Department of EECS University of California, Berkeley
EECS 105 Spring 2017, Module 3
Prof. Ali M. Niknejad Prof. Rikky Muller
Metal Oxide Semiconductor Field
Effect Transistors (MOSFETs)
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
2
Announcements
l Pick up Midterm 1 if you haven’t already! l Two options: (1) from my office hours, or (2) from
lecture today
University of California, Berkeley
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
3
MOSFET Cross Section
l Add two junctions around MOS capacitor l The regions forms PN junctions with substrate l MOSFET is a four terminal device l The body is usually grounded (or at a DC potential) l For ICs, the body contact is at surface
University of California, Berkeley
p-type substrate
n+ n+
source drain diffusion regions
L
gate body
p+
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
4
MOSFET Layout
l Planar process: complete structure can be specified by a 2D layout
l Design engineer can control the transistor width W and L
l Process engineer controls tox, Na, xj, etc. University of California, Berkeley
p-type substrate
n+ n+
S D G
B
p+
B S D
L
L
W
contact poly gate
jx
G
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
5
PMOS & NMOS
l A MOSFET by any other name is still a MOSFET: – NMOS, PMOS, nMOS, pMOS – NFET, PFET – IGFET – Other flavors: JFET, MESFET
l CMOS technology: The ability to fabricate NMOS and PMOS devices simultaneously
University of California, Berkeley
p-type substrate
n+ n+
S D B
p+ L jx
n-type substrate
p+ p+
S D B
n+ L jx
NMOS PMOS
G G
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
6
CMOS
l Complementary MOS (CMOS): Both P and N type devices l Create a n-type body in a p-type substrate through
compensation. This new region is called a “well”. l To isolate the PMOS from the NMOS, the well must be
reverse biased (p-n junction) University of California, Berkeley
n+ n+
S D B
p+ L jx
n-type well
p+ p+
S D B
n+ L jx
NMOS PMOS
G G
p-type substrate
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
7
Circuit Symbols
l The symbols with the arrows are typically used in analog applications
l The body contact is often not shown l The source/drain can switch depending on how the
device is biased (the device has inherent symmetry) University of California, Berkeley
NMOS PMOS
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
8
Circuits!
l When the inventors of the bipolar transistor first got a working device, the first thing they did was to build an audio amplifier to prove that the transistor was actually working!
University of California, Berkeley
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
9
A Simple Circuit: An MOS Amplifier
University of California, Berkeley
DSI
GSV
sv
DR DDV
GS GS sv V v= +
ov
Input signal
Output signal
Supply “Rail”
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
10
Plot of Output Waveform (Gain!)
University of California, Berkeley
output
input
mV
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
11
Observed Behavior: ID-VGS
l Current zero for negative gate voltage l Current in transistor is very low until the gate
voltage crosses the threshold voltage of device (same threshold voltage as MOS capacitor)
l Current increases rapidly at first and then it finally reaches a point where it simply increases linearly
University of California, Berkeley
GSV
DSI
TV
GSV
DSIDSV
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
12
Observed Behavior: ID-VDS
l For low values of drain voltage, the device is like a resistor l As the voltage is increases, the resistance behaves non-linearly
and the rate of increase of current slows l Eventually the current stops growing and remains essentially
constant (current source) University of California, Berkeley
DSV
/DSI k
“constant” current
resistor region
non-linear resistor region
2GSV V=
3GSV V=
4GSV V=
GSV
DSIDSV
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
13
Operating Points
l Which bias voltages you operate the MOSFET at will make a big difference in how it functions.
