chapter 5access.ee.ntu.edu.tw/course/vlsi_design_92first/ppt/chapter5 10-29 … · objective of...
TRANSCRIPT
-
Inverter
Chapter 5Chapter 5
The InverterThe InverterV1. April 10, 03V1.1 April 25, 03
-
Inverter
Objective of This ChapterObjective of This Chapter
Use Inverter to know basic CMOS Circuits OperationsWatch for performance Index such as
Speed (Delay calculation)Optimal Transistor Sizing for speed and EnergyPower Consumption and Dissipation
-
Inverter
The CMOS Inverter: A First GlanceThe CMOS Inverter: A First Glance
Vin Vout
CL
VDD
-
Inverter
CMOS InverterCMOS Inverter
Polysilicon
In Out
VDD
GND
PMOS
Metal 1
NMOS
OutIn
VDD
PMOS
NMOS
Contacts
N Well
-
Inverter
Two InvertersTwo Inverters
Connect in Metal
Share power and ground
Abut cells
VDD
Vin Vout
Vin
Vout
-
Inverter
CMOS InverterCMOS InverterFirstFirst--Order DC AnalysisOrder DC Analysis
VOL = 0VOH = VDD
VDD VDD
Vin = VDD Vin = 0
VoutVout
Rn
Rp
-
Inverter
Delay DefinitionsDelay Definitions
Vout
tf
tpHL tpLH
trt
Vin
t
90%
10%
50%
50%
-
Inverter
CMOS Inverter: Transient ResponseCMOS Inverter: Transient Response
tpHL = f(Ron.CL)= 0.69 RonCL
V outVout
R n
R p
V DDV DD
V in = V DDV in = 0
(a) Low-to-high (b) High-to-low
CLCL
ln(2)=0.69
-
Inverter
Voltage TransferVoltage TransferCharacteristicCharacteristic
-
Inverter
PMOS Load LinesPMOS Load Lines
VDS,p
IDp
VGSp=-2.5
VGSp=-1VDS,p
IDnVin=0
Vin=1.5
Vout
IDnVin=0
Vin=1.5
V in = V DD +VGSpIDn = - I Dp
Vout = V DD+VDSp
Vout
IDn
V in= V DD+VGS,pID,n= - I D,p
V out= V DD+VDS,p
(Vdd = 2.5V in 0.25um CMOS Process)(Vt = 0.4V as shown in Table 3-2)
-
Inverter
CMOS Inverter Load CharacteristicsCMOS Inverter Load Characteristics
IDn
Vout
Vin = 2.5
Vin = 2
Vin = 1.5
Vin = 0
Vin = 0.5
Vin = 1
NMOS
Vin = 0
Vin = 0.5
Vin = 1Vin = 1.5
Vin = 2
Vin = 2.5
Vin = 1Vin = 1.5
PMOS
-
Inverter
CMOS Inverter VTCCMOS Inverter VTC
Vout
Vin0.5 1 1.5 2 2.5
0.5
11.
