Linear
v
o
(t)
v
i
(t)
August 2008 2006 by Fabian Kung Wai Lee 8
SmallSignal Amplifier (SSA)
All amplifiers are inherently nonlinear.
However when the input signal is small, the input and output
relationship of the amplifier is approximately linear.
This linear relationship applies also to current and power.
An amplifier that fulfills these conditions: (1) smallsignal operation (2)
linear, is called SmallSignal Amplifier (SSA). SSA will be our focus.
If a SSA amplifier contains BJT and FET, these components can be
replaced by their respective smallsignal model, for instance the
hybridPi model for BJT.
( ) ( ) ( ) ( ) ( ) t v a T O H t v a t v a t v a t v
i i i i o 1
3
3
2
2 1
. . . + + + =
When v
i
(t)0 (< 2.6mV)
( ) ( ) t v a t v
i o 1
(1.1)
Linear relation
5
August 2008 2006 by Fabian Kung Wai Lee 9
Example 1.1  An RF Amplifier
Schematic (1)
Z
L
V
s
Z
s Amplifier
Input
Matching
Network
Output
Matching
Network
DC supply
L
L3
R=
L=100.0 nH
L
L2
R=
L=100.0 nH
R
RD1
R=100 Ohm
C
CD1
C=0.1 uF
C
CD2
C=100 pF
Port
VCC
Num=3
L
L4
R=
L=12.0 nH
C
C2
C=0.68 pF
C
Cc1
C=100.0 pF
R
RB2
R=1.5 kOhm
R
RB1
R=1 kOhm
L
LC
R=
L=100.0 nH
R
RC
R=470 Ohm
Port
Input
Num=1
Port
Output
Num=2
L
L1
R=
L=4.7 nH
C
C1
C=3.3 pF
C
Cc2
C=100.0 pF
pb_phl _BFR92A_19921214
Q1
RF power flow
August 2008 2006 by Fabian Kung Wai Lee 10
Example 1.1 Cont
Under AC and smallsignal conditions, the BJT can be replaced with
linear hybridpi model:
BJT_NPN
BJT1
BJT_NPN
BJT1
HybridPi equivalent
circuit for BJT
R
rbpc
L
LC
R=
L=100.0 nH
L
L4
R=
L=12.0 nH
C
C2
C=0.68 pF
Port
Output
Num=2
C
Cc2
C=100.0 pF
L
Lepkg
L
Lcpkg
L
Lbpkg
R
rbbp
I_DC
gmvbpe
R
rce
R
rbpe
C
Cc
C
Ce
R
RB1
R=1kOhm
R
RB2
R=1.5 kOhm
L
L2
R=
L=100.0 nH
L
L3
R=
L=100.0 nH
C
Cc1
C=100.0 pF
R
RC
R=470 Ohm
Port
Input
Num=1
L
L1
R=
L=4.7nH
C
C1
C=3.3pF
6
August 2008 2006 by Fabian Kung Wai Lee 11
Typical RF Amplifier Characteristics
To determine the performance of an amplifier, the following
characteristics are typically observed.
1. Power Gain.
2. Bandwidth (operating frequency range).
3. Noise Figure.
4. Phase response.
5. Gain compression.
6. Dynamic range.
7. Harmonic distortion.
8. Intermodulation distortion.
9. Third order intercept point (TOI).
Important to smallsignal
amplifier
Important parameters of
largesignal amplifier
(Related to Linearity)
Will elaborate in HighPower Circuits
August 2008 2006 by Fabian Kung Wai Lee 12
Power Gain
For amplifiers functioning at RF and microwave frequencies, usually
of interest is the input and output power relation.
The ratio of output power over input power is called the Power Gain
(G), usually expressed in dB.
There are a number of definitions for power gain as we will see
shortly.
Furthermore G is a function of frequency and the input signal level.
dB log 10
Power Input
Power Output
10

