pole (the lowest-frequency pole) in the transfer function NTUEE Electronics – L.H.
evaluated by the dominant pole (the lowest-frequency pole) in the transfer function NTUEE Electronics – L.H.

NTUEE

Electronics

L.H. Lu

8-12

Simplified analysis technique

Assuming the gain is nearly constant (A M )

Find the equivalent capacitance of C gd at the input (with identical I gd )

I gd

I

C

gd

eq

sC

sC

sC

( V

gs

V

o

)

sC

gd

[

V

gs

 g RV

(

m

L

)

gs

]

o )  sC gd [ V gs   g R  V ( m

gd

gd

(1g R)V

m

gs

eq

V

gs

m

R)C

L

gd

(1g

Miller effect

Neglect the small current I gd at the output

o

A

 

R

G

  R  

g

m

L

 

R

sig

R

G

1

R

s

{

R

sig

G

R

G

R

sig

[(1

(

g

g

m

m

R

L

)

R

C

L

)

gd

C

gs

]}

C

gs

]

1

 

sig

[(1

g

m

R

L

)

C

gd





R

 g m R L ) C gd     R V V sig

V

V sig

A

M



1

s

H

M

H

The dominant pole is normally determined by C eq

The frequency response of the common-source amplifier is approximated by a STC

NTUEE

Electronics

L.H. Lu

8-13

The common-emitter amplifier

The common-emitter amplifier  Simplified analysis for frequency response   Miller effect: replacing C 
The common-emitter amplifier  Simplified analysis for frequency response   Miller effect: replacing C 

Simplified analysis for frequency response

Miller effect: replacing C with C eq

The response is approximated by a STC

V

o

A

M

V

sig

1

H

A 

M

H

{[

R

B

r

R

C

B

R

(1

sig

r

g

m

R

R

)

||

L

B

C

R

sig

][

r

g

m

R

L

R

B

||

R

sig

]}

1

g m  R R ) ||  L B C  R sig ][ r

NTUEE

Electronics

L.H. Lu

8-14

8.4 Useful Tools for the Analysis of the High-Frequency Response of Amplifiers

Determining the upper 3-dB frequency

General transfer function of the amplifier

 

A

F

 

A

(1

/

s

Z

1  (1

)(1

s

/

Z

2

)

s

/

Zm

)

)

)

M

A

H

F

(

(

s

)

)

A

M

(1

(1

/

s

/

s

P

Z

1  

s

P

2

)(1

/

)

(1

1 )(1

s

/

Z

2

)

s

/

Pn

)

M

2

H

s

1

(1

M

(1

/

s

P

1 )(1

s

/

P

2

)

2

/

H

2

Z

1  

H

Z 1

)(1

2

/

2

)

1

2

H

1

2

Z

1

1

2

Z 2

4

H

1

2

Z

1

2

1

Z 2

 

(

j

H

)

 

2

 

(1

2

/

H

1

2

P

1  

H

)(1

2

/

2

P 1

)

1

2

H

1

2

P

1

1

2

P 2

4

H

 

1

2

P

1

2

1

P 2

1

2

P

1

1

2

P

2

2   

1

2

Z

1

1

2

Z 2

 

(

A s

F

H

H

(

A s

The upper 3-dB frequency

dominant-pole response

1

2

H

1

2

Z

1

1

Z 2

2

1

2

H

1

2

P

1

1

P 2

2

One of the poles is of much lower frequency than any of the other poles and zeros

A dominant pole exists if the lowest-frequency pole is at least 4away from the nearest pole or zero

H

 

1

1

2

P

1

1

2

P

2

2

  

1

2

Z

1

1

2

Z 2

P

1

NTUEE

Electronics

L.H. Lu

8-15

Open-circuit time constant method to evaluate amplifier bandwidth

General transfer function of the amplifier

F

H

( s

)

(1

s

/

Z

1 /

)(1

s

Z

2

)

 (1

s

/

Zm

)

(1

s

/

P

1 /

)(1

s

P

2

)

 (1

s

/

Pn

)

1

a s

1

a s

2

2



a

m

s

m

1

b s

1

b s

2

2



b s

n

n

a m s m 1  b s 1  b s 2 2  b

b 1

1

P

1

1

P

1



1

Pn

Open-circuit time constant (exact solution):

b

1

n

i 1

C R

i

i

Dominant pole approximation:

b

1

1

 

1

 

1

 

1

 
 



P

1

P

2

Pn

P

1

 

1

1

 

H

P 1

b

1

i

C R

i

i

Miller’s Theorem

A technique to replace the bridging capacitance The equivalent input and output impedances are:

V

2

KV

1

equivalent input and output impedances are: V 2  KV 1 Z Z  1 1

Z

Z

1 1 K

Z

2

Z

1

1/ K

and output impedances are: V 2  KV 1 Z Z  1 1  K

NTUEE

Electronics

L.H. Lu

8-16

Time-constant method*

A technique used to determine the coefficients of the transfer function from the circuit

F

H

(

s

)

(1

s

/

Z

1 )(1

s

/

Z

2

)

 (1

s

/

Zm

)

(1