52
2.5
NMOS resPMOS off
NMOS satPMOS sat
NMOS offPMOS res
NMOS satPMOS res
NMOS resPMOS sat
VM: Vin = VoutSwitching Threshold Voltage
-
Inverter
Switching Threshold as a Function of Switching Threshold as a Function of Transistor RatioTransistor Ratio
NMOS and PMOS are in Saturation Modes
For r = 1, and saturated velocity NMOS = 2 PMOS, Wp = 2Wn
),,when (1 TpTnDSATDD
DDM VVVVr
rVV >>+
≈
-
Inverter
Switching Threshold as a Function of Switching Threshold as a Function of Transistor RatioTransistor Ratio
100
101
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8M
V(V
)
W p/W n
-
Inverter
Simulated VTCSimulated VTC
0 0.5 1 1.5 2 2.50
0.5
1
1.5
2
2.5
Vin
(V)
Vou
t(V)
-
Inverter
Impact of Process VariationsImpact of Process Variations
0 0.5 1 1.5 2 2.50
0.5
1
1.5
2
2.5
Vin (V)
V out
(V)
Good PMOSBad NMOS
Good NMOSBad PMOS
Nominal
Good: Smaller oxide thickness, smaller L, higher W, smaller VT
-
Inverter
Propagation DelayPropagation Delay
-
Inverter
CMOS InvertersCMOS Inverters
Polysilicon
InOut
Metal1
VDD
GND
PMOS
NMOS
1.2 µm=2λ
-
Inverter
CMOS Inverter Propagation DelayCMOS Inverter Propagation Delay
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
-
Inverter
The Transistor as a SwitchThe Transistor as a SwitchVGS ≥ VT
RonS D
ID
VDS
VGS = VD D
VDD/2 VDD
R0
Rmid
-
Inverter
The Transistor as a SwitchThe Transistor as a Switch
0.5 1 1.5 2 2.50
1
2
3
4
5
6
7x 10
5
VDD
(V)
Req
(Ohm
)
-
Inverter
The Transistor as a SwitchThe Transistor as a Switch
-
Inverter
0 0.5 1 1.5 2 2.5
x 10-10
-0.5
0
0.5
1
1.5
2
2.5
3
t (sec)
Vou
t(V)
Transient ResponseTransient Response
tp = 0.69 CL (Reqn+Reqp)/2
?
tpLH tpHL
-
Inverter
Design for PerformanceDesign for Performance
Keep loading capacitances (CL) smallIncrease transistor sizes (add CMOS gain)
Watch out for self-loading (for the previous stage)!
Increase VDD (????) Power consumption??
-
Inverter
Delay (speed degrade) as a function of VDelay (speed degrade) as a function of VDDDD
0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.41
1.5
2
2.5
3
3.5
4
4.5
5
5.5
VDD
(V)
t p(n
orm
aliz
ed)
-
Inverter
-
Inverter
Propagation delay v.s. Transistor Propagation delay v.s. Transistor sizesize
NMOS-to-PMOS ratio: will symmetrical tpHL and tpLH have better propagation delay (tp), or a better ratio for overall tp can be found?
Consider two identical cascaded CMOS inverters. The approximated load cap of the 1st gate is
WgngpdndpL CCCCCC ++++= )()( 2211
-
Inverter
Propagation delay v.s. Transistor sizePropagation delay v.s. Transistor size
When the PMOS device is made β times larger than the NMOS ThenCL becomes From (5.20), we have
where is the resistance ratio of equal-size NMOS
and PMOS
n
p
LWLW
)/()/(
=β
2211 and gngpdndp CCCC ββ ≈≈WgndnL CCCC +++= ))(1( 21β
[ ]
[ ] )1())(1(345.0
)())(1(269.0
21
21
βγβ
ββ
++++=
++++=
eqnWgndn
eqpeqnWgndnp
RCCC
RRCCCt
eqn
eqp
RR
=γ
-
Inverter
1 1.5 2 2.5 3 3.5 4 4.5 53
3.5
4
4.5
5x 10
-11
β
t p(s
ec)
NMOS/PMOS ratioNMOS/PMOS ratio
tpLH tpHL
tp β = Wp/Wn
(See pp. 204)
Fig. 5-18
)(1(21 gndn
Wopt CC
C+
+= γβ
-
Inverter
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 SizingDevice Sizing
(for fixed load)
Self-loading effect:Intrinsic capacitancesdominate
(Fig. 5-19)
-
Inverter
Inverter SizingInverter Sizing
-
Inverter
Inverter ChainInverter Chain
CL
If CL is given:- How many stages are needed to minimize the delay?- How to size the inverters?
May need some additional constraints.
In Out
1 f2f1
-
Inverter
Inverter DelayInverter Delay• Minimum length devices, L=0.25µm• 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 RRW
WRWWRR ==
≈
=
−− 11
tpHL = (ln 2) RNCL tpLH = (ln 2) RPCLDelay (D):
2W
W
unitunit
gin CWWC 3=Load for the next stage:
-
Inverter
Inverter with LoadInverter with Load
Load (CL)
Delay
CL
tp = k RWCL
RW
RW
•k is a constant, equal to 0.69•Assumptions: no load zero delay
2W
W
-
Inverter
Inverter with Load and Para. Cap.Inverter with Load and Para. Cap.
