Aim of this chapter
EIE209 Basic Electronics Basic Transistor Amplifiers
Contents
- Biasing • Amplification principles
- Small-signal model development for BJT
Basic idea — biasing
Step 1: Set the transistor at a certain DC level Step 2: Inject a small signal to the input and get a bigger output
— coupling
7V
0.6V
amplifier
- –
R L = 0.5kΩ
V BE
) / R
B
= (10 –
0.7 )
/ R
B
I C
= b
I B
= 100 (10 –
0.7 )
/ R
B = 10 mA
So, R B
= 94kΩ
Also,
V CE
= 10 – R
L
I C
Hence, 5 = 10 – 10 R
L So,
= (10 –
I B
Then,
I C
Biasing the transistor
To set the transistor to a certain DC level = To set
V CE and I
C
R
L
R
B
V BE
V CE
I B
V BE ≈ 0.7 V.
Transistor: b = 100
V CC =10V
Suppose we want the following biasing condition:
I C = 10 mA and
V CE = 5 V
Find R
B
and R
L
Start with
- –
b dependent biasing — bad biasing
=10V Now, let’s go to the lab and try using = 94kΩ and
V R R
CC B L = 0.5kΩ, and see if we get what we want.
I C
R
L …totally wrong! We don’t get I = 10mA and V = 5V
R
C CE B
- I
B
V
- CE
This is a bad biasing circuit!
- – V
BE because it relies on the accuracy of , b
- – but can be ±50% different from what is
b
Transistor: given in the databook. b = 100
A slightly better biasing method
Again, our objective is to find the resistors such that = 10mA and = 5V .
I V C CE
=10V
V CC First, if
I is small, we can approximately write
B
I C
R R R
94 L B 2 B1
R
B1 0.6 = 10 ¥
= fi
R R R
6
+
+ B1 B2 B 2
I B
V CE
- Suppose we get = 10mA. Then R = 0.5kΩ.
I
C L- – V
BE
R
B2
We can start with = 940Ω and = 60Ω. Such R R
- – resistors will make sure
I is much smaller than B
the current flowing down R and R , which is
B1 B2 consistent with the assumption.
What we need in practice is to fine tune R or R
B1 B2 such that is exactly 5V.
V CE A much better biasing method — emitter degeneration
Again, our objective is to find the resistors such that = 10mA and = 5V .
I V C CE
=10V
V CC Set V = 2V, say. Then, R = 2V/10mA = 0.2kΩ.
E E
I C
R
L
R
B1
Surely, R = 0.5kΩ in order to get V = 5V.
L CE
- I
B
V CE
Finally, we have = + 0.6. Therefore, if is
V V
I B E B
- – small compared to and , we have
I I
- RB1 RB2
R
B2
R R
74 V E B1 E =
- – R B 2
- I
- – + Use ≈ 0.6 to get .
- –
R
B B 1 10 - V = - I
- – Slope=–1/
- L
- V
- – V
+
- – V
- –
–
- V
- –
- – V
- a smaller = 0.6
- and down.
- – V
- – Typically, when
- m
- – D
- – Hence,
- v
- – increase the gain by
- – Output waveform is anti- phase.
- v
- –
- –
- = v
- could actually be
- – BE
- –
- – characteristic!!
- because DC current must be zero
- – R
- – –
- – +
- coupling capacitor
- –
- – ±20mV
- – +
- – ~ v
- – –
- – in
- – coupling capacitors (large enough so that they become short-circuit at signal frequencies)
- – –
- –
- –
- –
- v
- – in
- v =
- – CE C L
- v
- – where is the BJT’s transconductance g
- Including BJT’s Early effect
- – C
- g D v R r
- – where is the Early resistor of the BJT.
- – o
- – I
- –
- – Similar to BJT, but input resistance is ∞.
- –
- v
- – Assume the coupling caps are large enough to be considered as short-circuit at signal frequency
- – + v
- v
- – ~
- – + v
- – + v
- v
- – ~
- – + v
- – + v
- – ~
- in B1 B1 BE o p
- – – –
- – ~
- in B1 B1 BE o p
- – – –
- –
- 10V
- v
- – +
- –
speaker
10Ω - – The effective gain drops to 200 ¥
- – + v
- v
- 10V
- 1kΩ
- –
- v
- o
- – –
- – –
- v
- –
- 10V
- v
- – speaker 10Ω
- – 5mA
- 10V
- R
- in
- – v
- – DV = DV
- 10V
- v
- v
- v
- – R
- –
- – B C E E
- v
- v
- v
- v
- – r
- v
- (1 + b )R
- v
- v
- i
- g
- g
- g
- v
- 1 r
- –
- g
- g
- g
- – very large
- – + v
- – +
- v
- (1+b)R
- 10V
- This circuit is also called CLASS A R
- B2
- –
o
in second year EC2. - – m
- 10V
- –
- v
- v
- – R
- – speaker 10Ω
- 10V
- –
- v
- v
- – R
- – speaker 10Ω
- –g R
- which does not depend on .