\

= G
Power Gain
(1.2)
7
August 2008 2006 by Fabian Kung Wai Lee 13
Why Power Gain for RF and Microwave
Circuits? (1)
Power gain is preferred for high frequency amplifiers as the
impedance encountered is usually low (due to presence of parasitic
capacitance).
For instance if the amplifier is required to drive 50 load the voltage
across the load may be small, although the corresponding current
may be large (there is current gain).
For amplifiers functioning at lower frequency (such as IF frequency), it
is the voltage gain that is of interest, since impedance encountered is
usually higher (less parasitic).
For instance if the output of an IF amplifier drives the demodulator
circuits, which are usually digital systems, the impedance looking into
the digital system is high and large voltage can developed across it.
Thus working with voltage gain is more convenient.
Power = Voltage x Current
August 2008 2006 by Fabian Kung Wai Lee 14
Why Power Gain for RF and Microwave
Circuits? (2)
Instead of focusing on voltage or current gain, RF engineers focus on
power gain.
By working with power gain, the RF designer is free from the
constraint of system impedance. For instance in the simple receiver
block diagram below, each block contribute some power gain. A
large voltage signal can be obtained from the output of the final block
by attaching a high impedance load to its output.
LO
IF Amp.
BPF
LNA
BPF
RF Portion
(900 MHz)
IF Portion
(45 MHz)
RF signal
power
1 W
15 W
IF signal
power
75 W
7.5 mW
400
t
v(t)
4.90 V
R
V
average
P
2
2
=
8
August 2008 2006 by Fabian Kung Wai Lee 15
Derivation of Input and Output Power
Relationship for SmallSignal Operation
R
L
V
s
Z
s
Z
in
= R
( ) ( ) t v a t v
i o 1
( )
( ) ( ) ( )
( ) dBm P a dBm P
P a P
P a P
P a P
v a v
in o
in o
in o
in o
i
R
o
R
+
+
2
1
2
1
2
1
2
1
2
2
1 2
1
2
2
1
log 10
mW 1 / log 10 log 10 mW 1 / log 10
mW 1 / mW 1 /
= R
Assume that the input impedance
to the amplifier is transformed to
R at the operating frequency
For smallsignal
operation
P
o
dBm
P
in
dBm
10log(a
1
2
)
Slope of 1 Unit
Usually we express
power in logarithmic
scale, i.e. the dB
or dBm scale.
Here the relation
between input and output
power is in dB.
Power gain
P
in
P
o
August 2008 2006 by Fabian Kung Wai Lee 16
Harmonic Distortion (1)
Z
L
V
s
Z
s
When the input driving signal is
small, the amplifier is linear.
Harmonic components are
almost nonexistent.
f
f
1
harmonics
f
f
1
2f
1
3f
1
4f
1
0
Smallsignal
operation
region
P
out
P
in
9
August 2008 2006 by Fabian Kung Wai Lee 17
Harmonic Distortion (2)
Z
L
V
s
Z
s
When the input driving signal is
too large, the amplifier becomes
nonlinear. Harmonics are
introduced at the output.
Harmonics generation reduces the gain
of the amplifier, as some of the output
power at the fundamental frequency is
shifted to higher harmonics. This result in
gain compression seen earlier!
f
f
1
f
harmonics
f
1
2f
1
3f
1
4f
1
0
P
out
P
in
August 2008 2006 by Fabian Kung Wai Lee 18
Power Gain, Dynamic Range and Gain
Compression
Dynamic range (DR)
Input 1dB compression
Point (P
in_1dB
)
Saturation
Device
Burn
out
Ideal amplifier
1dB
Gain compression
occurs here
Noise Floor
70 60 50 40 30 20 10 0 10
P
in
(dBm)
20
P
out
(dBm)
20
10
0
10
20
30
30
40
50
60
Power gain G
p
=
P
out
(dBm)  P
in
(dBm)
= 30(43) = 13dB
Linear Region
Nonlinear
Region
P
in
P
out
Input and output at same frequency
10
August 2008 2006 by Fabian Kung Wai Lee 19
Bandwidth
Power gain G versus frequency for smallsignal amplifier.
f / Hz 0
G/dB
3 dB
Bandwidth
P
o
dBm
P
i
dBm
P
o
dBm
P
i
dBm
August 2008 2006 by Fabian Kung Wai Lee 20
Intermodulation Distortion (IMD)
( ) ( ) ( ) ( ) . . .
3
3
2
2 1
T O H t v a t v a t v a t v
i i i o
+ + + =
( ) t v
i
( ) t v
o
ignored
f
f
1
f
2
V
i

These are unwanted components, caused by
the term
3
v
i
3
(t), which falls in the operating
bandwidth of the amplifier.
f
f
1
f
2
2f
1
f
2
2f
2
f
1
f
2
f
1
2f
1
f
1+
f
2
2f
2
3f
1
3f
2
2f
1
+f
2
2f
2
+f
1
Vo
IMD
Operating bandwidth
of the amplifier
More will be said about
this later in largesignal
amplifier design
Two signals v
1
, v
2
with similar
amplitude, frequencies f
1
and f
2
near each other
Usually specified
in dB
11
August 2008 2006 by Fabian Kung Wai Lee 21
Noise Figure (F)
V
s
The amplifier also introduces noise into the output in
addition to the noise from the environment.
Assuming smallsignal operation.
Noise Figure (F)= SNR
in
/SNR
out
Since SNR
in
is always larger
than SNR
out
, F > 1 for an
amplifier which contribute noise.
More will be said about this
later in smallsignal amplifier
design.
SNR:
Signal to Noise
Ratio
Larger SNR
in
Smaller SNR
out
Z
s
Z
L
August 2008 2006 by Fabian Kung Wai Lee 22
Phase Response (1)
Phase consideration is important for amplifier working with wideband
signals.
For a signal to be amplified with no distortion, 2 requirements are
needed (from linear systems theory).
1. The magnitude of the power gain transfer function must be a
constant with respect to frequency f.
2. The phase of the power gain transfer function must be a linear
function of f.
A 2Port
Network
H()
V
1
() V
2
()
Z
L
( )
( )
( )
( )
( )
j
V
V
e H H = =
1
2
Transfer function
H()
Magnitude response
(related to power
gain)
()
Phase response
Bode
Plot
12
August 2008 2006 by Fabian Kung Wai Lee 23
Phase Response (2)
A linear phase produces a constant time delay for all signal
frequencies, and a nonlinear phase shift produces different time delay
for different frequencies.
Property (1) means that all frequency components will be amplified by
similar amount, property (2) implies that all frequency components will
be delayed by similar amount.
()
Linear
phase response
Z
L
A 2Port
Network
H()
V
1
(t) V
2
(t)
t
V
1
(t)
()
Nonlinear
phase response
t
V
2
(t)
t
V
2
(t)
August 2008 2006 by Fabian Kung Wai Lee 24
2.0 SmallSignal Amplifier
Power Gain Expressions
13
August 2008 2006 by Fabian Kung Wai Lee 25
The Theory of Maximum Power
Transfer (1)
Z
s
Z
L
V
s
I
L
V
L
L L L
s s s
jX R Z
jX R Z
+ =
+ =
{ }
*
2
1
Re
L L L
I V P =
Time averaged power dissipated across
load Z
L
:
L s
s
L s
L s
Z Z
V
L
Z Z
Z V
L
I V
+ +
= =
( ) { } { }
( ) ( )
2 2
2
2
2
2
1
2
1
*
2
1
Re Re
L s L s
L s
L s
L s
L s
s
L s
L s
X X R R
R V
L
Z Z
Z V
Z Z
V
Z Z
Z V
L
P
P
+ + +
+
+ +
=
= =
( )
L L L L
X R P P , =
0 = =
L
L
L
L
X
P
R
P
Letting
We find that the value for R
L
and X
L
that would maximize P
L
is
R
L
= R
s
, X
L
= X
s
.
In other words: Z
L
= Z
s
*
To maximize power transfer to the load
impedance, Z
L
must be the complex
conjugate of Z
s
, a notion known as
Conjugate Matched.
where
Basic sourceload network
August 2008 2006 by Fabian Kung Wai Lee 26
The Theory of Maximum Power
Transfer (2)
Z
s
Z
L
= Z
s
*
V
s
P
L
= P
A
( ) A
s
R
s
V
L
P P = =
8
2
max
Under conjugate match condition:
Z
s
Z
L
Z
s
*
V
s
P
L
Under nonconjugate match condition:
Reflect
8
2
P P P
A
s
R
s
V
L
= <
P
A
P
reflect
We can consider the load power P
L
to
consist of the available power P
A
minus
the reflected power P
reflect
.
Available
Power

