• Nenhum resultado encontrado

Lecture 4: DC & Transient Response

N/A
N/A
Protected

Academic year: 2022

Share "Lecture 4: DC & Transient Response"

Copied!
71
0
0

Texto

(1)

Introduction to CMOS VLSI

Design

Lecture 4:

DC & Transient Response

Greco/Cin-UFPE

(Material taken/adapted from Harris’ lecture notes)

(2)

Outline

 DC Response

 Logic Levels and Noise Margins

 Transient Response

 Delay Estimation

(3)

Activity

1)     If the width of a transistor increases, the current will increase decrease not change

2)     If the length of a transistor increases, the current will increase decrease not change

3)     If the supply voltage of a chip increases, the maximum transistor current will

increase decrease not change

4)     If the width of a transistor increases, its gate capacitance will increase decrease not change

5)     If the length of a transistor increases, its gate capacitance will increase decrease not change

6)     If the supply voltage of a chip increases, the gate capacitance of each transistor will

increase decrease not change

(4)

Activity

1)     If the width of a transistor increases, the current will increase decrease not change

2)     If the length of a transistor increases, the current will increase decrease not change

3)     If the supply voltage of a chip increases, the maximum transistor current will

increase decrease not change

4)     If the width of a transistor increases, its gate capacitance will increase decrease not change

5)     If the length of a transistor increases, its gate capacitance will increase decrease not change

6)     If the supply voltage of a chip increases, the gate capacitance

(5)

DC Response

 DC Response: Vout vs. Vin for a gate

 Ex: Inverter

– When Vin = 0 -> Vout = VDD – When Vin = VDD -> Vout = 0 – In between, Vout depends on

transistor size and current

– By KCL, must settle such that Idsn = |Idsp|

– We could solve equations

– But graphical solution gives more insight

Idsn Idsp

Vout VDD

Vin

(6)

Transistor Operation

 Current depends on region of transistor behavior

 For what Vin and Vout are nMOS and pMOS in – Cutoff?

– Linear?

– Saturation?

(7)

nMOS Operation

Cutoff Linear Saturated

Vgsn < Vgsn >

Vdsn <

Vgsn >

Vdsn >

Idsn Idsp

Vout VDD

Vin

(8)

nMOS Operation

Cutoff Linear Saturated

Vgsn < Vtn Vgsn > Vtn

Vdsn < Vgsn – Vtn

Vgsn > Vtn

Vdsn > Vgsn – Vtn

Idsp

Vout VDD

Vin

(9)

nMOS Operation

Cutoff Linear Saturated

Vgsn < Vtn Vgsn > Vtn

Vdsn < Vgsn – Vtn

Vgsn > Vtn

Vdsn > Vgsn – Vtn

Idsn Idsp

Vout VDD

Vin

Vgsn = Vin Vdsn = Vout

(10)

nMOS Operation

Cutoff Linear Saturated

Vgsn < Vtn Vin < Vtn

Vgsn > Vtn Vin > Vtn

Vdsn < Vgsn – Vtn Vout < Vin - Vtn

Vgsn > Vtn Vin > Vtn

Vdsn > Vgsn – Vtn Vout > Vin - Vtn

Idsp

Vout VDD

Vin

Vgsn = Vin V = V

(11)

pMOS Operation

Cutoff Linear Saturated

Vgsp > Vgsp <

Vdsp >

Vgsp <

Vdsp <

Idsn Idsp

Vout VDD

Vin

(12)

pMOS Operation

Cutoff Linear Saturated

Vgsp > Vtp Vgsp < Vtp

Vdsp > Vgsp – Vtp

Vgsp < Vtp

Vdsp < Vgsp – Vtp

Idsp

Vout VDD

Vin

(13)

pMOS Operation

Cutoff Linear Saturated

Vgsp > Vtp Vgsp < Vtp

Vdsp > Vgsp – Vtp

Vgsp < Vtp

Vdsp < Vgsp – Vtp

Idsn Idsp

Vout VDD

Vin

Vgsp = Vin - VDD Vdsp = Vout - VDD

Vtp < 0

(14)

pMOS Operation

Cutoff Linear Saturated

Vgsp > Vtp

Vin > VDD + Vtp

Vgsp < Vtp

Vin < VDD + Vtp Vdsp > Vgsp – Vtp Vout > Vin - Vtp

Vgsp < Vtp

Vin < VDD + Vtp Vdsp < Vgsp – Vtp Vout < Vin - Vtp

Idsp

Vout VDD

Vin

Vgsp = Vin - VDD V = V

Vtp < 0

(15)