l We will explore these regions of operation in this lecture. l If we operate with a sufficiently high VGS AND a sufficiently
high VDS, we can make a very good small-signal amplifier! University of California, Berkeley
DSI
GSV
sv
DR DDV
GS GS sv V v= +
ov
Input signal
Output signal
Supply “Rail”
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
14
Small Signal vs. Large Signal
l We observe IDS vs. VGS to be quadratic for VGS > VT
l Large changes in vGS result in quadratic changes at the output l However, for small changes in vGS (denoted as vgs) will
produce linear changes at the output! l We can show this using Taylor series expansion:
University of California, Berkeley
VGS
DSI
TV
GSV
DSIDSV
iDS = k(vGS −VT )2;vGS =VGS + vgs
TS : iDS (vGS ) VGS = k(VGS −VT )2 + 2k(VGS −VT )vgs + kvgs
2
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
15
Simple Small Signal Model for MOSFET
l This is a simplified, 3-terminal small-signal model for a MOSFET
l In later lectures we will develop a more complete model l gm = transconductance
– defined as dids/dvgs, units [Ohms]-1
l ro = output resistance – defined as [dids/dvds]-1, units Ohms
University of California, Berkeley
NMOS
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
16
Small Signal Gain Example
University of California, Berkeley
DSI
GSV
sv
DR DDV
GS GS sv V v= +
ov
Input signal
Output signal
Supply “Rail”
l Steps to analyze small signal amplifiers: l 1. Calculate bias points using DC sources
– As you will see in later lecture, you will use these bias points to determine the MOSFET region of operation as well as to calculate small-signal parameters
l 2. Turn off DC sources l 3. Plug in the small-signal model for a MOSFET
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
17
Small Signal Gain Example
University of California, Berkeley
l 4. Calculate the gain (vout/vin) of the circuit
Department of EECS University of California, Berkeley
EECS 105 Spring 2017
MOSFET Large Signal Models and
Regions of Operation
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
19
Cut-off, VGS < VT
l This structure should look familiar! It is an MOS capacitor with two n+ diffusion regions on each side.
l When VGS < VT, the device is either in accumulation or in depletion.
l Since there are no (or few) inversion charges at the surface, therefore no current will flow regardless of the value of VDS. University of California, Berkeley
p-type
n+ n+ p+
Depletion Region
VGS <VT
100mVDSV ≈G
D S
NMOS
x
y
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
20
“Linear” Region Current
l If the gate is biased above threshold, the surface is inverted l This inverted region forms a channel of inversion charges
(in this case electrons) that connects the drain and source – inversion charges originate from n+ diffusion
l If a drain-source voltage (VDS) is applied positive, electrons will flow from source to drain
l Note: electrons flow S à D, current flows D à S University of California, Berkeley
p-type
n+ n+ p+
Inversion layer “channel”
VGS >VT
100mVDSV ≈G
D S
NMOS
x
y
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
21
MOSFET “Linear” Region
l The current in this channel is given by
l The charge proportional to the voltage applied across the oxide over threshold
l If the channel is uniform density, only drift current flows
University of California, Berkeley
DS y NI Wv Q= −
( )N ox GS TnQ C V V= −
( )DS y ox GS TnI Wv C V V= − −
y n yv Eµ= − DSy
VEL
= −
GS TnV V>( )DS n ox GS Tn DSWI C V V VLµ= − 100mVDSV ≈
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
22
MOSFET: Variable Resistor
l Notice that in the linear region, the current is proportional to the voltage
l Can define a voltage-dependent resistor
l This is a nice variable resistor, electronically tunable!
University of California, Berkeley
( )DS n ox GS Tn DSWI C V V VLµ= −
1 ( )( )
DSeq GS
DS n ox GS Tn
V L LR R VI C V V W Wµ
⎛ ⎞= = =⎜ ⎟− ⎝ ⎠W
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
23
Finding ID = f (VGS, VDS) l Approximate inversion charge QN(y): drain
voltage is higher than the source à less charge at drain end of channel
University of California, Berkeley
p-type
n+ n+ p+
GS TnV V>
DSVG
D S
NMOS
GS TnV V−
y=L y=0
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
24
Inversion Charge at Source/Drain
University of California, Berkeley
)()0()( LyQyQyQ NNN =+=≈
)()0( TnGSoxN VVCyQ −−== == )( LyQN)( TnGDox VVC −−
DSGSGD VVV −=
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
25
Average Inversion Charge
l Charge at drain end is lower since the vertical field is lower at that point
University of California, Berkeley
( ) ( )( )2
ox GS T ox GD TN
C V V C V VQ y − + −≈ −
Source End Drain End
( ) ( )( )2
ox GS T ox GS SD TN
C V V C V V VQ y − + − −≈ −
(2 2 )( ) ( )2 2
ox GS T ox SD DSN ox GS T
C V V C V VQ y C V V− −≈ − = − − −
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
26
Drift Velocity and Drain Current
University of California, Berkeley
Use mobility to find velocity v
( ) ( ) ( / ) n DSn n
Vv y E y V yL
µµ µ= − ≈ − −Δ Δ =
Substituting:
( )2
DS DSD N ox GS T
V VI WvQ W C V VL
µ= − ≈ − −
ID =WLµCox (VGS −VT −
VDS2)VDS
Family of Inverted Parabolas
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
27
Square-Law Characteristics
University of California, Berkeley
Boundary: what is ID,SAT?