\

=
2
1
L A L
P P or
L
14
August 2008 2006 by Fabian Kung Wai Lee 27
Power Components in an Amplifier
Z
L
V
s
Z
s Amplifier
V
s
Z
s
Z
L
Z
1
Z
2
V
Amp
+

P
Ao
P
L
P
Ro
P
As
P
Rs
P
in
2 basic source
load networks
Approximate
Linear circuit
August 2008 2006 by Fabian Kung Wai Lee 28
Power Gain Definition
From the power components, 3 types of power gain can be defined.
G
P
, G
A
and G
T
can be expressed as the Sparameters of the amplifier
and the reflection coefficients of the source and load networks. Refer
to Appendix 1 for the derivation.
As
L
T
As
Ao
A
in
L
p
P
P
G
P
P
G
P
P
G
= =
= =
= =
power Input Available
load to delivered Power
Gain Transducer
power Input Available
Power load Available
Gain Power Available
Amp. power to Input
load to delivered Power
Gain Power
The effective power gain
(2.1a)
(2.1b)
(2.1c)
15
August 2008 2006 by Fabian Kung Wai Lee 29
Naming Convention
Z
L
V
s
Z
s Amplifier
2  port
Network
1
2
Source
Network
Load
Network
L
(
22 21
12 11
s s
s s
In the spirit of high
frequency circuit design,
where frequency response
of amplifier is characterized
by Sparameters and
reflection coefficient is
used extensively
instead of impedance,
power gain can be expressed
in terms of these parameters.
August 2008 2006 by Fabian Kung Wai Lee 30
Summary of Important Power Gain
Expressions and the Gain Dependency
Diagram
s
s
L
L
s
D s
s
D s
=
=
11
22
2
22
11
1
1
1

