MOS Transistor Models
Lecture Outline
MOS Transistors
- I-V curve (Square-Law Model)
– Small Signal Model (Linear Model)
Observed Behavior: ID-VGS
C u r r e n t zero for negative gate voltage
Current in transistor is very low until the gate
voltage crosses the threshold voltage of device
(same threshold voltage as MOS capacitor)
Current increases rapidly at first and then it finally
reaches a point where it simply increases linearly
VGS
IDS
T
V
VGS
IDS
DS
V
Observed Behavior: ID-VDS
VDS
For low values of drain voltage, the device is like a resistor
As the voltage is increases, the resistance behaves non-linearly
and the rate of increase of current slows
E v e n t u a l l y t h e c u r r e n t stops growing and
remains essentially constant (current source)
IDS /k
“constant” current
resistor region
non-linear resistor region
VGS  2V
VGS  3V
VGS  4V
GS
V
IDS
DS
V
“Linear” Region Current
I f t h e gate is biased above threshold, the
surface is inverted
This inverted region forms a channel that connects
the drain and gate
I f a d r a i n voltage is applied positive,
electrons will flow from source to drain
p-type
n+ n+
p+
Inversion layer
“channel”
VGS VTn
DS
V 100mV
G
D
S
NMOS
x
y
MOSFET: Variable Resistor
N o t i c e t h a t in the linear region, the
current is proportional to the voltage
Can define a voltage-dependent resistor
T h i s i s a n i c e variable resistor,
electronically tunable!
DS n ox GS Tn DS
L
I 
W
C (V V )V
Finding ID = f (VGS, VDS)
Approximate inversion charge QN(y): drain is
higher than the source € less charge at drain end
of channel
Inversion Charge at Source/Drain
QN (y)  QN (y  0)  QN (y  L)
QN (y  0)  Cox (VGS VTn ) QN ( y  L)
 Cox (VGD  VTn)
GD  VGS VDS
Average Inversion Charge
Charge at drain end is lower since field is lower
Simple approximation: In reality we should
integrate the total charge minus the bulk depletion
charge across the channel
VT )  Cox (VGDVT)
2
N
Q (y)  
Cox (VGS
Source End Drain End
2
N
VT )  Cox (VGSVSD VT )
Q (y)  
Cox (VGS
 2VT) CoxVSD
2 2
N ox GS T
Q (y)  
Cox (2VGS
 C (V V VDS
)
Drift Velocity and Drain Current
“Long-channel” assumption: use mobility to find v
n DS
n n V / y) 
v(y)   E(y)   (
Substituting:
L
 V
2
VDS
)
D N ox GS T
L
I  WvQ W
VDS
C (V V
2
D ox GS T DS
L
I 
W
C (V V 
VDS
)V
Inverted Parabolas
Square-Law Characteristics
Boundary: what is ID,SAT?
TRIODE REGION
SATURATION REGION
The Saturation Region
When VDS > VGS – VTn, there isn’t anyinversion
charge at the drain … according to our simplistic model
Why do curves
flatten out?
Square-Law Current in Saturation
Current stays at maximum (where VDS = VGS – VTn =VDS,SAT)
Measurement: ID increases slightly with increasingVDS
model with linear “fudge factor”
2
D ox GS T DS
L
I 
W
C (V V 
VDS
)V
DS,sat
2
ox GS T GS T
VGS VT
)(V V )
L
I 
W
C (V V
DS,sat GS T
L 2
I 
W Cox
(V V )2
2
DS,sat GS T DS
V )
L 2
I 
W Cox
(V V ) (1
Pinching the MOS Transistors
When VDS > VDS,sat, the channel is “pinched” off at drain end (hence the
name “pinch-off region”)
Drain mobile charge goes to zero (region is depleted), the remaining elecric
field is dropped across this high-field depletion region
As the drain voltage is increases further, the pinch off point moves back
towards source
Channel Length Modulation: The effective channel length is thus reduced
higher IDS
p-type
n+ n+
p+
Pinch-Off Point
VGS VTn
DS
V
G
D
S
NMOS
Depletion Region
VGS VTn
Linear MOSFET Model
Channel (inversion) charge: neglect reduction at drain
Velocity saturation defines VDS,SAT = Esat L =constant
- vsat /n
Drain current:
ID,SAT  WvQN  W (vsat)[Cox (VGS VTn )],
|Esat| = 104 V/cm, L = 0.12 m € VDS,SAT = 0.12 V!
ID,SAT  vsatWCox (VGS VTn)(1 nVDS )
Why Find an Incremental Model?
Direct substitution into iD = f(vGS, vDS) is
tedious AND doesn’t include charge-storage
effects … pretty rough approximation
Signals of interest in analog ICs are often of the
form:
vGS (t)  VGS vgs (t)
Fixed Bias Point Small Signal
Which Operating Region?
 3V
 3V
VGS
VDS
TRIODE
SAT
OFF
Changing One Variable at a Time
IDS /k
VDS  3V
VT 1V
Square Law
Saturation
Region
Linear
Triode
Region
Slope of Tangent: Incremental current increase
VGS
Assumption: VDS > VDS,SAT = VGS – VTn (square law)
The Transconductance gm
Defined as the change in drain current due to a change in the
gate-source voltage, with everything else constant
GS VGS ,VDS

