1

LNA LinearizationUsing

Bipolar Transistors

By Edgar Sánchez-SinencioThanks to Drs Chunyu Xin and Alberto Garcia for providing

part of the material for this presentation.

Analog & Mixed Signal CenterElectrical and Computer Engineering Department

Texas A&M University

ECEN 665

2

Review: Conventional CMOS LNA Topologies

M1

M2

Rs

VBB

INPUT

ZLOUTPUT

VDD

M1

Rf

INPUT

RL

OUTPUT

VDD

R1

1

1RR

RZ

L

fin

+≈

sin RZ =1

1

min g

Z =

M1INPUT

RL

OUTPUT

VDD

VBB

INPUT

VBB

C1

Lg

ZL

OUTPUT

VDD

Ls

( ) sgs

m

gssgin L

Cg

CjLLjZ +++=

ωω 1

sm RgF

1

142αγ

+≥αγ

+≥1F

4.8 dB < 2dB2~5 dB

Resistive Termination Common Gate Shunt-series Feedback

Source Degeneration

> 6dBNF:

LSmLm RRgsmRg RgF 2

1

1

1 11 +++=+

αγ

3

Input Impedance Matching

CGS1

LG

LS

LM

M1

M2VB

*

*

M

ZIN

( ) ( )MLCg

sCLLsZ S

gs

m

gsSGin ±+++=

11

1

( ) 1

1

gsSGo CLL +=ω ( )MLR STin ±=ω

At resonant Frequency:

( )1121

1thgs

gs

mT VV

LCg

−∝=ω

113 thgs VVIIP −∝

( )211 thgsD VVLWI −∝

( ) ( )MLIVVIIP SDthgs −↑→↑→−↑→ 113

( ) ( )MLIIPVVI STthgsD +↓→↓→↓→−↓→ ω311

Design trade-offs:

The goal: Input impedance match to 50 or 75 Ohm

A Mutual Coupled Degenerated LNA

4

Inter-stage Impedance

CGS1

LG

+

_

LS

LM

11 GSm Vg1GSV

*

*ro1

Zo

Zin2Zo1Zin2'

CGS2

2

1

mg

sI

1oI

1

2

11 22 ooT

os

T

ooo rZjLrZ ≈++=

ωω

ωω

Mg

Z TT

o

T

omin ω

ωω

ωω

2

2

2

2

'2

1

11⎟⎟⎠

⎞⎜⎜⎝

⎛±

⎟⎟⎠

⎞⎜⎜⎝

⎛+

≈

2

'2

1

min g

Z ≤

More current into cascoded stage

Reduced inter-stage gain

Reduced Miller effect

Improved reverse isolation

Verification of reduced Miller effect

5

Noise

CGS1

LG

LS

LM

M1*

*

21ngi

21ndi

CGS1

LS

LM

M1*

*

2ni

2nv LG

Noise model

2

112

⎟⎟⎠

⎞⎜⎜⎝

⎛+=

T

ondngn jiii

ωω

2

1

12

⎟⎟⎠

⎞⎜⎜⎝

⎛−=

gso

ngn Cj

ijvω

M does not appear!

Loss due to inferior inductor

Substrate loss

Noise figure degradation:

6

A GSM LNA Using the Proposed Matching Method

CGS1

LS

LM

M1

M2VB

*

*

MCC1

M3

LD1

LD2

RD

CC2

CC3

RS

VDD

RF IN

RF OUT

LG

C. Xi and E. Sanchez-Sinencio, “ A GSM LNA Using Mutual-Coupled Degeneration.”IEEE Microwave and Wireless Component Letters, Vol. 15, No. 2, pp 68 -70, Feb. 2005

so

Tm R

jG 1⎟⎟⎠

⎞⎜⎜⎝

⎛−=

ωωEffective transconductance

7

Why Linearity So Important ?

Unwanted non-linearity will :Compress amplified signalDesensitize front-endGenerate harmonics (filter out)Generate in-band interference (IMD)Cross-modulation

Communication system always deals with interferences.

Wanted Signal-102dBm

In-band blocker-23dBm

f1 f2

Wanted Signal

IM3f=2f1-f2

In-band blocker

f1 f2

Channel SelectionFiltering Profile

IM3f=2f2-f1

8

Linearity Metrics

P1dB@input

P1dB@output1dB

Pout (dB)

Pin (dB)

Fund

amen

tal

IM3 IM2

IIP2

OIP2

OIP3

IIP3

Pout(dBm)

Pin (dBm)

1dB compression: Measure gain compression for large input signal

IIP3/IIP2: Measure inter-modulation behavior

Relationships between IIP3 and P1dBFor one tone test: IIP3-P1dB=10dBFor two tone test: IIP3-P1dB=15dB

tGSodds

VDS

VVVVwhere

v dso

−==

=( +θμμ 1)

Next we will consider the mobility not constant, i.e. ,

9

Non-linearity Terms of MOS Device

( ) 33

22 gsgsgsmgsds vgvgvgvi ++=

Intrinsic MOS I-V characteristic:

-0.22

-0.16

-0.10

-0.04

0.02

0.08

0.14

0.20

0.26

0.32

0.20 0.28 0.36 0.44 0.52 0.60 0.68 0.76 0.840.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

g 2(A/

V )