Load
Delay
Cint CL
Delay = kRW (Cint + CL) = kRWCint + kRWCL = kRW Cint(1+ CL /Cint)= Delay (Internal) + Delay (Load)
CN = Cunit
CP = 2Cunit
2W
W
-
Inverter
Delay FormulaDelay Formula( )
( ) ( )γ/1/1
~
0int ftCCCkRt
CCRDelay
pintLWp
LintW
+=+=
+
Cint = γCgin with γ ≈ 1f = CL/Cgin : Effective fanout
RW = Runit / W ; Cint =WCunittp0 = 0.69RunitCunit
-
Inverter
Apply to Inverter ChainApply to Inverter Chain
CL
In Out
1 2 N
tp = tp1 + tp2 + …+ tpN
+ +
jgin
jginunitunitpj C
CCRt
,
1,1~γ
LNgin
N
i jgin
jginp
N
jjpp CC C
Cttt =
+== +
=
+
=∑∑ 1,
1 ,
1,0
1, ,1 γ
-
Inverter
Optimal Tapering for Given NOptimal Tapering for Given NDelay 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
1,1,, +−= jginjginjgin CCC
-
Inverter
Optimum Delay and Number of StagesOptimum Delay and Number of Stages
1,/ ginLN CCFf ==
When each stage is sized by f and has same effective fanout f
N Ff =
( )γ/10 Npp FNtt +=Minimum path delay
Effective fanout of each stage:
-
Inverter
ExampleExample
CL= 8 C1
In Out
C1 1 f f2
283 ==f
CL/C1 has to be evenly distributed across N = 3 stages:
-
Inverter
Optimum Number of StagesOptimum Number of StagesFor a given load, CL and given input capacitance CinFind optimal sizing f
( )
+=+=
fffFtFNtt pNpp lnln
ln1/ 0/10
γγ
γ
0ln
1lnln2
0 =−−
⋅=∂
∂
fffFt
ft pp γ
γ
For γ = 0, f = e, N = lnF
fFNCfCFC in
NinL ln
ln with ==⋅=
( )ff γ+= 1exp
-
Inverter
Optimum Effective Optimum Effective Fanout Fanout ffOptimum f for given process defined by γ
( )ff γ+= 1expfopt = 3.6for γ=1
fopt = 2.718for γ=0
-
Inverter
Normalized delay function of Normalized delay function of FF
( )γ/10 Npp FNtt +=
-
Inverter
Buffer DesignBuffer 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
Without considering the internal capacitance
-
Inverter
Power DissipationPower Dissipation
-
Inverter
Where Does Power Go in CMOS?Where Does Power Go in CMOS?
• Dynamic Power Consumption
• Short Circuit Currents
• Leakage
Charging and Discharging Capacitors
Short Circuit Path between Supply Rails during Switching
Leaking diodes and transistors
-
Inverter
Dynamic Power DissipationDynamic Power Dissipation
Energy/transition = CL * Vdd2
Power = Energy/transition * f = CL * Vdd2 * f
Need to reduce CL, Vdd, and f to reduce power.
Vin Vout
CL
Vdd
Not a function of transistor sizes!
