28

Frequency Response of Transistor Amplifiers

■ Simplest case: CE short-circuit current gain Ai(jω) as a function of frequency

Cµ Io

+
Iin rπ Vπ Cπ gmVπ

Kirchhoff’s current law at the output node:

I o = g m V π – V π jωC µ
Kirchhoff’s current law at the input node:
Vπ
Z π = r π  -------------
1
I s = ------ + V π jωC µ where
Zπ  jωC 
π

■ Solving for Vπ at the input node:

Is
V π = --------------------------------------
( 1 ⁄ Z π ) + jωC µ

EECS 105 Fall 1998

Lecture 28
 Short-Circuit Gain Frequency Response

■ Substituting Vπ into the output node equation--

 jωC µ
g m Z π  1 – -------------
Io  gm 
---- = ------------------------------------------
Is 1 + jωC µ Z π

■ Substituting for Zπ and simplifying --

 jωC µ  jωC µ
ω
g m r π  1 – ------------- β o  1 – ------------- 1 – j ------
Io  g m   g m  ωz
---- = ----------------------------------------------- = ----------------------------------------------- = β o ------------------
Is 1 + jωr π ( C π + C µ ) 1 + jωr π ( C π + C µ ) ω
1 + j ------
ωp

Current gain has one pole:

–1
ωp = ( rπ ( Cπ + Cµ ) )
and one zero
–1 –1
ωz = ( gm Cµ ) » ωp

EECS 105 Fall 1998

Lecture 28
 Bode Plot of Short-Circuit Current Gain

■ Note low frequency magnitude of gain is βo

Io
Iin
βo = gmrπ

Cµ
Cπ + Cµ

1 gm gm ω
rπ(Cπ + Cµ) Cπ + Cµ Cµ log scale

(a)

I
∠ o
Iin

−45
−90
−135

−180

1 gm ω
log scale
rπ(Cπ + Cµ) Cµ

(b)

■ Frequency at which current gain is reduced to 0 dB is defined as the transition

frequency ωT. Neglecting the zero,
gm
ω T = -------------------------
( Cπ + Cµ )

EECS 105 Fall 1998

Lecture 28
 Transition Frequency of the Bipolar Transistor

■ Dependence of transition time τT = ωT-1 on the bias collector current IC:

1 Cπ + Cµ g m τ F + C jE + C µ
τ T = ------- = -------------------- = ------------------------------------------
ωT gm gm

 C jE + C µ V th
τ T = τ F +  ---------------------- = τ F + -------- ( C jE + C µ )
 gm  IC

■ If the collector current is increased enough to make the second term negligible,
then the minimum τT is the base transit time, τF.. In practice, the ωT decreases at
very high values of IC due to other effects and the minimum τT may not be
achieved.
■ Numerical values of fT = (1/2π)ωT range from 10 MHz for lateral pnp’s to 10
GHz for oxide-isolated npn’s
Note that the small-signal model is not valid above fT (due to distributed effects
in the base) and the zero in the current gain is not observed

EECS 105 Fall 1998

Lecture 28
 Common-Source Current Gain

■ CS amplifier has a non-infinite input impedance for ω > 0 and we can measure
its small-signal current gain.

Cgd
Io

+
Iin Vgs Cgs gmVgs
−

■ Analysis is similar to CE case; result is

 jωC gd
g m  1 – ----------------
Io  gm  gm
------ = -------------------------------------- ≈ ----------------------------------
I in jω ( C gs + C gd ) ω ( C gs + C gd )

■ Transition frequency for the MOSFET is

gm
ω T ≈ -------------------------
C gs + C gd

EECS 105 Fall 1998

Lecture 28
 Transition Frequency of the MOSFET

■ Substitution of gate-source capacitance and transconductance:

2 W
C gs = --- WLC ox » C gd and g m = ----- µ n C ox ( V GS – V Tn )
3 L
W
g m ---- - µ C ( V – V Tn ) ( V GS – V Tn )
L n ox GS 3
ω T ≈ --------- = ---------------------------------------------------- = --- µ n ------------------------------ L
C gs 2 2 L
--- WLC ox
3

