digital integrated circuits -...
TRANSCRIPT
Digital IC Introduction
Digital Integrated
Circuits YuZhuo Fu
contact:[email protected]
Office location:417 room
WeiDianZi building,No 800 DongChuan
road,MinHang Campus
Digital IC Introduction
3.CMOS Inverter
Digital IC 3
outline
• CMOS at a glance
• CMOS static behavior
• CMOS dynamic behavior
• Power, Energy, and Energy Delay
• Perspective tech.
Digital IC 4
CMOS dynamic characteristic
• CMOS capacitances mosaic
• CMOS propagation delay
• Optimizing inverter sizing
Digital IC 5
Circuit Under Design
VDD VDD
VinVout
M1
M2
M3
M4
Vout2
This two-inverter circuit will be manufactured in a twin-well process.
Digital IC
CMOS Inverter: Transient Response
6
VDD
Vout
Vin = VDD
Ron
CL
tpHL = f(Ron.CL)
= 0.69 RonCL
t
Vout
VDD
RonCL
1
0.5
ln(0.5)
0.36
Digital IC 7
CMOS capacitance mosaic
• Wire capacitance
• Junction(diffusion)
capacitance
• Gate capacitance
Most of them are nonlinear functions!
Digital IC 8
Computing the Capacitances
VDDVDD
VinVout
M1
M2
M3
M4Cdb2
Cdb1
Cgd12
Cw
Cg4
Cg3
Vout2
Fanout
Interconnect
VoutVin
CL
SimplifiedModel
Polysilicon
In Out
Metal1
V DD
GND
PMOS
NMOS
1.2 m m =2l
Digital IC 9
Capacitance model
DS
G
B
CGDCGS
CSB CDBCGBDdiffDB
SdiffSB
GCBGB
GCDGDOGD
GCSGSOGS
CC
CC
CC
CCC
CCC
Digital IC 10
The Miller Effect
V in
M 1
C gd 1
V out
D V
D V
V in
M 1
V out D V
D V
2 C gd 1
“A capacitor experiencing identical but opposite
voltage swings at both its terminals can be
replaced by a capacitor to ground, whose value is
two times the original value.”
Digital IC
Diffusion capacitances
Slide 11
0jeqeq CK=C
])V-Φ(-)V-Φ[(m)-)(1V-(V
Φ-=K m-1
low0
m-1
high0
lowhigh
m
0
eq
Digital IC
Miller effect
12
Digital IC 13
Computing the Capacitances
Cox
(fF/um2)
Co
(fF/um)
Cj
(fF/um2)
mj Φb
(V)
Cjsw
(fF/um)
Mjsw
Φbsw
(V)
NMOS 6 0.31 2 0.5 0.9 0.28 0.44 0.9
PMOS 6 0.27 1.9 0.48 0.9 0.22 0.32 0.9
CDG0
CJ CJSW
W/L AD(um2) PD(um) AS(um2) PS(um)
NMOS 3/2 19 15 19 15
PMOS 9/2 45 19 45 19
Digital IC
Computing the Capacitances
Slide 14
Polysilicon
In Out
Metal1
V DD
GND
PMOS
NMOS
1.2 m m =2l
AD=4*4+3*1=16+3=19λ2
PD=1+4+4+4+1+1=15λ
Digital IC
Computing the Capacitances
• Vhigh=-2.5V,Vlow=-1.25V[NMOS,{2.5V->1.25V} HL]
• Bottom plate:Keqn(m=0.5,Φ0=0.9)=0.57
• Sidewall:Keqwn(m=0.44,Φ0=0.9)=0.61
• Vlow=0V,Vhigh=-1.25V [NMOS,{0V->1.25V}LH]
• Bottom plate:Keqn(m=0.5,Φ0=0.9)=0.79
• Sidewall:Keqwn(m=0.44,Φ0=0.9)=0.81
• Vhigh=-1.25V,Vlow=0V [PMOS,{2.5V->1.25V}HL]
• Bottom plate:Keqp(m=0.48,Φ0=0.9)=0.79
• Sidewall:Keqwp(m=0.32,Φ0=0.9)=0.86
• Vhigh=-2.5V,Vlow=-1.25V [PMOS,{0V->1.25V}LH]
• Bottom plate:Keqp(m=0.48,Φ0=0.9)=0.59
• Sidewall:Keqwp(m=0.32,Φ0=0.9)=0.7
15
Digital IC 16
Computing the Capacitances
capacitor expression Value(fF)
(H->L)
Value(fF)
(L->H)
Cgd1 2CGD0n*Wn 0.23 0.23
Cgd2 2CGD0p*Wp 0.61 0.61