2)(
2
00
DDLout
outLoutVDDC
VCdtvdt
dvCdtvtiE ∫∫∞∞
=== Energy in CL
-
Inverter
Node Transition Activity and PowerNode Transition Activity and PowerConsider switching a CMOS gate for N clock cycles
EN CL Vdd•2 n N( )•=
n(N): the number of 0->1 transition in N clock cycles
EN : the energy consumed for N clock cycles
Pavg N ∞→lim
ENN-------- fclk•=
n N( )N------------N ∞→
lim C•
LVdd•
2 fclk•=
α0 1→n N( )
N------------N ∞→lim=
Pavg = α0 1→ C• LVdd•
2 fclk•
e)Capacitanc Effective:(
)( 2210
Eff
CLKDDEffCLKDDLAVG
C
fVCfVCP ⋅⋅=⋅⋅⋅= →α
-
Inverter
Switching Activity (Example 5.12)Switching Activity (Example 5.12)
25.08/210 ==→α
-
Inverter
Transistor Sizing for Minimum EnergyTransistor Sizing for Minimum EnergyGoal: Minimize Energy of whole circuit while maintaining the speed speed performance
Design parameters: f and VDDtp ≤ tp,ref of referenced circuit with f=1 and Vdd =Vref
1Cg1
In
fCext
Out
TEDD
DDp
pp
VVVt
fFftt
−∝
++
+=
0
0 11 γγ)/( 1gext CCF =
)2/( DSATTTE VVV +=
-
Inverter
Transistor Sizing (2)Transistor Sizing (2)Performance Constraint (γ=1) Vdd(f)
Energy for single transition
Energy ratio of the design and reference circuit
( ) ( ) 13
2
3
2
0
0 =+
++
−
−=
+
++
=F
fFf
VVVV
VV
FfFf
tt
tt
TEDD
TEref
ref
DD
refp
p
pref
p
+++
=
FFf
VfV
EE
ref
DD
ref 422)(
2
[ ] ])1)(1[(1 1212 FfCVFffCVE gDDgDD +++=++++= γγγ
-
Inverter
1 2 3 4 5 6 70
0.5
1
1.5
2
2.5
3
3.5
4
f
vdd
(V)
1 2 3 4 5 6 70
0.5
1
1.5
f
norm
aliz
ed e
nerg
y
Transistor Sizing (4)Transistor Sizing (4)
F=1
2
5
10
20
VDD=f(f) E/Eref=f(f)
Required Supply Voltage Energy v.s. Sizing factor
-
Inverter
Sizing factor for Speed and EnergySizing factor for Speed and Energy
Device sizing, combined with supply voltage reduction, is a very effective way in reducing energy consumption of a logic network.
The gain can be up to 10 for large fanout.
Oversizing beyond the optimal value comes at a hefty price in energy.Optimal size for energy is smaller than the optimal sizing for performance.
For example, f(energy) = 3.53, f(performance) = 4.47= , for F=20
20
-
Inverter
Short Circuit Currents (during switching)Short Circuit Currents (during switching)
Vin Vout
CL
Vdd
I VD
D (m
A)
0.15
0.10
0.05
Vin (V)5.04.03.02.01.00.0
-
Inverter
Minimizing ShortMinimizing Short--Circuit PowerCircuit Power
-
Inverter
Neil Weste Textbook
-
Inverter
Leakage CurrentLeakage Current
Vout = Vdd
Vdd
Sub-ThresholdCurrent
Drain JunctionLeakage Current
Sub-threshold current is one of most compelling issuesin low-energy circuit design!!
Vdd
-
Inverter
ReverseReverse--Biased Diode LeakageBiased Diode Leakage
Np+ p
+
Reverse Leakage Current
+
-Vdd
GATE
IDL = JS × A
JS = 10-100 pA/µm2 at 25 deg C for 0.25µm CMOSJS doubles for every 9 deg C!
-
Inverter
SubthresholdSubthreshold Leakage ComponentLeakage Component
-
Inverter
Static Power ConsumptionStatic Power Consumption
Vin=5V
Vout
CL
Vd d
Istat
Pstat = P(In=1).Vdd . Istat
Wasted energy …Should be avoided in almost all cases.
-
Inverter
Principles for Power ReductionPrinciples for Power ReductionPrime choice: Reduce voltage!
Recent years have seen an acceleration in supply voltage reductionDesign at very low voltages still open question (0.6, … , 0.9 V by 2010!)
Reduce switching activity (at different levels)Reduce physical capacitance
Device Sizing: for example, for F = 20fopt(energy)=3.53, fopt(performance)=4.47.