\


\

=
2
1
2
22
2 2
21
1 1
1
L
L
P
s
s
G

\


\

=
2
2
2
11
2
21
2
1 1
1
s
s
A
s
s
G
2
1
2
22
2 2
21
2
1 1
1 1
s L
s L
T
s
s
G

\


\

=
Note:
All G
T
, G
P
, G
A
,
1
and
2
depends on the S
parameters.
(2.2a)
21 12 22 11
s s s s D =
(2.2b)
(2.2c)
(2.2d)
(2.2e)
s
L
1
2
G
A
G
P
G
T
(
22 21
12 11
s s
s s
The Gain Dependency Diagram
16
August 2008 2006 by Fabian Kung Wai Lee 31
Transducer Power Gain G
T
(1)
G
T
is the relevant indicator of the amplifying capability of the amplifier.
Whenever we design an amplifier to a specific power gain, we refer to
the transducer power gain G
T
.
G
P
and G
A
are usually used in the process of amplifier synthesis to
meet a certain G
T
.
An amplifier can have a large G
P
or G
A
and yet small G
T
, as illustrated
in the next slide.
August 2008 2006 by Fabian Kung Wai Lee 32
Transducer Power Gain G
T
(2)
Z
s
Z
L
Amplifier
with large
G
P
P
L
P
in
P
As
s
1
s
*
G
T
=
Small
P
L P
As
Z
s
Z
L
Amplifier
with large
G
P
Input
Matching
Network
P
in
P
in
s
1
=
s
*
G
T
=
Large
Note that G
P
remains
unchanged in both
cases.
17
August 2008 2006 by Fabian Kung Wai Lee 33
Essence of SmallSignal Amplifier
Design
In essence, designing a smallsignal amplifier with
transistor or monolithic microwave integrated circuit
(MMIC) implies finding the suitable load and source
impedance to be connected to the output and input
port, so that you get the required transducer power
gain G
T
, bandwidth and other characteristics.
Unilateral Condition (1)
In certain cases, when operating frequency is low, s
12
 0. Such
condition is known as Unilateral.
Under unilateral condition:
From (2.2d):
We see that the Transducer Power Gain G
T
consists of 3 parts that are
independent of each other.
August 2008 2006 by Fabian Kung Wai Lee 34
L o s
L
L
s
s
T
G G G
s
s
s
G =
=
2
22
2
2
21 2
11
2
1
1
1
1
11
22
22 11 11
22
11
1
1 1
s
s
s s s
s
D s
L
L
L
L
=
=
=
(2.3)
G
s
G
o G
L
18
Unilateral Condition (2)
In smallsignal amplifier design for unilateral condition, we can find
suitable source and load impedance for a required G
T
by optimizing G
s
and G
L
independently, and this simplifies the design procedures.
However in most case, s
12
is typically not zero especially at frequency
above 1 GHz. Thus we will not pursue design techniques for unilateral
condition.
August 2008 2006 by Fabian Kung Wai Lee 35
August 2008 2006 by Fabian Kung Wai Lee 36
Example 2.1 Familiarization with the
Gain Expressions
An RF amplifier has the following Sparameters at f
o
: s
11
=0.3<70
o
,
s
21
=3.5<85
o
, s
12
=0.2<10
o
, s
22
=0.4<45
o
. The system is shown below.
Assuming reference impedance (used for measuring the Sparameters)
Z
o
=50, find:
(a) G
T
, G
A
, G
P
.
(b) P
L
, P
A
, P
inc
.
Amplifier
(
22 21
12 11
s s
s s
Z
L
=73
40
5<0
o
19
August 2008 2006 by Fabian Kung Wai Lee 37
Example 2.1 Cont...
Step 1  Find
s
and
L
.
Step 2  Find
1
and
2
.
Step 3  Find G
T
, G
A
, G
P
.
Step 4  Find P
L
, P
A
.
151 . 0 146 . 0
22
1
11
1
j
L
s
L
D s
= =
358 . 0 265 . 0
11
1
22
2
j
s
s
s
D s
= =
111 . 0 = =
+
o s
o s
Z Z
Z Z
s
187 . 0 = =
+
o L
o L
Z Z
Z Z
L
742 . 13
1 1
1
2
1
2
22
2 2
21
=

\


\

=
L
L
P
s
s
G
739 . 14
1 1
1
2
2
2
11
2
21
2
=

\


\

=
s
s
A
s
s
G
562 . 12
1 1
1 1
2
1
2
22
2 2
21
2
=

\


\

=
s L
s L
T
s
s
G
[ ]
W P
s
s
Z
V
A
078 . 0
Re 8
2
= =
W P P
s
s
Z Z
Z Z
A in
0714 . 0 1
2
1
1
= 
\

