0% found this document useful (0 votes)
24 views

Lecture8 Ratioed

The document discusses ratioed logic and differential cascode voltage switch logic (DCVSL). It explains how DCVSL works using cross-coupled transistors and can eliminate one inverter delay compared to regular logic. Sizing transistors for DCVSL involves setting the output voltage less than or equal to the supply voltage plus the threshold voltage.

Uploaded by

조동올
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
24 views

Lecture8 Ratioed

The document discusses ratioed logic and differential cascode voltage switch logic (DCVSL). It explains how DCVSL works using cross-coupled transistors and can eliminate one inverter delay compared to regular logic. Sizing transistors for DCVSL involves setting the output voltage less than or equal to the supply voltage plus the threshold voltage.

Uploaded by

조동올
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 29

Lecture 8: Ratioed logic

Hyeon-Min Bae

Department of Electrical Engineering


KAIST, Daejeon, Korea

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 1


Ratioed Logic

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 2


Ratioed Logic

xVDD

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 3


Active Loads

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 4


Pseudo-NMOS

MP Vin=0 à Vout=VDD
Vin=VDDà VOL>0
vout
MN
vi

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 5


Pseudo-NMOS

Vin=VDDà VOL>0
MP IDP NMOS: Linear
PMOS: Vel. Sat
vout
MN IDN
vi

I DN + I DP = 0
2 2
VOL VDSAT
K N [(VDD - VTn )VOL - ] + K P [(-VDD - VTp )VDSAT - ]=0
2 2

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 6


Pseudo-NMOS

• Impact of Kn &Kp on VOL


• Assume VOL<<VDD-VTn ; |VDSAT|<<VDD+VTp

K N (VDD - VTn )VOL + K P [(-VDD - VTp )VDSAT ] = 0


K P [(VDD + VTp )VDSAT µ pW p
VOL = » VDSATp
K N (VDD - VTn ) µ nWn

VOL ¯ as Wp/Wn ¯ à PMOS Rp >> Rn

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 7


Pseudo-NMOS

• Pstatic =PDC=VDD x ILOW


• =VDDKPVDSAT[(-VDD-VTP)-VDSATp/2]
• Pstatic reduces as Kp àWp ¯
• VOL & Pstatic move in the same direction
• Why not simply make Wp = Wmin

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 8


Pseudo-NMOS VTC

3.0

2.5

2.0 W/Lp = 4

1.5
Vout [V]

W/Lp = 2
1.0

W/Lp = 0.5 W/Lp = 1


0.5

W/Lp = 0.25
0.0
0.0 0.5 1.0 1.5 2.0 2.5
Vin [V]

tPLH will suffer à Trade off between (VOL, Pstatic) vs tPLH

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 9


Pseudo-NMOS

Vin = VDD
MP IDP If Kp ↑ to reduce tPLH
Then more/increased contention
vout @ output node
MN IDN
vi Thus, tPHL ↑

Let’s verify in the next slide

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 10


Psueto-NMOS

Rising transient response


$
#% &
𝑉! 𝑡 = 𝑉"" (1 − 𝑒 ! )

Falling transient response


$
𝑉"" #
𝑉! 𝑡 = (𝑅( 𝑒 (%! ||%" )& + 𝑅' )
𝑅' + 𝑅(
% ,-%"
Time to reach -(%# ,% )
𝑉"" à (V(0)+V(final))/2
# "

T=0.69(𝑅( | 𝑅' 𝐶
Transient response is determined only by RC.
Large Kp increases total C. à delay increases
Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 11
Improved Loads

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 12


Improved Loads (2)

VDD VDD

M1 M2

Out Out

A
A PDN1 PDN2
B
B

VSS VSS

Differential Cascode Voltage Switch Logic (DCVSL)

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 13


Improved Loads (2)

VDD VDD

M1 M2

Out Out

A
A PDN1 PDN2
B
B

VSS VSS
PDN1 & PDN2 are mutually exclusive
Only 1 is on for any input combination
Vout and Vout are logical inverse
Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 14
Example AND/NAND

VNAND VAND
B

A A B

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 15


DCVSL Example

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 16


Why differential in logic (DCVSL)

• Can eliminate one inverter delay


• Note however in DCVSL usually tPHL¹tPLH and thus
differential outputs are not perfectly synchronized.
à careful design is required

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 17


DCVSL Transient Response

2.5

AB
V ol tage [V]

1.5
AB
A,B
0.5 A,B

-0.5 0 0.2 0.4 0.6 0.8 1.0


Time [ns]

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 18


The operation of DCVSL

MP1 MP2 PDN1 = ON,


PDN2 = OFF
Vo Vo
VDD

Vo ¯ à MP2 is on à MP1 is turned off (contention is removed)

VO=VDD VOL=0

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 19


DCVSL

• Rail to rail swing. Zero static power consumption


• Ratioed logic
• Short circuit power + dynamic power

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 20


How to size

Vout £ VDD + VTP


0
NFET à vel.sat
Vo
PFET à linear
VDD
Equation?

Do the same thing for PDN2


Keep MP1 and MP2 identical.

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 21


How to size

𝑉! ≤ 𝑉"" + 𝑉.(
0
NFET à vel.sat
Vo
PFET à linear
VDD

-
𝑉"01.
𝐾' 𝑉"" − 𝑉./ 𝑉"01. −
2
(𝑉3 − 𝑉"" )-
+𝐾( −𝑉"" − 𝑉.2 (𝑉! −𝑉"" ) − =0
2

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 22


Cross coupled transistors

Vy
Vinit,y
R: output resistance Vx R C
C: gate capacitance

Vx
Vinit,x
Vy R C

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 23


Latch – latching phase

Vy
g mVx + + sCLV y - C LV y ,init = 0 (1)
R
Vx
g mVy + + sCLVx - C LVx ,init = 0 (2)
R
By taking (1) – (2)

DV
g m DV - - sCL DV + CL DVinit = 0
R

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 24


Latch – latching phase

g m R -1
t
DV = DVinit e RC

For a large gain,

wT t
DV » DVinit e
Cross coupled positive feedback circuit converges with the
maximum possible rate if the gate capacitances are the
dominating ones

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 25


Transit Frequency

𝑔4
𝜔. =
𝐶50

• Transit frequency, fT, is defined as the


frequency where the current gain from input
to output drops to 1.

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 26


MOS device fT

Cgd
vgs
iout

iin Cgs gmvgs ro

iout gm
=
iin jw (Cgs + Cgd )
At unity gain frequency wT, iout=iin

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 27


MOS Device fT

gm
fT =
2p (Cgs + Cgd )
W ∆=Vgs-Vth
µCox D
Leff 3µD
= »
2
2p ( WLeff Cox + 2Col ) 4 pL eff
2

Increase overdrive and decrease gate length

Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 28


MOS Device fT

Ft
3µD
1.00E+12
fT =
4pLeff
2
Frequency (Hz)

1.00E+11

Mobility degradation
1.00E+10
0 0.2 0.4 0.6 0.8
Vgs

Mobility degradation from vertical field


µn
µeff =
1 + q (Vgs - Vth )
Copyright 2022, Nanoscale advanced integrated systems lab, KAIST 29

You might also like