- o
- – v
- – resistance. This will affect the gain.
- Therefore, the gain decreases to
- v
- – o bias p
- – ||
- bias
- – –
- –
- v
- v
- – R
- – + v
- – 50Ω
- – + v
- v
- – v
- – 200Ω
- v g R
- o m L
- o
- 1 + g R
- – m E
- – B2
- g - R R
- 1 g R R
- v
- o
- – v
- – B2
- v
- – g
- but with infinite input
- – Therefore, the FET can be
- R g v r R v v
- – – –
- v
- o
- – Gain =
- –g R v
- – Input resistance = R || (quite large — desirable)
- v
- – + v
- – Gain = 1 Input resistance = R
- –
- (1+b)R
- – R
- v
- v
- – + v
26 Hence, R = 740Ω and R = 260Ω.
NOTE: b is never used in calculation!! Stable (good) biasing
Summary of biasing with emitter degeneration: =10V
V CC Choose V , I and V .
E C CE
I C
R
L
R
B1
B R R
V E L
B
V CE
V V R
BE B
B2R
V E E
Then use
R
B 2 BV
to choose R and R such that I is much smaller the
B1 B2 B current flowing in and .
R R
Terminology
The following are the same: Biasing point Quiescent point Operating point (OP) DC point Alternative view of biasing
I C
V CC
C
R R
V L L R
Load line
R
CC
CE
I V
operating point
BE C CE
V CE
V CC What controls the operating point?
I C
V CC
Load line
I C
R Slope=–1/
L
R
L a bigger
V BE
CE
a smaller
V R
operating point
BE L
CE
V CC V or I controls the OP
BE B CONCLUSION:
R also controls the OP L V What happens if dances up and down?
BE
I C The OP also
V CC
dances up and Load line down along the
I Slope=–1/ C R
R
L L a bigger = 0.65
V BE load line.
V BE
V also moves up
V CE CE
BE
V BE moves a little bit,
V V CE CE
V CC moves a lot! Animation to show amplifier action
Derivation of voltage gain
DV DV
o CE
=
Question: what is ?
DV DV
in BE
V Clearly, Ohm’s law says that CC
V R = V - I fi DV = -R .DI
I CE CC C L CE L C
CR
L ? Then, what relates DI and DV +
C BE D
I V
C CE
g =
Last lecture: transconductance
V BE
V BE
D
V CE
= -
g R m LD
V BE Common-emitter amplifier The one we have just studied is called COMMON-EMITTER amplifier.
V CC SUMMARY:
R
L
I C
R
Small-signal voltage gain B1
= – g R m L
CE That means we can
R v
B2 BE increasing g and/or R . m L
or
R
CE
D v
~
CE ~
V CE
±20mV =
B2 ? v in
B1
R
V
CCI C
v CE
V BE
L
R
How do we inject signal into the amplifier?
= + a DC point
~
V v
A
total signal operating point small signal Small signal
(large signal) or or
DC value ~ ac signal
a or Da
or quiescent point Solution: Add the same biasing DC level
But, it is impossible to find a
V CC
voltage source which is equal
Exactly the
to the exact biasing voltage
same biasing
R
L
R
I B1 C across B-E.
V BE
V +
v
0.621234V, which is
CE
determined by the network ~
R v
RB1, RB2 and the transistor
v
in
±20mV How to apply the exact ?
V BE The wonderful voltage source: capacitor
V CC
R
L The capacitor voltage is
R
I B1 C exactly equal to
V BE
0A
V CE
V B2 BE
V C
V CC to exactly the same biasing
V BE
R
L
R
I B1 C
This is called a
v CE
~
R v
v in
V CC
R
L
R
I B1 C
CE v
o
R v
~ v
The DC bias does not matter!
Can we create a simple circuit just to look at ac signals?
common emitter ~
~ v v amplifier in o
common emitter amplifier ?
?
What is the loading (resistance) seen here?
What is the Thévenin or Norton equivalent circuit seen here?