TRIODE REGION TRIODE REGION
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
28
The Saturation Region
University of California, Berkeley
When VDS > VGS – VTn, there isn’t any inversion charge at the drain … what happens?
Why do curves flatten out?
SATURATION REGION
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
29
Why does current saturate?
l The charge at drain end goes to zero once VGD < VT
l We say that the drain end is “pinched off” – If you pinch a hose, water flow stops ! – But then how does current flow?
University of California, Berkeley
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
30
Pinch Off
l Excess field beyond Edsat drops across tiny region between drain and channel
– Huge field means that electrons flow at very high velocity across the “high field” region.
– They are injected from source end and are collected at the drain end
l Increasing the drain voltage does not increase current (appreciably) because the current is limited by the supply of electrons from channel side
University of California, Berkeley
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
31
Square-Law Current in Saturation
University of California, Berkeley
Current stays at maximum (where VDS = VGS – VTn = VDS,SAT)
( )2DS
D ox GS T DSVWI C V V V
Lµ= − −
, ( )( )2
GS TDS sat ox GS T GS T
V VWI C V V V VLµ
−= − − −
2, ( )
2ox
DS sat GS TCWI V V
Lµ
= −
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
32
Actual Saturation Current l Measurement: ID increases slightly with increasing VDS: l The physics is complicated, but a simple way to see this is that
the channel is getting shorter as the drain voltage depletes away more electrons from the drain end
l We model this with an additional linear factor:
University of California, Berkeley
2, ( ) (1 )
2ox
DS sat GS T DSCWI V V V
Lµ
λ= − +
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
33
Channel Length Modulation l When vDS = vGS-VT, the channel pinches off near the drain. With
further increase in vDS, the pinch-off point moves toward the source, effectively reducing the channel length from L to L-ΔL.
University of California, Berkeley
ID =12µnCox
WL −ΔL
VGS −VT( )2
The continual increase of ID with VDS is modeled by
ID =12µnCox
WLVGS −Vtn( )2 1+λVDS( )
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
34
Summary: Regions of Operation
l Cut-off: VGS < VT – IDS = 0 – Note: this is an approximation we will make in EE105, in later
courses you will learn about sub-threshold conduction
l Linear: VGS > VT, VDS << VGS – VT
l Triode: VGS > VT, VDS < VGS – VT
l Saturation: VGS > VT, VDS > VGS – VT
University of California, Berkeley
2, ( ) (1 )
2ox
DS sat GS T DSCWI V V V
Lµ
λ= − +
IDS =WLµnCox (VGS −VT −
VDS2)VDS
IDS =WLµnCox (VGS −VT )VDS
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
35
PMOS Device
l So far, we’ve derived all of our equations for an NMOS device
l PMOS devices work exactly the same way, but with an n-type body and a channel made of positive charges (holes)
l The direction of the voltages and currents are inverted, for example:
University of California, Berkeley
ISD,sat =WLµCox2(VSG −VTp )
2(1+λVSD )
EE 105 Spring 2017 Prof. A. M. Niknejad Prof. Rikky Muller
36
Next Lecture
l In the next lecture we will learn how to analyze small-signal amplifiers, including
– Computation of bias points – How to derive and compute small signal model
parameters (gm, ro) – How to calculate small signal gain
University of California, Berkeley