iD
m
GS VGS ,VDS
ox GS T DS
V )(1 V )
g 
iD
v v L
 C
W
(V
2
ox
DS,sat
L 2
GS T DS
V )
(V  V ) (1 
W C
I 
m ox GS T
L
g  C
W
(V V )
 0
m ox ox DS
ox
2IDS
W
L
W
I
L
 2C
g  C
L
W
C
2IDS
(VGS VT )
m
g 
Gate Bias
Drain Current Bias
Drain Current Bias and
Gate Bias
Output Resistance ro
Defined as the inverse of the change in drain current due
to a change in the drain-source voltage, with everything
else constant
Non-Zero Slope
VDS
IDS
Evaluating ro
o
DS
iD

r 
1
v VGS ,VDS
2
ox
L 2
D GS T DS
V )
(V V ) (1 
W C
i 
0
2
1
ox
(VGS
L 2 T
V ) 
r 
W C
0
1
DS
r 
I
Total Small Signal Current
iDS (t)  IDS ids

iDS  iDS
v
ds gs ds
gs ds
i v
v v
ds m gs ds
o
r
i  g v 
1
v
Transconductance
Conductance
Putting Together a Circuit Model
ds m gs ds
o
i
r
 g v 
1
v
Role of the Substrate Potential
Need not be the source potential, but VB <VS
Effect: changes threshold voltage, which
changes the drain current … substrate acts
like a “backgate”
vBS Q

iD
vBS Q
gmb 
iD
Q = (VGS, VDS, VBS)
Backgate Transconductance
Result:  gm
mb
BS BS
Tn Q
VTn

iD
BS p
2 V 2
Q Q
g 
iD
v V v

VSB  2p
VT  VT0    2p 
Four-Terminal Small-Signal Model
ds m gs mb bs ds
o
i  g v
r
 g v 
1
v
MOSFET Capacitances in Saturation
Gate-source capacitance: channel charge is not
controlled by drain in saturation.
Gate-Source Capacitance Cgs
Wedge-shaped charge in saturation € effective area is (2/3)WL
(see H&S 4.5.4 for details)
Cgs  (2/3)WLCox  Cov
Overlap capacitance along source edge of gate €
Cov  LDWCox
(Underestimate due to fringing fields)
Gate-Drain Capacitance Cgd
Not due to change in inversion charge in channel
Overlap capacitance Cov between drain and source
is Cgd
Junction Capacitances
Drain and source diffusions have (different) junction
capacitances since VSB and VDB = VSB+ VDS aren’t
the same
Complete model (without interconnects)
P-Channel MOSFET
Measurement of –IDp versus VSD, with VSG as a parameter:
Square-Law PMOS Characteristics
Small-Signal PMOS Model
MOSFET SPICE Model
Many “levels” … we will use the square-law
“Level 1” model
See H&S 4.6 + Spice refs. on reserve for details.