■ The transition time is the inverse of ωT and can be written as the average time for
electrons to drift from source to drain
L L
τ T = ---------------------------------------------- = ----------
2 ( V GS – V Tn ) v dr
µ n --- ------------------------------
3 L

velocity saturation causes τT to decrease linearly with L; however,submicron

MOSFETs have transition frequencies that are approaching those for oxide-
isolated BJTs

EECS 105 Fall 1998

Lecture 28
 Frequency Response of Voltage Amplifiers

■ Common-emitter amplifier:

V+

iSUP

iOUT

RS
+
+ RL vOUT
V
− s
−
+
V
− BIAS

V−

Procedure: substitute small-signal model and perform phasor analysis

EECS 105 Fall 1998

Lecture 28
 Brute Force Phasor Analysis

■ “Exact” analysis: transform into Norton form at input to facilitate nodal analysis

RS Cµ

+
+
Vs + Vπ gmVπ ro roc RL Vout
rπ Cπ
−
−
−

(a)

Cµ

+
+
Is R'in Vπ Cπ gmVπ R'out Vout

−
−

R'in = RS rπ R'out = ro roc RL

(b)

Note that Ccs is omitted, along with rb

Details: see Section 10.4
Factor (approximately!) into one high-frequency zero and two poles ω1 << ω2
–1
ω 1 = ( r π R S ) ( C π + ( 1 + g m r o r oc R L )C µ ) + ( r o r oc R L )C µ

( r o r oc R L ) ( r π R S )C µ C π
–1
ω 2 = -------------------------------------------------------------------------------------------------------------------------------------------------
( r π R S ) ( C π + ( 1 + g m r o r oc R L )C µ ) + ( r o r oc R L )C µ

EECS 105 Fall 1998

Lecture 28
 The Miller Approximation

■ The “exact” analysis is not particularly helpful for gaining insight into the
frequency response ... consider the effect of Cµ on the input only

It Cµ

+
Vt gmVt R'out = ro roc RL Vout
−

neglect the feedforward current Iµ in comparison with gm Vπ ... a good

approximation
It = (Vt - Vo) / Zµ
Vo = - gmVt RL / (RL + Rout’) = AvCµ Vt
where AvCµ is the low frequency voltage gain across Cµ
It = Vt (1- Av) / Zµ
Zeff = Vt / It = Zµ / (1 - Av)

Z eff = -------------  --------------------- = -------------------------------------------- = ---------------

1 1 1 1
jωC µ  1 – A vC  jω ( C µ ( 1 – A vC ) ) jωC M
µ µ

C M = ( 1 – A vC )C µ is the Miller capacitor

µ

EECS 105 Fall 1998

Lecture 28
 Generalized Miller Approximation

■ An impedance Z connected across an amplifier with voltage gain AvZ can

be replaced by an impedance to ground ... multiplied by (1-AvZ)

+ Avo Avo
+ + +
Vi Vo Vi Zeff Vo
− − − −

Zeff = Z /(1 − Avo)

■ Common-emitter and common-source:

AvZ = large and negative for Cµ or Cgd --> capacitance at the input is magnified
■ Common-collector and common-drain:
A vZ ≈ 1 --> capacitance at the input due to Cπ or Cgs is greatly reduced

EECS 105 Fall 1998

Lecture 28
 Voltage Gain vs. Frequency for CE Amplifier
Using the Miller Approximation

■ The Miller capacitance is lumped together with Cπ, which results in a single-
pole low-pass RC filter at the input

RS

+
+
Vs + Vπ Cπ CM gmVπ R'out Vout
rπ
−
−
−
CM = Cµ(1 + gm R'out)

Transfer function has one pole and no zero after Miller approximation:
–1
ω 3dB = ( r π R S ) ( C π + C M )

–1
ω 3dB = ( r π R S ) [ C π + ( 1 + g m r o r oc R L )C µ ]

–1 –1
ω 3dB ≈ ω 1 from the exact analysis (final term R out ′C µ is missing)

EECS 105 Fall 1998

Lecture 28
EECS 105 Fall 1998
Lecture 28