I-V Characteristics

 Make pMOS is wider than nMOS such that n = p

Vgsn5

Vgsn4 Vgsn3 Vgsn2 Vgsn1

Vgsp5 Vgsp4 Vgsp3 Vgsp2 Vgsp1

VDD -VDD

Vdsn -Vdsp

-Idsp Idsn

0

(16)

Current vs. V

out

, V

in

Vin5

Vin4

Vin3 Vin2 Vin1 Vin0

Vin1

Vin2 Vin3 Vin4 Idsn, |Idsp|

(17)

Load Line Analysis

Vin5

Vin4

Vin3 Vin2 Vin1 Vin0

Vin1

Vin2 Vin3 Vin4 Idsn, |Idsp|

Vout VDD

 For a given Vin:

– Plot Idsn, Idsp vs. Vout

– Vout must be where |currents| are equal in

Idsn Idsp

Vout VDD

Vin

(18)

Load Line Analysis

Vin0 Vin0

Idsn, |Idsp|

 Vin = 0

(19)

Load Line Analysis

Vin1 Vin1

Idsn, |Idsp|

Vout VDD

 Vin = 0.2VDD

(20)

Load Line Analysis

Vin2 Vin2

Idsn, |Idsp|

 Vin = 0.4VDD

(21)

Load Line Analysis

Vin3 Vin3

Idsn, |Idsp|

Vout VDD

 Vin = 0.6VDD

(22)

Load Line Analysis

Vin4

Vin4 Idsn, |Idsp|

 Vin = 0.8VDD

(23)

Load Line Analysis

Vin5 Vin0

Vin1

Vin2 Vin3 Vin4 Idsn, |Idsp|

Vout VDD

 Vin = VDD

(24)

Load Line Summary

Vin5

Vin4

Vin3 Vin2 Vin1 Vin0

Vin1

Vin2 Vin3 Vin4 Idsn, |Idsp|

(25)

DC Transfer Curve

 Transcribe points onto Vin vs. Vout plot

Vin5

Vin4

Vin3 Vin2 Vin1 Vin0

Vin1

Vin2 Vin3 Vin4

Vout VDD

C Vout

0

Vin

VDD VDD

A B

D E

Vtn VDD/2 VDD+Vtp

(26)

Operating Regions

 Revisit transistor operating regions

C Vout

0

V

VDD VDD

A B

D E

Vtn VDD/2 VDD+Vtp

Region nMOS pMOS A

B C D E

(27)

Operating Regions

 Revisit transistor operating regions

C Vout

0

Vin

VDD VDD

A B

D E

Vtn VDD/2 VDD+Vtp

Region nMOS pMOS

A Cutoff Linear

B Saturation Linear

C Saturation Saturation

D Linear Saturation

E Linear Cutoff

(28)

Beta Ratio

 If p / n  1, switching point will move from VDD/2

 Called skewed gate

 Other gates: collapse into equivalent inverter

Vout

0 VDD

0.5 21

p 10

n

p 0.1

n

(29)

Noise Margins

 How much noise can a gate input see before it does not recognize the input?