=
+
W P G P
in P L
9814 . 0 = =
Try to derive
These 2 relations
Again note that this is an
analysis problem.
August 2008 2006 by Fabian Kung Wai Lee 38
3.0 Stability Analysis of
SmallSignal Amplifier
20
August 2008 2006 by Fabian Kung Wai Lee 39
Introduction
An amplifier is a circuit designed to enlarged electrical signals.
When there is no input, there should be no output, this condition is
known as stable.
On the contrary, if the amplifier produces an output when there is no
input, it is unstable. In fact the amplifier becomes an oscillator!
Thus a stability analysis is required to determine whether an amplifier
circuit is stable or not.
Stability analysis is also carried out by assuming a smallsignal
amplifier, since the initial signal that causes oscillation is always very
small.
Stability of an amplifier is affected by the load and source impedance
connected to its two ports.
An unstable or marginally stable amplifier can be made more stable.
August 2008 2006 by Fabian Kung Wai Lee 40
Stability Concept (1)  Perspective of
Oscillation from Wave Propagation
An incident wave V
+
propa
gating towards port 1 will suffer
multiple reflection.
If 
s
1
>1, then the magnitude
of the incident wave towards port 1
will increase indefinitely,
leading to oscillation.
Z
1
or
1
b
s
b
s
1
b
s
1
b
s
1
2
b
s
s
2
1
2
b
s
s
2
1
3
Source
2port
Network
Z
s
or
s
Port 1 Port 2
s
s
s s s s s
b
a
b b b a
=
+ + + =
1
1
2 2
1 1 1
1
...
s
s
s s s s s
b
b
b b b b
=
+ + + =
1
1
1
2 3
1
2
1 1 1
1
...
Only if
1
1
<
s
A geometric series
b
s
s
3
1
3
b
s
s
3
1
4
a
1
b
1
21
August 2008 2006 by Fabian Kung Wai Lee 41
Stability Concept (2)
Thus oscillation will occur when 
s
1
 > 1.
Since the source network is usually passive, 
s
 < 1. Thus for
instability to occur, 
1
 > 1, this condition represents the potential for
oscillation.
Similar argument can be applied to port 2, and we see that the
condition for oscillation is 
2
 > 1.
Since input and output power of a 2port network are related, when
either port is stable, the other will also be stable.
1 always since 1
1
1
1 1
< >
> =
s
s s
1 always since 1
1
2
2 2
< >
> =
L
L L
Port 1:
Port 2:
August 2008 2006 by Fabian Kung Wai Lee 42
Perspective of Oscillation from Circuit
Theory (1)
Source Network Port 1
R
s
jX
s
R
1
jX
1
Z
s
Z
1
V
( )
s
s
s
s s
V
Z Z
Z
V
X X j R R
jX R
V
1
1
1 1
1 1
+
=
+ + +
+
=
(3.1)
22
August 2008 2006 by Fabian Kung Wai Lee 43
Perspective of Oscillation from Circuit
Theory (2)
Using Laplace transformation:
( )
( )
( ) ( )
( ) s V
s Z s Z
s Z
s V
s
s 1
1
+
=
For a system to oscillate, the denominator of the equation must
has a conjugate poles at the frequency
o
where oscillation occurs,
this means:
( ) ( ) 0 
1
= +
=
o
j s s
s Z s Z
0 
1
= +
o
R R
s
0 
1
= +
o
X X
s
As will be shown, this requirement is similar to 
s
1
=1
This fact will be used in
later chapter to create
oscillating circuit.
(3.2a)
(3.2b)
(3.2c)
August 2008 2006 by Fabian Kung Wai Lee 44
Similarity Between Both Perspectives
(1)
From: ( )
( ) jX Z R
jX Z R
o
o
+ +
+
=
( )
( )
( )
( )
( ) ( ) ( )
( ) ( ) ( )
s o s s s o s o s
s o s s s o s o s
s o s
s o s
o
o
s
X X Z X R X R j X X Z R R Z R R
X X Z X R X R j X X Z R R Z R R
jX Z R
jX Z R
jX Z R
jX Z R
o
+ + + + + + +
+ + + + +
=
+ +
+
+ +
+
=
1 1 1 1
2
1 1
1 1 1 1
2
1 1
1 1
1 1
1
( ) ( ) ( ) ( )
( ) ( ) ( ) ( )
2
1 1 1
2
1
2
1 1
2
1 1 1
2
1
2
1 1
1
s o s s s o s o s
s o s s s o s o s
s
X X Z X R X R X X Z R R Z R R
X X Z X R X R X X Z R R Z R R
o
+ + + + + + +
+ + + + +
=
23
August 2008 2006 by Fabian Kung Wai Lee 45
Similarity Between Both Perspectives
(2)
We see that when R
1
+R
s

o
= 0, X
1
+X
s

o
=0 :

1
s

o
= 1
And when R
1
+R
s

o
<0, X
1
+X
s

o
= 0 we have 
1
s

o
> 1
The above requirement also applies to port 2, in which case when
R
2
+R
L

o
=0, X
2
+X
L

o
=0

2
L

o
= 1
And when R
2
+R
L

o
<0, X
2
+X
L

o
= 0 we have 
2
L

o
> 1
August 2008 2006 by Fabian Kung Wai Lee 46
Important Summary on Oscillation
Assuming that 
s
 and 
L
 always < 1 for passive components, we
conclude that:
For a 2port network to be stable, it is necessary that the load and
source impedance are chosen in such a way that 
1
 < 1and 
2
 <
1.
To prevent oscillation:
( )
( ) 1
1
2
1
<
<
1
2
Source
Network
Load
Network
s
L
(
22 21
12 11
s s
s s
As seen previously, for stability 
1
 < 1 and 
2
 < 1.
Using (2.2a) this can be expressed as:
1
1
1
1
11
21 12
22 2
22
21 12
11 1
<
+ =
<
+ =
s
s
L
L
s
s s
s
s
s s
s
(3.3a)
(3.3b)
August 2008 2006 by Fabian Kung Wai Lee 48
3.1 Stability Circles
25
August 2008 2006 by Fabian Kung Wai Lee 49
Load Stability Circle (LSC)
Setting 
1
 =1, we can determine all the corresponding values of
L
.
The
L
happens to fall on the locus of a circle on Smith Chart.
( )
2 2
22
21 12
2 2
22
*
*
11 22
22
21 12
11
1
1
D s
s s
D s
Ds s
s
s s
s
L
L
L
+
L L L
R C =
Load stability circle
Center of circle
Radius of circle
(3.4)
For detailed derivation,
See [2], Chapter 10
or [1], Chapter 11
Re
L
Im
L
R
L
C
L
0
August 2008 2006 by Fabian Kung Wai Lee 50
Source Stability Circle (SSC)
Setting 
2
 =1, we can determine all the corresponding values of
s
.
The
s
also happens to fall on the locus of a circle on Smith Chart.
( )
2 2
11
21 12
2 2
11
*
*
22 11
22
21 12
22
1
1
D s
s s
D s
Ds s
s
s s
s
s
s
s
+
s s s
R C = Source stability circle
Center of circle
Radius of circle
(3.5)
For detailed derivation,
See [2], chapter 10
or [1], Chapter 11
Re
s
Im
s
R
s
C
s
0
26
August 2008 2006 by Fabian Kung Wai Lee 51
Stable Regions (1)
The source and load stability circles only indicate the value of
s
and
L
where 
2
 = 1 and 
1
 = 1. We need more information to show the
stable regions for
s
and
L
plane on the Smith Chart.
For example for LSC, when
L
= 0, 
1
 = s
11
 (See (2.2a)).
Assume LSC does not encircle s
11
= 0 point. If s
11
 < 1 then
L
= 0 is a
stable point, else if s
11
 > 1 then
L
=0 is an unstable point.
LSC
s
11
<1
Stable
Region
LSC
s
11
>1
L
plane
August 2008 2006 by Fabian Kung Wai Lee 52
Stable Region (2)
Now let the LSC encircles s
11
= 0 point. Similarly if s
11
 < 1 then
L
= 0
is an stable point, else if s
11
 > 1 then
L
= 0 is an unstable point.
This argument can also be applied for SSC, where we consider s
22