1
2 Two basic questions: ~
~ Thévenin form
or Norton form
in
in
v
m
A
in
v
m
G
m
A
o
R
r
Small-signal model of BJT: objectives r
o
R
in
r
m
G
o
R
in
To find: r
o
R
in Derivation of the small-signal model Input side:
v v
B E B E
r = =
in
i i
i
/ b
B C
B
r
For small-signal, p
Dv Dv BE
r = =
Di Di / b
B C
b b = =
r = b/g
(Di / Dv ) g
C BE m p m
where is the BJT’s transconductance g
m Derivation of the small-signal model Output side:
V – I R CE CC C L
V CC
~
R v
For small-signal, L CE
~ R
L
I D = -D ¥ C
v i R
g v
m BE = - D ¥ g v R
m BE L
CE
m Derivation of the small-signal model
Output side:
~
r R v
o L CE
V Dv Dv CE CE CC
~ = -Di +
R r g v
L o m BE
R
L
I C
D = -D v i ( R r )
CE C L o
= - ( )
m BE L o
vCE
r
o Recall: r = V /I , where V is typically about 100V. o A C A A very rough approx. is = ∞. r o Initial small-signal model for BJT
Small-signal
“MUST” REMEMBER
BJT parameters:
I I C C g = = m kT / q)
V ( B
C T
b r = p
~ ~
g
~
m r v g v
o CE
r v
m BE p
BE
V A r =
C E V is thermal voltage
T ª 25mV
BJT model
V is Early voltage A typically ~ 100V
Small-signal FET parameters: g
is a semiconductor parameter FET model
K
parameter
l l is the channel length modulation
1
o =
r
2 K I D
m =
~
Initial small-signal model for FET g
GS
G S D v
o
r
~
v
DS
GS ~
v
m
All amplifier configurations using BJT can be likewise constructed using FET.
B C E E g
V CC is ac 0V.
L
R
o
r
p
r
BE ~
v
m
o
in
B2
R
B1
R
V CC
BE
L v
R
Example: common-emitter amplifier
R B1
|| R
B1|| r
R
in
= R
B1
|| R
B1
|| r
p
Total output resistance R
o
= R
L
o
Voltage gain v
o
v
in
= - g
m
( R
L
|| r
o
) ª - g
m
R
Simplified model: Total input resistance
BE
o
in
Complete model for common-emitter amplifier
Complete model: g
m
v
BE ~
r
p
r
o
R
L R B1 || R B1
o
in
BE
g
m
v
BE ~
R
L
|| r
o
R B1|| R B1 ||
r
p
L Alternative model for common-emitter amplifier Output in Thévenin form:
Total input resistance ||
R r
L o
R = R || R || r
in B1 B1 p
v v v
R || R ||
r
Total output resistance
||
~ R = R r o L o
( || ) g R r v
L m o BE
Voltage gain v
o
= - g ( R || r )
m L o
v
in
ª - g R
m L More about common-emitter amplifier
Because the output resistance is quite large (equal to R || r ≈ R ), the
L o L
common-emitter amplifier is a POOR voltage driver. That means, it is not a good idea to use such an amplifier for loads which are smaller than R . This
L makes it not suitable to deliver current to load.
1kΩ, for example ||
R r
L o practically no output!!
v v v
R || R || 10Ω
r
~
( || ) g R r v
L
m o BE
Transconductance g m =
o
1kΩ 10Ω
in
200 v
But the output circuit is:
= (0.2)(1k) = 200 or 46dB
I C / (25mV) = 5/25 = 0.2 A/V Expected gain = g m R L
in
5mA
o
B2
R
B1
R
v BE
1kΩ
Bad idea — wrong use of common-emitter amplifier
10 1000 + 10 = 1.98 Proper use of common-emitter amplifier The load must be much larger than .
R
L
5mA 1kΩ R
B1
Now the output circuit is:
nearly open circuit
10MΩ
v
10MΩ + o
200 v R v
in
v
in
The effective gain is 7
10 ª 200 200 ¥ 7 1000 + 10
How can we use the amplifier in practice?
1kΩ
v BE
R
B1
R
B2
How to connect the output to load?
o
?