Indeterminate Region NML

NMH

Input Characteristics Output Characteristics

VOH VDD

VOL

GND VIH

VIL

Logical High Input Range

Logical Low Input Range Logical High

Output Range

Logical Low Output Range

(30)

 By definition, VIH and VIL are the operational points of the inverter where

Vout

VOH

VDD Unity Gain Points

Slope = -1

p/n > 1 Vin Vout

A high gain (g) is desirable

(31)

Logic Levels

 To maximize noise margins, select logic levels at

VDD Vin Vout

VDD

p/n > 1 Vin Vout

0

(32)

Logic Levels

 To maximize noise margins, select logic levels at – unity gain point of DC transfer characteristic

Vout

VOH

VDD Unity Gain Points

Slope = -1

p/n > 1 Vin Vout

(33)

Transient Response

DC analysis tells us Vout if Vin is constant

Transient analysis tells us Vout(t) if Vin(t) changes – Requires solving differential equations

 Input is usually considered to be a step or ramp – From 0 to VDD or vice versa

(34)

Inverter Step Response

 Ex: find step response of inverter driving load cap

(

0

)

(

) )

(

o i

ut n

out

V t t t

V t V

d d

t

 

Vin(t)

Vout(t) Cload Idsn(t)

(35)

Inverter Step Response

 Ex: find step response of inverter driving load cap

0 0

( )

( )

( )

( )

ou

DD in

t out

u t t V

d d

t

t t V t V

V

t

Vin(t)

Vout(t) Cload Idsn(t)

(36)

Inverter Step Response

 Ex: find step response of inverter driving load cap

0

(

0

( )

) ( (

) )

DD D

o i

D o t

n

ut u

V t

u t t V V

d d

t

t V

V

t t

 

Vin(t)

Vout(t) Cload Idsn(t)

(37)

Inverter Step Response

 Ex: find step response of inverter driving load cap

0

0

( )

( )

( ) (

(

) )

DD DD

loa d ou

i

d t

o n

ut sn

V V

u t t V t t

V t

V d

dt C

t

I t

 

 

0

( ) out DD t

ou ds

t DD t

I n t V V

V

V V

V

t t

Vin(t)

Vout(t) Cload Idsn(t)

(38)

Inverter Step Response

 Ex: find step response of inverter driving load cap

0

0

( )

( )

( ) (

(

) )

DD DD

loa d ou

i

d t

o n

ut sn

V V

u t t V t t

V t

V d

dt C

t

I t

 

 

 

0 2

2

0 )

( DD out DD t

n

Ids V

t t

V V V V

t

Vin(t)

Vout(t) Cload Idsn(t)

(39)

Inverter Step Response

 Ex: find step response of inverter driving load cap

0

0

( )

( )

( ) (

(

) )

DD DD

loa d ou

i

d t

o n

ut sn

V V

u t t V t t

V t

V d

dt C

t

I t

 

 

 

0 2

2

0

)2 )

( ( )

( DD DD t

DD

out

out out out D t

n

t ds

D

I V

t t

V V V V

V V V V V

t

V t V t

   

 

Vout(t) Vin(t)

t0 t

Vin(t)

Vout(t) Cload Idsn(t)

nMOS – “1” transfer is bad

(40)

Delay Definitions

tpdr: rising propagation delay

– From input to rising output crossing VDD/2 (upper bound)

tpdf: falling propagation delay

– From input to falling output crossing VDD/2 (upper bound)

tpd: average propagation delay – tpd = (tpdr + tpdf)/2

tr: rise time

– time required for a node to charge from the 10% point to 90% ( 0.1 VDD to 0.9 VDD )

tf: fall time

(41)

Tpdr = TpHL Tcdf = TpHL

tr tf

Delay Definitions

(42)

CMOS Inverter Switching Characteristics

Define:

– Falling delay tdf = delay time with output falling – Rising delay tdr = delay time with output rising

(43)

Trajectory of a n-transistor operating point

a. Transistor is Cut off

a. Applying a Vgs=VDD b. Vout -> 0,9 to (VDD-Tth) a. VDD-Tth -> to 0.1V

(44)

Fall-time equivalent circuit

Rise-time equivalent circuit

The fall-time is faster than the rise time,

tf=tr/2

primarily due to different carrier mobilities associated with the p and n- devices:

µn = 2 µp

(45)

 If the designer wants to have both n an p transistor with the same rise and fall time:

βn / βp = 1

 This implies that channel width for the p-device must be increased to approximately two to three times of the n-device:

wp = 2 – 3 wn

n+ n+

p-type body

W

L tox

SiO2 gate oxide (good insulator, ox = 3.90) polysilicon

gate

Trajectory of a n-transistor operating

point

(46)

Delay Estimation

 We would like to be able to easily estimate delay – Not as accurate as simulation

– But easier to ask “What if?”

