Transcript Lecture 1

Lecture 20
ANNOUNCEMENTS
• HW#11 is due in 2 weeks, on 11/20.
• Review session: Fri. 11/9, 3-5PM in 306 Soda (HP Auditorium)
• Midterm #2 (Thursday 11/15 in Sibley Auditorium):
• Material of Lectures 11-18 (HW# 7-10; Chapters 6,9,11)
• 4 pgs of notes (double-sided, 8.5”×11”), calculator allowed
OUTLINE
• Review of MOSFET Amplifiers
• MOSFET Cascode Stage
• MOSFET Current Mirror
Reading: Chapter 9
EE105 Fall 2007
Lecture 20, Slide 1
Prof. Liu, UC Berkeley
Review: MOSFET Amplifier Design
• A MOSFET amplifier circuit should be designed to
1. ensure that the MOSFET operates in the saturation region,
2. allow the desired level of DC current to flow, and
3. couple to a small-signal input source and to an output “load”.
 Proper “DC biasing” is required!
(DC analysis using large-signal MOSFET model)
• Key amplifier parameters:
(AC analysis using small-signal MOSFET model)
– Voltage gain Av  vout/vin
– Input resistance Rin  resistance seen between the input node
and ground (with output terminal floating)
– Output resistance Rout  resistance seen between the output
node and ground (with input terminal grounded)
EE105 Fall 2007
Lecture 20, Slide 2
Prof. Liu, UC Berkeley
MOSFET Models
• The large-signal model is used to determine the DC
operating point (VGS, VDS, ID) of the MOSFET.
• The small-signal model is used to determine how the
output responds to an input signal.
EE105 Fall 2007
Lecture 20, Slide 3
Prof. Liu, UC Berkeley
Comparison of Amplifier Topologies
Common Source
• Large Av < 0
- degraded by RS
• Large Rin
Common Gate
• Large Av > 0
-degraded by RS
• Small Rin
- decreased by RS
– determined by biasing
circuitry
• Rout  RD
• Rout  RD
• ro decreases Av & Rout
but impedance seen
looking into the drain
can be “boosted” by
source degeneration
EE105 Fall 2007
• ro decreases Av & Rout
but impedance seen
looking into the drain
can be “boosted” by
source degeneration
Lecture 20, Slide 4
Source Follower
• 0 < Av ≤ 1
• Large Rin
– determined by
biasing circuitry
• Small Rout
- decreased by RS
• ro decreases Av &
Rout
Prof. Liu, UC Berkeley
Common Source Stage
 0
R1 || R2
 RD
Av 

RG  R1 || R2 1  R
S
gm
Rin  R1 || R2
Rout  RD
EE105 Fall 2007
0
Rout  RD rO  g m rO RS 
Lecture 20, Slide 5
Prof. Liu, UC Berkeley
Common Gate Stage
 0
RS || 1 / g m 
Av 
 g m RD
RS || 1 / g m   RG
1
Rin 
RS
gm
0
Rout  RD
Rout  RD rO  g m rO RS 
EE105 Fall 2007
Lecture 20, Slide 6
Prof. Liu, UC Berkeley
Source Follower
0
 0
Av 
RS
Av 
1
 RS
gm
Rin  RG
Rout
1
 rO || RS
gm
Rin  R G
1

|| RS
gm
EE105 Fall 2007
rO || RS
Rout
Lecture 20, Slide 7
1

|| ro || RS
gm
Prof. Liu, UC Berkeley
CS Stage Example 1
• M1 is the amplifying device; M2 and M3 serve as the load.
Equivalent circuit for small-signal analysis,
showing resistances connected to the drain
 1

Av   g m1 
|| rO 3 || rO 2 || rO1 
 g m3

1
Rout 
|| rO 3 || rO 2 || rO1
g m3
EE105 Fall 2007
Lecture 20, Slide 8
Prof. Liu, UC Berkeley
CS Stage Example 2
• M1 is the amplifying device; M3 serves as a source (degeneration)
resistance; M2 serves as the load.
Equivalent circuit for small-signal analysis
1  0
rO 2
Av  
1
1

|| rO 3
g m1 g m3
EE105 Fall 2007
Lecture 20, Slide 9
Prof. Liu, UC Berkeley
CS Stage vs. CG Stage
• With the input signal applied at different locations, these circuits
behave differently, although they are identical in other aspects.
Common gate amplifier
Common source amplifier
1  0
2  0
Av   g m1(1  g m 2 rO 2 ) RS  rO 2  || rO1
EE105 Fall 2007
Lecture 20, Slide 10
Av 
rO1
1
 RS
gm2
Prof. Liu, UC Berkeley
Composite Stage Example 1
• By replacing M1 and the current source with a Thevenin
equivalent circuit, and recognizing the right side as a CG stage,
the voltage gain can be easily obtained.
1  0
EE105 Fall 2007
2  0
Av 
Lecture 20, Slide 11
RD
1
1

g m 2 g m1
Prof. Liu, UC Berkeley
Composite Stage Example 2
• This example shows that by probing different nodes in a circuit,
different output signals can be obtained.
• Vout1 is a result of M1 acting as a source follower, whereas Vout2
is a result of M1 acting as a CS stage with degeneration.
1
|| rO 2
vout1
gm2