Cdb1 KeqnADnCJ+KeqwnPDnCJSW 0.66 0.90
Cdb2 KeqnADnCJ+KeqwnPDnCJSW 1.5 1.15
Cg3 (CGD0n+CGSOn)Wn+CoxWnLn 0.76 0.76
Cg4 (CGD0p+CGSOp)Wp+CoxWpLp 2.28 2.28
Cw 0.12 0.12
CL 6.1 6.0
Digital IC Introduction
Computing it more
simple by estimation
Slide 17
Digital IC
Capacitance
• Any two conductors separated by an insulator have
capacitance
• Gate to channel capacitor is very important
• Creates channel charge necessary for operation
• Source and drain have capacitance to body
• Across reverse-biased diodes
• Called diffusion capacitance because it is associated
with source/drain diffusion
Slide 18
Digital IC Slide 19
Gate Capacitance
• Approximate channel as connected to source
• Cgs = eoxWL/tox = CoxWL = CpermicronW
• Cpermicron is typically about 2 fF/mm
n+ n+
p-type body
W
L
tox
SiO2 gate oxide
(good insulator, eox
= 3.9e0)
polysilicon
gate
Digital IC Slide 20
Diffusion Capacitance
• Csb, Cdb
• Undesirable, called parasitic capacitance
• Capacitance depends on area and perimeter
• Use small diffusion nodes
• Comparable to Cg
for contacted diff
• ½ Cg for uncontacted
• Varies with process
Digital IC Slide 21
Effective Resistance
• Shockley models have limited value
• Not accurate enough for modern transistors
• Too complicated for much hand analysis
• Simplification: treat transistor as resistor
• Replace Ids(Vds, Vgs) with effective resistance R
• Ids = Vds/R
• R averaged across switching of digital gate
• Too inaccurate to predict current at any given time
• But good enough to predict RC delay
Digital IC Slide 22
RC Delay Model
• Use equivalent circuits for MOS transistors
• Ideal switch + capacitance and ON resistance
• Unit nMOS has resistance R, capacitance C
• Unit pMOS has resistance 2R, capacitance C
• Capacitance proportional to width
• Resistance inversely proportional to width
kg
s
d
g
s
d
kCkC
kCR/k
kg
s
d
g
s
d
kC
kC
kC
2R/k
Digital IC
Reason of 2R
Slide 23
)2
V-VV(
L
Wk=I
2
DSAT
DSATGT
'
DSAT
Digital IC Slide 24
RC Values
• Capacitance
• C = Cg = Cs = Cd = 2 fF/mm of gate width
• Values similar across many processes
• Resistance
• R 6 KW*mm in 0.6um process
• Improves with shorter channel lengths
• Unit transistors
• May refer to minimum contacted device (4/2 l)
• Or maybe 1 mm wide device
• Doesn’t matter as long as you are consistent
Digital IC Slide 25
Inverter Delay Estimate
• Estimate the delay of a fanout-of-1 inverter
2
1A
Y 2
1
Digital IC Slide 26
Inverter Delay Estimate
• Estimate the delay of a fanout-of-1 inverter
C
CR
2C
2C
R
2
1A
Y
C
2C
Y2
1
Digital IC Slide 27
Inverter Delay Estimate
• Estimate the delay of a fanout-of-1 inverter
C
CR
2C
2C
R
2
1A
Y
C
2C
C
2C
C
2C
RY
2
1
Digital IC Slide 28
Inverter Delay Estimate
• Estimate the delay of a fanout-of-1 inverter
C
CR
2C
2C
R
2
1A