 The step response usually looks like a 1st order RC response with a decaying exponential.

 Use RC delay models to estimate delay – C = total capacitance on output node – Use effective resistance R

– So that tpd = RC

 Characterize transistors by finding their effective R

(47)

RC Delay Models

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

k g

s d

g

s d

kC kC

R/k kC

k g

s d

g

s

d kC

kC kC 2R/k

Capacitance

proportional to width

Resistance inversely

proportional to width

nMOS pMOS

(48)

Resistance of a uniform slab:

– R =  (l/A) = (/t) (L/W) where  is the resistivity in ohm-cm, t is the thickness in cm, L is the length, W is the width, and A is the cross-sectional area

– Using the concept of sheet resistance, R = Rs (L/W) where Rs is called the sheet resistance and given in ohms per square

• Rs =  / t

Gate Capacitance -

Cg = oxWL/tox  = oxA/tox 

n+ n+

p-type body

W

L tox

SiO2 gate oxide (good insulator, ox = 3.90) polysilicon

gate

RC Delay Models

(49)

MOS device capacitances

MOS Device capacitances

– Cgs, Cgd = gate-to-channel capacitances, which are lumped at the source and the drain regions.

– Csb, Cdb = source and drain-duiffusion capacitances to bulk (or substrate) – Cgb = gate –to-bulk capacitance

(50)

1. Off region, where Vgs < Vt. There is no channel Cgs, Cgd = 0.

2. Non-saturated region, where Vgs - Vt. > Vds. Cgb = 0

3. Saturated region, where Vgs <-Vt. < Vds. Channel is heavely inverted.

The drain is pinched off causing Cgd = 0.

MOS device capacitances

(51)

RC Delay Models

pMOS capacitance

= 2x nMOS capacitance

(52)

Example: 3-input NAND

 Sketch a 3-input NAND with transistor widths chosen to achieve effective rise and fall resistances equal to a unit inverter (R).

(53)

Example: 3-input NAND

 Sketch a 3-input NAND with transistor widths chosen to achieve effective rise and fall resistances equal to a unit inverter (R).

(54)

Example: 3-input NAND

 Sketch a 3-input NAND with transistor widths chosen to achieve effective rise and fall resistances equal to a unit inverter (R).

3 3 2 2

2

3

(55)

3-input NAND Caps

 Annotate the 3-input NAND gate with gate and diffusion capacitance.

2 2 2

3 3 3

(56)

3-input NAND Caps

 Annotate the 3-input NAND gate with gate and diffusion capacitance.

2 2 2

3 3

3 3C

3C 3C 2C

2C

2C 2C

2C 2C

3C 3C

2C 2C 2C

Resultant capacitance per gate = 2C+3C = 5C

Parallel capacitances 3x2C+3C = 9C

(57)

3-input NAND Caps

 Annotate the 3-input NAND gate with gate and diffusion capacitance.

9C 3C 3 3C

3 3 2 2

2 5C

5C

5C

(58)

Elmore Delay

 ON transistors look like resistors

 Pullup or pulldown network modeled as RC ladder

 Elmore delay of RC ladder

R1 R2 R3 RN

C C C C

   

nodes 

1 1 1 2 2

...