1
1
vin

|| rO 2
g m1 g m 2
1  0
vout 2
vin
EE105 Fall 2007
Lecture 20, Slide 12
1
|| rO 3 || rO 4
g m3

1
1

|| rO 2
g m1 g m 2
Prof. Liu, UC Berkeley
NMOS Cascode Stage
Rout  1  g m1rO1 rO 2  rO1
Rout  g m1rO1rO 2
• Unlike a BJT cascode, the output impedance is not limited by .
EE105 Fall 2007
Lecture 20, Slide 13
Prof. Liu, UC Berkeley
PMOS Cascode Stage
Rout  1  g m1rO1 rO 2  rO1
Rout  g m1rO1rO 2
EE105 Fall 2007
Lecture 20, Slide 14
Prof. Liu, UC Berkeley
Short-Circuit Transconductance
• The short-circuit transconductance is a measure of the
strength of a circuit in converting an input voltage signal into
an output current signal:
iout
Gm 
vin
• The voltage gain of a linear circuit is
vou t  0
Av  Gm Rout
(Rout is the output resistance of the circuit)
EE105 Fall 2007
Lecture 20, Slide 15
Prof. Liu, UC Berkeley
Transconductance Example
Gm  g m1
EE105 Fall 2007
Lecture 20, Slide 16
Prof. Liu, UC Berkeley
MOS Cascode Amplifier
Av  Gm Rout
Av   g m1 (1  g m 2 rO 2 )rO1  rO 2 
Av   g m1rO1 g m 2 rO 2
EE105 Fall 2007
Lecture 20, Slide 17
Prof. Liu, UC Berkeley
PMOS Cascode Current Source as Load
• A large load impedance can be achieved by using a PMOS
cascode current source.
RoN  g m 2 rO 2 rO1
RoP  g m 3rO 3 rO 4
Rout  RoN || RoP
EE105 Fall 2007
Lecture 20, Slide 18
Prof. Liu, UC Berkeley
MOS Current Mirror
• The motivation behind a current mirror is to duplicate a
(scaled version of the) “golden current” to other locations.
Current mirror concept
Generation of required VGS
1
W 
2
I REF  nCox   VX  VTH 
2
 L  REF
VX 
EE105 Fall 2007
2 I REF
 VTH 1


nCox W / L 1
Lecture 20, Slide 19
Current Mirror Circuitry
I copy1 
1
W 
nCox   VX  VTH 2
2
 L 1

W / L 1
I copy1 
I REF
W / L REF
Prof. Liu, UC Berkeley
MOS Current Mirror – NOT!
• This is not a current mirror, because the relationship between
VX and IREF is not clearly defined.
• The only way to clearly define VX with IREF is to use a diodeconnected MOS since it provides square-law I-V relationship.
EE105 Fall 2007
Lecture 20, Slide 20
Prof. Liu, UC Berkeley
Example: Current Scaling
• MOS current mirrors can be used to scale IREF up or down
– I1 = 0.2mA; I2 = 0.5mA
  0:
EE105 Fall 2007
Lecture 20, Slide 21
Prof. Liu, UC Berkeley
Impact of Channel-Length Modulation
0
1
W 
 nCox   VX  VTH 2 1   VDS1  VD , sat 
2
 L 1
1
W 
2
  nCox   VX  VTH  1   VDS1  VGS  VTH 
2
 L 1
I copy1 
1
W 
2
I REF   nCox   VX  VTH  1   VGS  VD , sat 
2
 L  REF
1
W 
2
  nCox   VX  VTH  1  VTH 
2
 L  REF
  VDS1  VGS  


1   VDS1  VGS  VTH 
W / L 1
W / L 1

I copy1 
I REF

I REF 1 
W / LREF
W / LREF
1  VTH
1  VTH 

EE105 Fall 2007
Lecture 20, Slide 22
Prof. Liu, UC Berkeley
CMOS Current Mirror
EE105 Fall 2007
Lecture 20, Slide 23
Prof. Liu, UC Berkeley