Y
C
2C
C
2C
C
2C
RY
2
1
d = 6RC
Digital IC 29
CMOS dynamic characteristic
• CMOS capacitances mosaic
• CMOS propagation delay
• Optimizing inverter sizing
Digital IC
Inverter Transient Response
-0.5
0
0.5
1
1.5
2
2.5
3
0 0.5 1 1.5 2 2.5
Vin
t (sec) x 10-10
VDD=2.5V
0.25mm
W/Ln = 1.5
W/Lp = 4.5
Reqn= 13 kW ( 1.5)
Reqp= 31 kW ( 4.5)
tpHL = 36 psec
tpLH = 29 psec
so
tp = 32.5 psec
tf tr tpHL
tpLH
From simulation: tpHL = 39.9 psec and tpLH = 31.7 psec
Digital IC 32
Propagation delay: first order analysis
• Propagation delay model of RC
2→/ 21
OHOLOLOH
VVvVVv
)2
--()(
)9
7-1(
4
3≈
)1(I
dv1
'
2
1 sat12
n
nn
n
DSATTDDDSATnnsat
DD
DSAT
DD
v
v
eq
VVVVk
LWI
VI
V
VvvR
l
l
Assuming transistor as saturation
Digital IC 33
Propagation delay: first order analysis
• Propagation delay model of RC
L
DSATp
DDeqppHL C
I
VCRt )V-1(
4
3*69.0)2ln( DDl
L
DSATn
DDeqnpLH C
I
VCRt )V-1(
4
3*69.0)2ln( DDl
)
)/2-('**
1
)/2-('**
1(*52.0≈
*)11
(**2/75.0*69.0
2*69.0
2
ppppp
pnnnn DSATGTDSATDSATGTDSAT
n
nDDL
L
dsatpdsatn
DD
L
eqpeqnpLHpHL
p
VVVkL
WVVVk
LW
VC
CII
V
CRRtt
t
Digital IC
Inverter Propagation Delay,
Revisited
To see how a designer can optimize the delay of a
gate have to expand the Req in the delay equation
1
1.5
2
2.5
3
3.5
4
4.5
5
5.5
0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4
VDD (V)
tpHL = 0.69 Reqn CL
= 0.69 (3/4 (CL VDD)/IDSATn )
0.52 CL / (W/Ln k’n VDSATn )
Digital IC 35
Design techniques for minimized
propagation delay
• Reduce CL
• Keep the drain diffusion
areas as small as possible
• Increase the W/L ratio of
the transistor
• Increase Vdd
)1()(
*52.0≈
'
DDDSATnn
LpHL
VVkL
W
Ct
nl
Digital IC
Design for Performance
• Reduce CL
• internal diffusion capacitance of the gate itself
• keep the drain diffusion as small as possible
• interconnect capacitance
• fanout
• Increase W/L ratio of the transistor
• the most powerful and effective performance optimization tool in the
hands of the designer
• watch out for self-loading! – when the intrinsic capacitance dominates
the extrinsic load
• Increase VDD
• can trade-off energy for performance
• increasing VDD above a certain level yields only very minimal
improvements
Digital IC 37
Define NMOS-to-PMOS ratio
In order to create an inverter with
a symmetrical propagate delays
Also create symmetrical VTC β =2.4 which Rn=Rp!
nDSATnn
L
neqLpHLVk
LW
CRCt
')(2)2ln(
p
p
DSATpp
LeqLpLH
VkL
W
CRCt
')(2)2ln(
)2
(
)2
(
)(
)(
'
'
p
p
n
n
DSAT
TpMDDDSATp
DSAT
TnMDSATn
n
p
VVVVVk
VVVVk
LW
LW
1)(
)(
'
'
p
n
n
p
DSATpp
DSATnn
eq
eq
VkL
W
VkL
W
R
R
Digital IC 38
1 1.5 2 2.5 3 3.5 4 4.5 53
3.5
4
4.5
5x 10
-11
b
t p(s
ec)
tpLH tpHL
tp
b = Wp/Wn
Which point is optimal delay?
p
p
DSATpp
LeqLpLH
VkL
W
CRCt
')(2)2ln(
nDSATnn
L
neqLpHLVk
LW
CRCt
')(2)2ln(
Digital IC 39
Which point is optimal delay?