instead and the Smith Chart corresponds to
s
plane.
LSC
s
11
<1
LSC
s
11
>1
Stable
Region
L
plane
27
August 2008 2006 by Fabian Kung Wai Lee 53
Summary for Stability Regions
For both Source and Load reflection coefficients (
s
and
L
) :
LSC or SSC
s
11
 or s
22
 <1
LSC or SSC
s
11
 or s
22
 >1
LSC or SSC
s
11
 or s
22
 <1
LSC or SSC
s
11
 or s
22
 >1
Stability circles does not
enclose origin
Stability circles
enclose origin
Stable
Region
August 2008 2006 by Fabian Kung Wai Lee 54
Unconditionally Stable Amplifier
There are times when the amplifier is stable for all passive source and
load impedance.
In this case the amplifier is said to be unconditionally stable.
Assuming s
11
 < 1 and s
22
 < 1, the stability region would look like this:
LSC
s
11
<1
L
can
occupy any
point in the
Smith chart
SSC
s
22
<1
s
can
occupy any
point in the
Smith chart
28
August 2008 2006 by Fabian Kung Wai Lee 55
Example 3.1
Use the Sparameters of the amplifier in Example 2.1, draw the load
and source stability circles and find the stability region.
SSC LSC
Hint:
Apply equations
(3.4) and (3.5) to
find the center
and radius of
the circles.
August 2008 2006 by Fabian Kung Wai Lee 56
Example 3.1 Cont
Macro to convert complex number in polar form to rectangular form
Polar R theta , ( ) R cos
theta
180

\


i R sin
theta
180

\


+ :=
Definition of Sparameters:
S11 Polar 0.3 70 , ( ) :=
S12 Polar 0.2 10 , ( ) :=
S21 Polar 3.5 85 , ( ) :=
S22 Polar 0.4 45 , ( ) :=
S11 0.1026 0.2819i = S12 0.197 0.0347i =
S21 0.305 3.4867i + = S22 0.2828 0.2828i =
D S11 S22 S12 S21 :=
D 0.2319 0.7849i =
Computation
Using the
Software
MATHCAD
29
August 2008 2006 by Fabian Kung Wai Lee 57
Example 3.1 Cont
Finding the Load Stability Circle parameters
RL
S12 S21
S22 ( )
2
D ( )
2
:=
CL
S22 D S11
( )
S22 ( )
2
D ( )
2
:=
CL 0.1674 0.2686i = RL 1.373 =
Finding the Source Stability Circle parameters
CS
S11 D S22
( )
S11 ( )
2
D ( )
2
:= RS
S12 S21
S11 ( )
2
D ( )
2
:=
CS 0.0928 9.8036i 10
3
+ = RS 0.3124 1.1661i + =
August 2008 2006 by Fabian Kung Wai Lee 58
3.2 Test for Unconditional
Stability The Stability Factor
30
August 2008 2006 by Fabian Kung Wai Lee 59
Stability Factor (1)
Sometimes it is not convenient to plot the stability circles, or we just
want a quick check whether an amplifier is unconditionally stable or
not.
In such condition we can compute the Stability Factor of the amplifier.
Rollette* has come up with a factor, the K factor that tell us whether or
not an amplifier (or any linear 2port network) is unconditionally stable
based on its Sparameters at a certain frequency.
A complete derivation can be found in reference [1], [4], [5], here only
the result is shown.
*Rollett J., Stability and power gain invariants of linear twoports,
IRE Transactions on Circuit Theory, 1962, CT9, p. 2932.
August 2008 2006 by Fabian Kung Wai Lee 60
Stability Factor (2)
The condition for a 2port network to be unconditionally stable is:
Otherwise the amplifier is conditional stable or unstable at all (it is an
oscillator !).
K is also known as the Rollette Stability Factor.
1
1
2
1
21 12
2 2
22
2
11
<
>
+
=
D
s s
D s s
K
(3.6)
Note that the K factor only tells us if an amplifier (or any linear 2port network)
is unconditionally stable. It doesnt indicate the relative stability of 2 amplifiers
which fail the test. A newer test, called the factor allows comparison of
2 conditionally stable amplifiers (See Appendix 2).
31
August 2008 2006 by Fabian Kung Wai Lee 61
What if Amplifier is Unstable, or Stable
Region is too Small?
Use negative feedback to reduce amplifier gain.
Redesign d.c. biasing, find new operating point (or Q point) that will
result in more stable amplifier.
Add some resistive loss to the circuit to improve stability.
Use a new component with better stability.
August 2008 2006 by Fabian Kung Wai Lee 62
3.3 Stabilization of Amplifier
32
August 2008 2006 by Fabian Kung Wai Lee 63
Stabilization Methods (1)