1 2

...

pd i to source i

i

N N

t R C

R C R R C R R R C

 

       

(59)

Example: 2-input NAND

 Estimate worst-case rising and falling delay of 2- input NAND driving h identical gates.

h copies

2 2 2 2

B A

x

Y

(60)

Example: 2-input NAND

 Estimate worst-case rising and falling delay of 2- input NAND driving h identical gates.

h copies

(61)

Example: 2-input NAND

 Estimate worst-case rising and falling delay of 2- input NAND driving h identical gates.

h copies

Parallel capacitances 2C+2C+2C= 6C

(62)

Example: 2-input NAND

 Estimate rising and falling propagation delays of a 2- input NAND driving h identical gates.

h copies

6C 2 2C

2 2 2

4hC B

A

x

Y

(63)

Example: 2-input NAND

 Estimate rising and falling propagation delays of a 2- input NAND driving h identical gates.

h copies

6C 2 2C

2 2 2

4hC B

A

x

Y

R

(6+4h)CY

t

pdr

(64)

Example: 2-input NAND

 Estimate rising and falling propagation delays of a 2- input NAND driving h identical gates.

h copies

6C 2 2C

2 2 2

4hC B

A

x

Y

R Y

t 6 4 h RC

Rising time: nMOS - “off”

pMOS - “on”

(65)

Example: 2-input NAND

 Estimate rising and falling propagation delays of a 2- input NAND driving h identical gates.

h copies

6C 2 2C

2 2 2

4hC B

A

x

Y

(66)

Example: 2-input NAND

 Estimate rising and falling propagation delays of a 2- input NAND driving h identical gates.

h copies

6C 2 2C

2 2 2

4hC B

A

x

Y

t

x R/2 Y

Consider that the nMOS resistance = (pMOS resistance)/2 = R/2

(67)

Example: 2-input NAND

 Estimate rising and falling propagation delays of a 2- input NAND driving h identical gates.

h copies

6C 2 2C

2 2 2

4hC B

A

x

Y

       

 

2 2 2

2 6 4

7 4

R R R

t

pdf

C h C

h RC

       

 

(6+4h)C R/2 2C

x R/2 Y

(68)

Delay Components

 Delay has two parts – Parasitic delay

• 6 or 7 RC

• Independent of load – Effort delay

• 4h RC

• Proportional to load capacitance

(69)

Contamination Delay

 Best-case (contamination) delay can be substantially less than propagation delay.

 Ex: If both inputs fall simultaneously

6C 2 2C

2 2 2

4hC B

A

x

Y

R

(6+4h)CRY

3 2

t

cdr

  h RC

(70)

3 7C 2 2

2 2C

2C

Isolated Contacted Diffusion Merged

Uncontacted Shared Contacted Diffusion

Diffusion Capacitance

 we assumed contacted diffusion on every source/drain

 Good layout minimizes diffusion area

 Ex: NAND3 layout shares one diffusion contact – Reduces output capacitance by 2C

– Merged uncontacted diffusion might help too

Capacitance = 2C+2C+3C = 7C

(71)

Layout Comparison

 Which layout is better?

VDD A

GND

B

Y

VDD A

GND

B

Y

Referências

Documentos relacionados

75 Figure 4.11: Rotation response to a step input signal applied to roll, pitch

Este trabalho tem como objectivo central o desenvolvimento de uma plataforma de suporte à actualização dinâmica e distribuída do modelo operacional as-is, baseada na

Metodologia: Trata-se de um relato de experiência da construção e implementação de uma paródia sobre neoplasias da mama, criada e apresentada para 66 Agentes Comunitários

grego é, assim, a estranha síntese entre a forma e o informe, de tal modo que se descobre como sendo o próprio mundo da arte, sem que nunca se perca a visão de que “a sua

O BJECTIVES : Using Nail Psoriasis Severity Index for evaluation in patients with plaque psoriasis during treatment with acitretin.. M ETHODS : Thirty volunteers with plaque

A differenza degli anni Trenta, dove lo Stato fascista, mosso dalla necessità di salvaguardare il proprio potere, ebbe l’esigenza di controllare ogni ambito della vita sociale

A contribuição da barragem subterrânea na sustentabilidade do agroecossistemas foi avaliada a partir de metodologias baseadas no Marco para la Evaluación de

Using market equity data to estimate industry size and industry concentration, I find evidence that the average stock returns on high-concentration industries are higher than