V DD V DD
V in V out
M 1
M 2
M 3
M 4 C db 2
C db 1
C gd 12
C w
C g 4
C g 3
V out 2
Interconnect
n
p
wgdwggddL
LW
LW
CCCCCCCCCnnpnpn
)(
)(
))(1(212211
b
b
)1()))(1((345.0
))())(1((2
2ln
2
21
21
b
b
bb
nnn
p
nnn
eqwgd
eq
eqwgd
pHLpLH
p
RCCC
RRCCC
ttt
Digital IC 40
Which point is optimal delay?
This r is different from before! It is the
resistor rate of the NMOS and PMOS
0
b
pt
0
)]1()))(1((345.0[21
b
b
b
nnn eqwgd RCCC
54.113
31≈)1(
25.05.2
21
um
VVdd
gndn
w
CC
Cb
Digital IC 41
Summary of ratio
Beta=1.6, we have minimum delay
Beta=2.4, we have equal delay tphl=tplh
Beta=3.5, we have VM=Vdd/2
Digital IC 42
CMOS dynamic characteristic
• CMOS capacitances mosaic
• CMOS propagation delay
• Optimizing inverter sizing
Digital IC
iref
extpp
irefextirefref
intextinteqextinteqp
SC
Ctt
SCCSCSR
CCCRCCRt
00 69.069.0
)1)()((69.0
)1(69.0)(69.0
43
Increasing inverter performance by sizing the
NMOS and PMOS
Intrinsic delay is independent
of the sizing of the gate
S>>0 will eliminate the
impact of any external load If no load
If load
Digital IC 44
2 4 6 8 10 12 142
2.2
2.4
2.6
2.8
3
3.2
3.4
3.6
3.8x 10
-11
S
t p(s
ec)
Device Sizing
(for fixed load)
Self-loading effect:
Intrinsic capacitances
dominate
Example 5.5
Digital IC 45
Inverter Chain
CL
If CL is given: - How many stages are needed to minimize the delay?
- How to size the inverters?
In Out
Digital IC 46
Inverter Delay
Minimum length devices, L=0.25mm, Assume
that for WP = 2WN =2W same pull-up and pull-down currents
approx. equal resistances RN = RP
approx. equal rise tpLH and fall tpHL delays
Analyze as an RC network
WNunit
Nunit
unit
PunitP RR
W
WR
W
WRR
11
tpHL = (ln 2) RNCL tpLH = (ln 2) RPCL Delay (D):
2W
W
unit
unit
gin CW
WC 3Load for the next stage:
Digital IC 47
Inverter with Load
Load (CL)
Delay
Assumptions: no load -> zero delay
CL
tp = k RWCL
RW
RW
Wunit = 1
k is a constant, equal to 0.69
Digital IC 48
Inverter with Load
Load
Delay
Cint CL
Delay = kRW(Cint + CL) = kRWCint + kRWCL = Delay
(Internal) + Delay (Load)= kRW Cint(1+ CL /Cint)
CN = Cunit
CP = 2Cunit
2W
W
Digital IC 49
Delay Formula
/1/1
~
0int ftCCCkRt
CCRDelay
pintLWp
LintW
Cint = Cgin with 1
f = CL/Cgin - effective fanout
R = Runit/W ; Cint =WCunit
tp0 = 0.69RunitCunit
Digital IC 50
Apply to Inverter Chain
CL
In Out
1 2 N
tp = tp1 + tp2 + …+ tpN
jgin
jgin
unitunitpjC
CCRt
,
1,1~
LNgin
N
i jgin
jgin
p
N
j
jpp CC C
Cttt
1,
1 ,
1,
0
1
, ,1
Digital IC
Optimal Tapering for Given N
Delay equation has N - 1 unknowns, Cgin,2 – Cgin,N
Minimize the delay, find N - 1 partial derivatives Result:
Cgin,j+1/Cgin,j = Cgin,j/Cgin,j-1
Size of each stage is the geometric mean of two
neighbors
each stage has the same effective fanout (Cout/Cin)