1
 > 1 and 
2
 > 1 can be written in terms of input and output
impedances:
This implies that Re[Z
1
] < 0 or Re[Z
2
] < 0.
Thus one way to stabilize an amplifier is to add a series resistance or
shunt conductance to the port. This should made the real part of the
impedance become positive.
In other words we deliberately add loss to the network.
( )
( )
1 1
1 1
1
jX
o
Z R
jX
o
Z R
+ +
+
=
1 and 1
2
2
2
1
1
1
>
+
= >
+
=
o
o
o
o
Z Z
Z Z
Z Z
Z Z
( )
( )
2
1
2
1
2
1
2
1
1
X
o
Z R
X
o
Z R
+ +
+
=
August 2008 2006 by Fabian Kung Wai Lee 64
Stabilization Methods (2)
2  port
Network
Z
1
Source
Network
Load
Network
Z
1
+R
1
22 21
12 11
s s
s s
R
1
R
2
Z
2
Z
1
+R
1
2  port
Network
Y
1
Source
Network
Load
Network
Y
1
+G
1
22 21
12 11
s s
s s
G
1
G
2
Y
2 Y
2
+G
2
33
August 2008 2006 by Fabian Kung Wai Lee 65
Effect of Adding Series Resistance on
Smith Chart
Suppose we have an impedance Z
L
and a load stability circle (LSC).
Assuming the LSC touches the R=10 circle. Thus by inserting a series
resistance of 10, we can limit Z
L
to the stable region on the Smith
Chart.
Z =
Z
L
+ 10
10
Z
L
LSC
R = 10
circle
Stable
Region
Unstable
Region
August 2008 2006 by Fabian Kung Wai Lee 66
Effect of Adding Shunt Resistance on
Smith Chart
Suppose we have an admittance Y
L
and a load stability circle (LSC).
Assuming the LSC touches the G=0.002 circle. Thus by inserting a
shunt resistance of 500, we can limit Y
L
to the stable region on the
Smith Chart.
Y =
Y
L
+ 1/500 500
Y
L
LSC
G = 0.002
circle
Unstable
Region
Stable
Region
34
August 2008 2006 by Fabian Kung Wai Lee 67
Summary for Stability Check of Single
Stage Amplifier
Set frequency range
and design d.c. biasing
Get Sparameters within
frequency range
K factor > 1
and D < 1 ?
Amplifier Unconditionally
Stable
Yes
Draw SSC and LSC
No
Find s
11
 and s
22