IDS MOS Equation (1).pptx

  • 1.
  • 2.
    Lecture Outline MOS Transistors -I-V curve (Square-Law Model) – Small Signal Model (Linear Model)
  • 3.
    Observed Behavior: ID-VGS Cu r r e n t zero for negative gate voltage Current in transistor is very low until the gate voltage crosses the threshold voltage of device (same threshold voltage as MOS capacitor) Current increases rapidly at first and then it finally reaches a point where it simply increases linearly VGS IDS T V VGS IDS DS V
  • 4.
    Observed Behavior: ID-VDS VDS Forlow values of drain voltage, the device is like a resistor As the voltage is increases, the resistance behaves non-linearly and the rate of increase of current slows E v e n t u a l l y t h e c u r r e n t stops growing and remains essentially constant (current source) IDS /k “constant” current resistor region non-linear resistor region VGS  2V VGS  3V VGS  4V GS V IDS DS V
  • 5.
    “Linear” Region Current If t h e gate is biased above threshold, the surface is inverted This inverted region forms a channel that connects the drain and gate I f a d r a i n voltage is applied positive, electrons will flow from source to drain p-type n+ n+ p+ Inversion layer “channel” VGS VTn DS V 100mV G D S NMOS x y
  • 6.
    MOSFET: Variable Resistor No t i c e t h a t in the linear region, the current is proportional to the voltage Can define a voltage-dependent resistor T h i s i s a n i c e variable resistor, electronically tunable! DS n ox GS Tn DS L I  W C (V V )V
  • 7.
    Finding ID =f (VGS, VDS) Approximate inversion charge QN(y): drain is higher than the source € less charge at drain end of channel
  • 8.
    Inversion Charge atSource/Drain QN (y)  QN (y  0)  QN (y  L) QN (y  0)  Cox (VGS VTn ) QN ( y  L)  Cox (VGD  VTn) GD  VGS VDS
  • 9.
    Average Inversion Charge Chargeat drain end is lower since field is lower Simple approximation: In reality we should integrate the total charge minus the bulk depletion charge across the channel VT )  Cox (VGDVT) 2 N Q (y)   Cox (VGS Source End Drain End 2 N VT )  Cox (VGSVSD VT ) Q (y)   Cox (VGS  2VT) CoxVSD 2 2 N ox GS T Q (y)   Cox (2VGS  C (V V VDS )
  • 10.
    Drift Velocity andDrain Current “Long-channel” assumption: use mobility to find v n DS n n V / y)  v(y)   E(y)   ( Substituting: L  V 2 VDS ) D N ox GS T L I  WvQ W VDS C (V V 2 D ox GS T DS L I  W C (V V  VDS )V Inverted Parabolas
  • 11.
    Square-Law Characteristics Boundary: whatis ID,SAT? TRIODE REGION SATURATION REGION
  • 12.
    The Saturation Region WhenVDS > VGS – VTn, there isn’t anyinversion charge at the drain … according to our simplistic model Why do curves flatten out?
  • 13.
    Square-Law Current inSaturation Current stays at maximum (where VDS = VGS – VTn =VDS,SAT) Measurement: ID increases slightly with increasingVDS model with linear “fudge factor” 2 D ox GS T DS L I  W C (V V  VDS )V DS,sat 2 ox GS T GS T VGS VT )(V V ) L I  W C (V V DS,sat GS T L 2 I  W Cox (V V )2 2 DS,sat GS T DS V ) L 2 I  W Cox (V V ) (1
  • 14.
    Pinching the MOSTransistors When VDS > VDS,sat, the channel is “pinched” off at drain end (hence the name “pinch-off region”) Drain mobile charge goes to zero (region is depleted), the remaining elecric field is dropped across this high-field depletion region As the drain voltage is increases further, the pinch off point moves back towards source Channel Length Modulation: The effective channel length is thus reduced higher IDS p-type n+ n+ p+ Pinch-Off Point VGS VTn DS V G D S NMOS Depletion Region VGS VTn
