0% found this document useful (0 votes)
29 views64 pages

B Logic

The document discusses combinational logic circuits and complementary CMOS design. It covers static and dynamic properties of CMOS, design techniques such as transistor sizing and input reordering, and optimizing delay. The delay of a complex gate depends on its intrinsic delay, electrical effort determined by output capacitance, and effort determined by the gate's topology and layout. Optimizing involves balancing these factors.

Uploaded by

wqy15902896758
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)
29 views64 pages

B Logic

The document discusses combinational logic circuits and complementary CMOS design. It covers static and dynamic properties of CMOS, design techniques such as transistor sizing and input reordering, and optimizing delay. The delay of a complex gate depends on its intrinsic delay, electrical effort determined by output capacitance, and effort determined by the gate's topology and layout. Optimizing involves balancing these factors.

Uploaded by

wqy15902896758
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/ 64

数字集成电路原理

Combinational Logic Circuits

刘佳欣
[email protected]
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

2
Logic Circuits

3
Combinational Logic Circuits

Complementary CMOS

(1) Static comb. logic Ratioed Logic

Pass-Transistor Logic

(2) Dynamic comb. logic

4
Static VS. Dynamic Logic
pStatic Logic
Ø At every point in time (except during
the switching transients), each gate
output is connected to either VDD or
VSS via a low-resistive path.
Ø The outputs of the gates assume
the value of the Boolean function.

pDynamic Logic
Ø Relies on temporary storage of
signal values on the capacitance of
high impedance circuit nodes.

5
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

6
Static Complementary CMOS

VDD
PUN:
In1 Ø Make a connection from VDD to F
In2 PUN when F(In1,In2, ... Inn) = 1

Ø PMOS only
InN
F(In1,In2,…InN)
In1
In2 PDN PDN:

Ø Make a connection from VSS to F


InN
when F(In1,In2, ... Inn) = 0
Ø NMOS only

PUN and PDN are dual (Complementary) logic networks

7
Construct the PUN and PDN Networks
p Threshold drop issues

VDD VDD
PUN
S D
Good! VDD Bad!

D 0  VDD S 0  VDD - VTn


VGS
CL CL

PDN VDD  0 VDD  |VTp|


VGS
D CL S CL
VDD

S Good! D Bad!

8
Construct the PDN Networks
p Transistors can be thought as a switch
p NMOS switch closes when its gate is high

A B

X Y Y = X if A and B

X B Y = X if A OR B
Y

NMOS Transistors pass a “strong” 0 but a “weak” 1

9
Construct the PUN Networks
p PMOS switch closes when its gate is low

10
Complementary CMOS Logic Style

11
Example Gate: NAND

12
Complex CMOS Gate

B
A
C

D
F = D + A • (B + C)
A
D
B C

13
Constructing a Complex Gate

VD D VD D

C
S N F1 S N 4 A
F
S N 2 B
A A
D D S N 3

B C B C D

( a ) l-d
wn
o n A e
D
s u b - ts
B C

( c )

14
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

15
Static Behavior
VTC of inverter
p The VTC of a complementary CMOS
gate is data-dependent

VTC of NAND2

16
Dynamic Behavior

q Delay is dependent on the inputs

q Low to high transition


§ both inputs go low
– delay is 0.69 Rp/2 CL
§ one input goes low
– delay is 0.69 Rp CL
q High to low transition
§ both inputs go high
– delay is 0.69 2Rn CL

17
Delay Dependence on Input Patterns

NMOS = 0.5m/0.25 m
PMOS = 0.75m/0.25 m
CL = 100 fF

18
Fan-In Considerations

19
Fan-In Considerations

1250

Gates with a fan- quadratic


1000
in greater than 4
should be
750 avoided.
tpHL tp
500

250 tpLH
linear
0
2 4 6 8 10 12 14 16

Propagation delay of CMOS NAND gate as a function of fan-in

20
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

21
Fast Complex Gates: Design Technique 1

q Transistor sizing
§ as long as fan-out capacitance dominates
q Progressive sizing

InN MN CL
M1 > M2 > M3 > … > MN
(the fet closest to the
output is the smallest)
In3 M3 C3
In2 M2 C2
In1 M1 C1

22
Fast Complex Gates: Design Technique 1

q Transistor sizing
§ as long as fan-out capacitance dominates
q Progressive sizing