Circles intersect
Smith Chart ?
Amplifier is
conditionally stable,
find stable regions
Yes
No
Amplifier is
not stable
Start
End
Perform
stabilization.
Redesign
d.c. biasing.
Choose a new
transistor.
August 2008 2006 by Fabian Kung Wai Lee 68
Chapter Summary
Here we have learnt the important concepts of smallsignal
amplifier and amplifier characteristics.
Here we have derived the three types of power gain expressions
for an amplifier using Sparameters.
We also studied how the various gains depend on either
s
and
L
(the dependency diagram).
We have looked at the concept of oscillation and how it applies
to stability analysis.
Learnt about stability circles and how to find the stable region for
source and load impedance.
Learnt about the Rolette Stability Factor test (K) for
unconditionally stable amplifier.
Learnt about elementary stabilization methods.
35
August 2008 2006 by Fabian Kung Wai Lee 69
Example 3.2  SParameters Measurement
and Stability Analysis Using ADS Software
DC
DC1
DC
S_Param
SP1
Step=1.0 MHz
Stop=1.0 GHz
Start=50.0 MHz
SPARAMETERS
C
Cc2
C=470.0 pF
C
Cc1
C=470.0 pF Term
Term1
Z=50 Ohm
Num=1
Term
Term2
Z=50 Ohm
Num=2
L
Lb2
R=
L=330.0 nH
L
Lb1
R=
L=330.0 nH
L
Lc
R=
L=330.0 nH
R
Rb1
R=10.0 kOhm
R
Rb2
R=4.7 kOhm
C
Ce
C=470.0 pF
R
Re
R=100 Ohm
pb_phl_BFR92A_19921214
Q1
V_DC
SRC1
Vdc=5.0 V
August 2008 2006 by Fabian Kung Wai Lee 70
Example 3.2 Stability Test
m1
freq=600.0MHz
K=0.956
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.0 1.0
0.6
0.4
0.2
0.0
0.2
0.4
0.6
0.8
1.0
0.8
1.2
freq, GHz
K
m1
D
Plotting K and D versus frequency
(from 50MHz to 1.0GHz):
This is the frequency
we are interested in
Amplifier is
conditionally
stable
36
August 2008 2006 by Fabian Kung Wai Lee 71
Example 3.2  Viewing S
11
and S
22
at
f=600MHz and Plotting Stability Circles
indep(SSC) (0.000 to 51.000)
S
S
C
indep(LSC) (0.000 to 51.000)
L
S
C
Since s
11
 < 1 @ 600MHz
Since s
22
 < 1 @ 600MHz
freq
600.0MHz
S(1,1)
0.263 / 114.092
S(2,2)
0.491 / 20.095
August 2008 2006 by Fabian Kung Wai Lee 72
Appendix 1 Derivation of
SmallSignal Amplifier
Power Gain Expressions
37
August 2008 2006 by Fabian Kung Wai Lee 73
Derivation of Amplifier Gain
Expressions in terms of Sparameters
When the electrical signals in the amplifier is small, the active component (BJT)
can be considered as linear.
Thus the amplifier is a linear 2port network, Sparameters can be obtained and
it is modeled by an S matrix.
The preceding gains G
P
, G
A
, G
T
can be expressed in terms of the
s
,
L
and
s
11
, s
12
, s
21
, s
22
.
(
22 21
12 11
s s
s s
Input Output
Vcc
Rb1
Rb2
Re
Cc1
Cc2
L1
L2
L3
Cd
L4
August 2008 2006 by Fabian Kung Wai Lee 74
Derivation of Gain Expression  1
Source
2port
Network
b
s
b
1
a
1
b
s
b
s
1
b
s
1
b
s
1
2
b
s
s
2
1
2
b
s
s
2
1
3
s
s
s s s s s
b
b
b b b b
=
+ + + =
1
1
1
2 3
1
2
1 1 1
1
...
1
1
1
=
s
s
b
a
s s
b b a + =
1 1
or
s
s
s s s s s
b
a
b b b a
=
+ + + =
1
1
2 2
1 1 1
1
...
This derivation is largely based
on the book [5].
38
August 2008 2006 by Fabian Kung Wai Lee 75
Derivation of Gain Expression  2
2  port
Network
1
2
Source
Network
Load
Network
s
L
(
22 21
12 11
s s
s s
b
1
a
1
b
2
a
2
From:
2 22 1 21 2
2 12 1 11 1
a s a s b
a s a s b
+ =
+ =
1 1
2 2
b a
b a
s
L
=
=
and
One obtains:
s
s
s
D s
a
b
= =
11
22
2
2
2
1
L
s
a s
b
=
22
1 21
2
1
s
s
a s
b
=
11
2 12
1
1
L
L
s
D s
a
b
= =
22
11
1
1
1
1
In a similar manner we
can also obtains:
21 12 22 11
s s s s D =
August 2008 2006 by Fabian Kung Wai Lee 76
Derivation of Gain Expression  3
Finding Transducer Power Gain G
T
:
( )
2 2
2
2
1
1
L L
b P = ( )
*
2
1
2
1
2
1
1
1
=
=
s
a P
A
*
1
1
1
1
=
=
s
s
s
b
a
2
1
2
2
1
1
=
s
A
b
P
( )( )
( )( )
2 2
2
2
2
*
2
1
2
2
2
2
1 1
1 1
Power Soure Available
Power Load
1
s L
s
T
L
s
A
L
T
b
b
G
b
b
P
P
G
s
=
= = =
=
From slide DGE1
Only for
available gain
39
August 2008 2006 by Fabian Kung Wai Lee 77
Derivation of Gain Expression  4
Finding Transducer Power Gain G
T
Cont :
L
s
s
a
b
=
22
21
1
2
1
( )
s s
a b =
1 1
1
( )( )
s L s
s
s
b
b
=
1 22
21 2
1 1
Using
( ) ( )
2
1
2
22
2 2
21
2
1 1
1 1
s L
s L
T
s
s
G
=
( ) ( )
2
2
2
11
2 2
21
2
1 1
1 1
L s
s L
T
s
s
G
=
( )( )
2 2
2
2
2
1 1
s L
s
T
b
b
G =
( ) ( )
( )( )
2
21 12 11 22
2 2
21
2
1 1
1 1
L s s L
s L
T
s s s s
s
G
=
From slide DGE2
From slide DGE1
L
L
s
D s
=
22
11
1
1
( )( )
( )( )
11 2
22 1
1 1
1 1
s
s
s L
L s
=
August 2008 2006 by Fabian Kung Wai Lee 78
Derivation of Gain Expression  5
The Available Power Gain G
A
can be obtained from G
T
when
L
=
2
*
Since
Power Source Available
matched) y conjugatel is output (when Power Load Available
=
A
G
( ) ( )
*
2
2
2
11
2 2
21
2
*
2
2
1 1
1 1
L
L
L s
s L
T A
s
s
G G
=
=
= =
( )
( )
2
2
2
11
2
21
2
1 1
1
=
s
s
A
s
s
G
We have
40
August 2008 2006 by Fabian Kung Wai Lee 79
Derivation of Gain Expression  6
( )
( )
2
1
2
22
2 2
21
1 1
1
Power Source
Power Load
= = =
L
L
in
L
P
s
s
P
P
G
( )
2 2
2
2
1
1
L L
b P = ( )
2
1
2
1
2
1
1 = a P
in
Finding Power Gain G
P
:
L
s
s
a
b
=
22
21
1
2
1
From slide DGE2
August 2008 2006 by Fabian Kung Wai Lee 80
Appendix 2 The Stability
Test
41
August 2008 2006 by Fabian Kung Wai Lee 81
The Stability Test
The Roulette Stability Test consist of 2 separate tests (the K and D
values).
This makes it difficult to compare the relative stability of 2
conditionally stable amplifiers.
A later development combines the 2 tests into one, known as the
factor. Larger value indicates greater stability.
For an amplifier to be unconditionally stable, it is necessary that:
( )
1
12 21
*
22 11
2
22
1
1
> =
+
s s s D s
s
Edwards, M. L., and J. H. Sinsky, A new criterion for linear twoport stability using
A single geometrically derived parameter, IEEE Trans. On Microwave Theory and
Techniques, Dec 1992.