each stage has the same delay
51
1,1,, jginjginjgin CCC
Digital IC 52
Optimum size for fixed Number
of Stages
When each stage is sized by f and has same eff. fanout f:
Minimum path delay
Effective fanout of each stage:
1,/ ginL
N CCFf
N Ff
/10N
pp FNtt
Digital IC 53
Example
CL= 8 C1
In Out
C1 1 f f2
283 f
CL/C1 has to be evenly distributed across N = 3 stages:
CL/C1 has to be evenly distributed across N = 4 stages:
?84 f
Digital IC 54
Optimum Number of Stages
For a given load, CL and given input
capacitance Cin Find optimal sizing f
ftf
FFNtt p
N
pp 0
/1
0ln
ln1/
0ln
1lnln2
0
f
ffFt
f
t pp
For = 0, f = e, N = lnF
f
FNCfCFC in
N
inLln
ln with
fef 1
Digital IC 55
Optimum Effective Fanout f
Optimum f for given process defined by
fef 1
fopt = 3.6[4] for =1
Digital IC 56
Impact of Self-Loading on tp
1.0 3.0 5.0 7.0u
0.0
20.0
40.0
60.0
u/l
n(u
)
x=10
x=100
x=1000
x=10,000
No Self-Loading, =0 With Self-Loading =1
0
1
2
3
4
5
6
7
1 1.5 2 2.5 3 3.5 4 4.5 5
Digital IC 57
Normalized delay function of F
( ) Fln6.3=)6.3+1(Fln78.0=γ/F+1Nt=t N0pp
F unbuffered Two stages Inverter chain
10 11 8.3 8.3
100 101 22 16.6
1000 1001 65 24.9
10000 10001 202 33
Fln78.0=6.3ln
Fln=
fln
Fln=N,F=f N
Digital IC 58
Buffer Design
1
1
1
1
8
64
64
64
64
4
2.8 8
16
22.6
N f tp
1 64 65
2 8 18
3 4 15
4 2.8 15.3
Digital IC
More general example
Slide 59
1 2 3 CLoutin
( )γ/f+1Nt=t 0pp
3
3
3
L
2
3
1
2
3
L
2
3
1
2
1
L
3
L
2
3
1
2
164=f
f16
1=
C
C
C
C4
C
C4
16
1=
C
C
C
C
C
C=
C
C=F
f=C
C=
C
C4=
C
C4
13
13
33
2
13
13
L
3
C16=C164
44*4=
f
C4=C
C44=C164
64=
f
C=C
Digital IC
Design Challenge
• A gate is never designed in isolation: its performance
is affected by both the fan-out and the driving strength
of the gate(s) feeding its inputs.
( 0.25)
• Keep signal rise times smaller than or equal to the gate
propagation delays
• good for performance
• good for power consumption
• Keeping rise and fall times of the signals small and of
approx. equal values is one of the major challenges in
high-performance designs(slope engineering.)
1-iii
stepstepptη+t=t
Digital IC
Input Signal Rise/Fall Time
• In reality, the input signal
changes gradually (and both
PMOS and NMOS conduct
for a brief time). This affects
the current available for
charging/discharging CL and
impacts propagation delay.
3.6
3.8
4
4.2
4.4
4.6
4.8
5
5.2
5.4
0 2 4 6 8ts(sec)
x 10-11
x 10-11
for a minimum-size inverter with a fan-
out of a single gate
tp increases linearly with increasing input slope, ts, once ts > tp
ts is due to the limited driving capability of the preceding gate
Digital IC
Rising-fall time of the input signal
Slide 62
1-iii
stepstepptη+t=t
Note: tp increases linearly with increasing input
slope,once ts>tp(ts=0)
1 2 3 CLoutin