  • 15.
    Linear MOSFET Model Channel(inversion) charge: neglect reduction at drain Velocity saturation defines VDS,SAT = Esat L =constant - vsat /n Drain current: ID,SAT  WvQN  W (vsat)[Cox (VGS VTn )], |Esat| = 104 V/cm, L = 0.12 m € VDS,SAT = 0.12 V! ID,SAT  vsatWCox (VGS VTn)(1 nVDS )
  • 16.
    Why Find anIncremental Model? Direct substitution into iD = f(vGS, vDS) is tedious AND doesn’t include charge-storage effects … pretty rough approximation Signals of interest in analog ICs are often of the form: vGS (t)  VGS vgs (t) Fixed Bias Point Small Signal
  • 17.
    Which Operating Region? 3V  3V VGS VDS TRIODE SAT OFF
  • 18.
    Changing One Variableat a Time IDS /k VDS  3V VT 1V Square Law Saturation Region Linear Triode Region Slope of Tangent: Incremental current increase VGS Assumption: VDS > VDS,SAT = VGS – VTn (square law)
  • 19.
    The Transconductance gm Definedas the change in drain current due to a change in the gate-source voltage, with everything else constant GS VGS ,VDS  iD m GS VGS ,VDS ox GS T DS V )(1 V ) g  iD v v L  C W (V 2 ox DS,sat L 2 GS T DS V ) (V  V ) (1  W C I  m ox GS T L g  C W (V V )  0 m ox ox DS ox 2IDS W L W I L  2C g  C L W C 2IDS (VGS VT ) m g  Gate Bias Drain Current Bias Drain Current Bias and Gate Bias
  • 20.
    Output Resistance ro Definedas the inverse of the change in drain current due to a change in the drain-source voltage, with everything else constant Non-Zero Slope VDS IDS
  • 21.
    Evaluating ro o DS iD  r  1 vVGS ,VDS 2 ox L 2 D GS T DS V ) (V V ) (1  W C i  0 2 1 ox (VGS L 2 T V )  r  W C 0 1 DS r  I
  • 22.
    Total Small SignalCurrent iDS (t)  IDS ids  iDS  iDS v ds gs ds gs ds i v v v ds m gs ds o r i  g v  1 v Transconductance Conductance
  • 23.
    Putting Together aCircuit Model ds m gs ds o i r  g v  1 v
  • 24.
    Role of theSubstrate Potential Need not be the source potential, but VB <VS Effect: changes threshold voltage, which changes the drain current … substrate acts like a “backgate” vBS Q  iD vBS Q gmb  iD Q = (VGS, VDS, VBS)
  • 25.
    Backgate Transconductance Result: gm mb BS BS Tn Q VTn  iD BS p 2 V 2 Q Q g  iD v V v  VSB  2p VT  VT0    2p 
  • 26.
    Four-Terminal Small-Signal Model dsm gs mb bs ds o i  g v r  g v  1 v
  • 27.
    MOSFET Capacitances inSaturation Gate-source capacitance: channel charge is not controlled by drain in saturation.
  • 28.
    Gate-Source Capacitance Cgs Wedge-shapedcharge in saturation € effective area is (2/3)WL (see H&S 4.5.4 for details) Cgs  (2/3)WLCox  Cov Overlap capacitance along source edge of gate € Cov  LDWCox (Underestimate due to fringing fields)
  • 29.
    Gate-Drain Capacitance Cgd Notdue to change in inversion charge in channel Overlap capacitance Cov between drain and source is Cgd
  • 30.
    Junction Capacitances Drain andsource diffusions have (different) junction capacitances since VSB and VDB = VSB+ VDS aren’t the same Complete model (without interconnects)
  • 31.
    P-Channel MOSFET Measurement of–IDp versus VSD, with VSG as a parameter:
  • 32.
  • 33.
  • 34.
    MOSFET SPICE Model Many“levels” … we will use the square-law “Level 1” model See H&S 4.6 + Spice refs. on reserve for details.