in
Emitter follower
Biasing conditions:
Base voltage ≈ 5.6V Emitter voltage ≈ 5V Collector current ≈ 10mA
R
I biased to 10mA B1 C
R = 500Ω
E
: ≈ 44:56 R R
V biased to 5V B1 B2
CESay, = 440kΩ R
B1
= 560kΩ R
B2
B2
v V = V – 0.6
E B
Thus, for small signal, R
o E
E B
= v or v
o in v / v
Gain = = 1 o in
v
B2
|| R
B1
R
E
R
o
r
p
r
BE ~
m
g
o
E
in
B2
R
B1
R
Small-signal model of emitter follower
o
B = v
in i
B r in
= v i n i
B = v
BE
E i
E which is quite large (good)!! Input resistance is
B E i
v
E i
B = r p
E i
B = r p
E i
E /(1 + b ) = r p
B
o
in
p
Small-signal model of emitter follower
B C E E g
m
v
BE ~
r
r
B2
o
R
E
R
B1
|| R
E
r
p
m
v
E
=
1
1 R
E
p
m
=
1
1 R
m
E
E
i
m
v
m
1 g
||
= R
b
m
E
1 R
1
ª
m
E
i
m
E
Small-signal model of emitter follower
B C E E g
m
v
BE ~
r
p
r
o
R
E r out which is quite small (good)!! Output resistance is
r
out
= v
m
i
E
= v
BE
v
m
B
i
m
BE
=
m
i
BE
=
Good for any load Draw no current from previous stage
) r
Large input resistance Small output resistance Voltage gain = 1
very small
o
in
E
p
m
||(1/ g
E
R
in
1 v
Small-signal model of emitter follower
Input resistance is very LARGE
because R = ∞.
E
R
B1 Output resistance is 1/ g . m Gain = 1.
v
output stage. Details to be studied ∞ in
v
1/ g
I
E
Common-emitter amplifier with emitter follower as buffer
v BE
R
B1
R
B2
in
L
o
common-emitter amplifier (high gain) emitter follower (unit gain)
∞
1 g m
I E
FET amplifiers (similar to BJT amplifiers)
v GS
R
G1
R
G2
in
L
o
common-source amplifier (high gain = – g m R L ) source follower (unit gain)
∞
1 g m
I S
Will the biasing resistors affect the gain? Seems not, because
R
L
R
bias
Gain =
m L
R
bias
v
in
However, a realistic voltage source has finite internal
Input source with finite resistance The input has a voltage divider network.
R r
|| bias
R
p L
v = v
BE in
R
bias
R || r R +
bias s p
R
s
o
v R || r
v
in = (- )
g R
m L
v R r R
in p s assuming r very large. o
R
s
R g v r R v v v
bias r m BE o L in BE o p
= 0.845 or - 1.463dB
g
R bias || r
p
50 + R bias || r p = 596|| 500 50 + 596||500) = 169
L
R
m
= 100/0.2 = 500Ω Voltage divider attenuation = Hence, the gain is reduced to 0.845( g
m
= b/g
p
= 5mA/25mV = 0.2A/V r
m
5mA
50Ω 94k||600 = 596Ω
bias
in
BE
p
10V By how much does the gain drop? r
600Ω
o
in
1kΩ 94kΩ
Example
Further thoughts Recall that the best biasing scheme should be b independent.
10V One good scheme is emitter degeneration, i.e.,
5mA
using to fix biasing current directly. Here, R
E
1kΩ since V is about 1.6V, as fixed by the base
84kΩ
B
resistor divider, V is about 1V.
E
o
Therefore, ≈
I V /R = 5mA (no b needed!) +
C E E
in
16kΩ
Question:
Will this biasing scheme affect the gain? Common-emitter amplifier with emitter degeneration
Exercise: Find the small-signal gain of this amplifier.
V CC R
L Answer:
R
B1
= Ê ˆ
v v
1 in
Á ˜
1
Á ˜
v
in
b
Ë ¯
R
R E
m L L ª ª
m E E The gain is MUCH smaller.
We have a good biasing, but a poor gain! Can we improve the gain?
Common-emitter amplifier with emitter by-pass
Add C such that the effective emitter
E
resistance becomes zero at signal frequency.
V CC R
L
So, this circuit has good biasing, and the R
B1
gain is still very high!
Gain = – g R
m L
in
R
which is unaffected by because R
R C E E E
effectively is shorted at signal R
E frequency. is called bypass capacitor. C E is Early voltage typically ~ 100V
V A
/
is thermal voltage ª 25mV
V T
I C
V A
o =
r
m
b g
p =
V T r
I C
) =
q
kT
Summary
(
I C
m =
BE
B E C E
p
r
o
r
BE
v
m
Basic BJT model (small-signal ac model): g
Basic FET model (small-signal ac model):
D G Similar to the BJT model,
∞ g v r v
m GS o
resistance. GS
S S used in the same way as amplifiers. Summary
Common-emitter (CE) amplifier small-signal ac model:
v
bias r
m BE o L in BE o pR
L
R
bias
m L in
r
bias p
Output resistance = || ≈ (large — undesirable) R r R
L o L
E
in
E
R
bias
R
E
) (small — desirable) R
m
|| (1/g
E
R
] (quite large — desirable) Output resistance =
|| [ r
p
bias
o
bias
in
BE
p
r
BE
v
m
Emitter follower (EF) small-signal ac model: g
Summary
o