InN MN CL M1 > M2 > M3 > … > MN


(the fet closest to the
output is the smallest)
In3 M3 C3
In2 M2 C2
In1 M1 C1

23
Fast Complex Gates: Design Technique 2

q Input re-ordering
critical path critical path

charged 01
In3 1 M3 CL In1 M3 CLcharged

In2 1 M2 In2 1 M2 C2 discharged


C2 charged
In1 In3 1 M1 C1 discharged
M1 C1 charged
01

delay determined by time to delay determined by time to


discharge CL, C1 and C2 discharge CL

24
Fast Complex Gates: Design Technique 3

q Alternative logic structures

F = ABCDEFGH

25
Fast Complex Gates: Design Technique 4

q Isolating fan-in from fan-out using buffer


insertion

CL CL

26
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

27
Review: Optimize the Dealy of Inverter Chain

pFor a given stage number N


C L 1/ N
f opt ( )
C g1
pIf N is not fixed

���� ≈ 4, �ℎ�� ��������� �

q Can we extend this design rule to a comb. gate?


28
Delay of Complex Gate
 CL 
pInverter delay t p,inv  t p 0 1    t p 0 1  f /  
  Cin 

pComplex gate delay t p  t p 0  p  gf  Assume γ=1

Ø h=gf: effort
Ø p: intrinsic (parasitic) delay
• Ratio of the intrinsic delays of the complex gate and the inverter
• p=1 for inverter
• Set by circuit topology and layout
Ø f: electrical effort = Cout/Cin
• Ratio of output to input capacitance
• We call it effective fanout in inverter
Ø g: logical effort
• presents the ratio of its input capacitance to the inverter
capacitance when sized to deliver the same current
• g=1 for inverter
• Depends only on circuit topology only
29
Intrinsic Delay

• Ignore the internal


capacitance

30
Logical Effort
pPresents the ratio of its input capacitance to the inverter
capacitance when sized to deliver the same current
pFor a given load, complex gates have to work harder than
an inverter to produce a similar response.
pCan estimate by counting transistor widths

31
Logical Effort

32
Delay: NAND2 VS. INV

t p  t p 0  p  gf  • d: normalized delay
d   p  gf    p  h • h: gate effort
33
Example: FO4 Inverter

pEstimate the delay of a fanout-of-4 (FO4) inverter

• Intrinsic Delay: p=1


• Logical Effort: g=1
• Electrical Effort: f=Cout/Cin=4
• Gate Delay: d=5

34
Multistage Logic Networks

N N

Optimal f for inverter chain D    pi  g i  f i     pi  hi 


N i 1 i 1
tp   t 1  f 
h1  h2  ...  hN
p0 i
1

f1  f 2  ...  f N  N
F

pEvery stage should have the same effort!


35
Multistage Logic Networks

N N N
1

Path effort: H  h   f  g
i 1
i
i 1
i
i 1
i  F G h1  h2  ...  hN  H N

N
Path electrical effort: F  f
i 1
i  C out _ path / C in _ path
N
Path logic effort: G  g
i 1
i

Path intrinsic delay: P


1
Path effort delay: D H  N  H N

Path delay: D   p i   hi  P  D H
36
Multistage Logic Networks

N N N
f  C out _ gate / C in _ gate
H   hi   f i   g i  F  G C in _ gate  C out _ gate / f
i 1 i 1 i 1

F  Cout / Cin  5 f1  h / g 1  1.93


c  5/ f 4  2.59
G  25 / 9 f 2  h / g 2  1.16
b  c / f 3  2.23
H  125 / 9 f 3  h / g 3  1.16
a  b / f 2  1.93
h 4
H  1.93 f 4  h / g 4  1.93

37
Multistage Logic Networks

N N N
H   hi   f i   g i  F  G
i 1 i 1 i 1

F  Cout / Cin  5 c  1  5/h  2.59


G  25 / 9 h  g  f  g  C out _ gate / C in _ gate
5
H  125 / 9 C in _ gate  g  C out _ gate / h b  c / h  2.23
3
h 4
H  1.93 5
a   b / h  1.93
3
38
Multistage Logic with Branches

H≠FG

p Introduce the branching effort:

Con  path  Coff  path


Gate branching effort : b 
Con  path
Path branching effort : B   bi
Path effort:H  FGB

39
Example: Sizing an Inverter Network
pDetermine the min delay form in to out, CL=64Cin
Ø Path electrical effort:
F=CL/Cin=64

Ø Path logical effort:


G=1

Ø Path branching effort:


B=4*4=16

Ø Path effort:
H=FGB=1024

Ø Optimal gate effort:


h =H1/N ≈ 10.08

Ø Path effort delay:


DH=N*h≈ 30.24
40
Example: Sizing an Inverter Network
pDetermine the min delay form in to out, CL=64Cin

Ø Gate intrinsic delay:


p=1

Ø Path intrinsic delay:


P=3

Path delay: D  P+N  h  3+3 10.08  33.24

41
Review of Definitions

Term Gate/Stage Path

Logical effort g �= ��
Electrical effort f=Cout/Cin F=Cout(patch)/Cin(path)

Branching effort b=(Con_path+Coff_patch)/Con_path �= ��


Effort h=gf H=GFB

Effort delay h �� = ℎ�
Number of stages 1 N

Intrinsic delay p �= ��
Delay d=p+h D=P+DH

42
Review of Definitions

f <-> h

F <-> H

43
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

44
Ratioed Logic: Concept

VDD VDD VDD

Resistive Depletion PMOS


Load RL Load VT < 0 Load
VSS
F F F
In1 In1 In1
In2 PDN In2 PDN In2 PDN
In3 In3 In3

VSS VSS VSS


(a) resistive load (b) depletion load NMOS (c) pseudo-NMOS

Goal: to reduce the number of devices over complementary CMOS

45
Resistive Load

46
Pseudo-NMOS Load

47
Pseudo-NMOS Inverter VTC

p Voltage-transfer curves of the pseudo-NMOS inverter as a function of the


PMOS size (with an NMOS size of 0.5um/0.25um)

48
Improved Load

VDD VDD
Ø Differential logic
Ø Positive feedback
M1 M2

Out Out

A
A PDN1 PDN2
B
B Ø Rail-to-rail
Ø No static power
VSS VSS

Differential Cascode Voltage Switch Logic (DCVSL)

49
Improved Load
VDD VDD

M1 M2 Ø Differential logic
Ø Positive feedback
Out Out

A
A PDN1 PDN2
B
B
Ø Rail-to-rail
VSS VSS
Ø No static power
Differential Cascode Voltage Switch Logic (DCVSL)

Ø Ratioed problem: Out (1->0)


Ø PDNs need to be stronger than load
50
DCVSL Example: XOR/NXOR

Out

Out

B B B B

A A
CMOS XOR Logic

XOR-NXOR gate

51
DCVSL Response: AND/NAND

2.5

AB

V olta ge [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]

52
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

53
Pass-Transistor Logic

54
NMOS-Only Pass-Transistor Logic
B

A
pExample: AND gate B
F A = B

x does not pull up to 2.5V, but 2.5V - VTN


Threshold voltage loss causes static power consumption
55
Solution 1: Level Restoring Transistor

p Full Swing
p No static power in INV

p Restorer adds capacitance, takes away pull down current at X


p Ratioed problem: Mn needs to be stronger than Mr to pull down X
56
Restorer Sizing

57
Solution 2: Multiple Threshold Transistor

58
Solution 3: Transmission Gate

pTransmission gate work as switch


pNeed 2 transistors, and more control logic
pRail to rail swing

59
Solution 3: Transmission Gate Based Logic

CMOS Logic: 8T CMOS Logic: 12T


60
Outline
pLogic Circuits Classification
pComplementary CMOS
pConcept
p Static & Dynamic Properties
p Design Techniques
p Effort
pRatioed Logic
pPass-Transistor Logic
pDynamic CMOS Design

61
Dynamic Gate:

Clk Mp Clk Mp
Out Out
In1 CL
A
In2 PDN
C
In3
B
Clk Me
Clk Me
Two phase operation
Precharge (CLK = 0)
Evaluate (CLK = 1)

62
Properties of Dynamic Gate

p Logic function is implemented by the PDN (or PUN, but barely use)
only

p Fewer transistors: N + 2 transistors for fan-in-N


p Full swing outputs (VOL = GND and VOH = VDD)
p Non-ratioed - sizing of the devices does not affect the logic levels
p Faster switching speeds
• reduced load capacitance due to lower input capacitance (Cin)
• reduced load capacitance due to smaller output loading (Cout)
• no Isc, so all the current provided by PDN goes into discharging CL

p Need careful design especially in advanced process nodes


• Due to leakage
p Possible higher power consumption although no Isc
• Frequent pull-up operations

63
Thank you!

You might also like