2g 3

(A/V

)3

g3

g2

gm

g m(A/

V)

Vth

Vgs(V)

mg 2g3g

( )( )21

2

od

odod

VVKV

θθ

++

( )31 odVKθ+ ( )41 odV

Kθθ+

−Strong/moderate

t

sIηφ

0

( )20

2 t

sIηφ ( )3

0

6 t

sIηφ

Weak

Inversion Level

θχχ+

=1

2

Kids

⎟⎟⎠

⎞⎜⎜⎝

⎛ −+=

t

thgst

VVηφ

ηφχ2

exp1ln2

10

Non-Linearity Analysis of Conventional Inductive Degenerated LNA

gL

sLsR

sv

dL dRdC

gsC

gdC

gsv

( )gsds vfi =

dL dRdC

3Z

2Z

gLsRsv

1Z

( )( ) ( ) ( )ωωεωω

ωω2,6

123 31

12Δ⋅⋅⋅

=−AHR

IIPs

( ) ( )ωωωωε 2,2, 3 Δ−=Δ oBgg3goBg

NegativePositive

The absolute value of these two quantities should be kept small in order to achieve high linearity.

( ) ( ) ( )⎥⎦⎤

⎢⎣

⎡+

+Δ+

=Δωω

ωω2

12322, 2

2 gggggg

mmoB ↓3g

↓oBg↑⇒ 3IIP

( )⋅g is a function of Z1, Z2 and Z3

11

Review of Multi-Gated-Transistor Linearization

VBmain

VBauxRB

RB

Cc

Cc

RFIN M1

M2

Original Configuration: Alternate Configuration:

VBmain

VBauxRB

RB

Cc

Cc

RFINM1

M2 M3

VDD

VB

33

22 gsgsgsmds vgvgvgi ++=

3,3

2,2, gsauxgsauxgsauxmaux vgvgvgi ++=

auxdsout iii +=

33

22 gsgsgsmds vgvgvgi ++=

3,3

2,2, gsauxgsauxgsauxmaux vgvgvgi −+−=

auxdsout iii −=678

910111213

141516

0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7678

910111213

141516

Auxilliary gate bias, Vaux (V)

IIP2IIP3

Vmain=0.74V

IIP3

(dB

m)

IIP2

(dB

m)

1

2

3

4

5

6

7

8

9

10

0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.700

5

10

15

20

25

30

35

IIP3

(dB

m)

IIP2

(dB

m)

IIP3

IIP2

Main gate bias at 0.74V

Auxiliary gate bias, Vaux (V)

12

Proposed Method: Hybrid LNA

reRB1

RFIN

RB2

VBNPN

VBMOS

VDD

Ld RFOUT

Bond wire

Bipolar replaces NMOSat weak inversion.M1

Q2

MOS in weak inversion has speed problem

MOS transistor in weak inversion acts like bipolar

Bipolar available in TSMC 0.18 technology (not a parasitic BJT)

Why not using that bipolar transistor to improve linearity ?

13

Linearity Analysis of the BJT

Bipolar is more non-linear than MOS

Degeneration used to match the 3rd order non-linear term of MOST

πCπv

ev⎟⎟⎠

⎞⎜⎜⎝

⎛ −=

t

eiQo

vvIiφ

exp

er

oiiv

brbi

μC

re

oi

iv

Weak memory effect

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

0.20

0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00Bias voltage (V)

3rd

orde

r coe

ffici

ent

3GHz

DC

( )( )em

em

m

Qbjt rg

rgg

Ig 21

161

5

3

2,3 −+

=

14

3rd Order Cancellation Effect

-0.30

-0.22

-0.14

-0.06

0.02

0.10

0.18

0.26

0.34

-0.16 -0.12 -0.08 -0.04 0.00 0.04 0.08 0.12 0.16

g 3(A

/V )3

g3,NPN

Input voltage (V)

g3,MOS g3

re

M1

Q2

MOS and BJT biased separately

MOS in moderate inversion, BJT in active region

The Goal!

15

Effects on Input Impedance Matching and Noise

LS

Lg

Ct gπ Vgs

gmVgs

Zin

-36

-34

-32

-30

-28

-26

-24

-22

-20

-18

2.80 2.85 2.90 2.95 3.00 3.05 3.10 3.15 3.20

BJT Enabled

BJT Disabled

Frequency (GHz)

S11(

dB)

πCCC gst += 1

πC shifts the matching point to a lower frequency

πg moves the impedance away from the intended value

BJT biased at low current: 320uA

BJT noise contribution: 2.4%

23.6%Other2.4%Bipolar Transistor14%MOS Transistor60%Source Resistance

Noise ratioDevice

16

Biasing Temperature Profile

0.66

0.69

0.72

0.75

0.78

0.81

0.84

-10 0 10 20 30 40 50 60 70 80 90 100

NPN

opt

imal

bia

s (V

)

Temperature ( C)

MOS biased by constant-gm

BJT biased by a PTAT circuit

17

Experimental Results of the Proposed Linearized LNA

-16

-12

-8

-4

0

4

8

2.5 2.55 2.6 2.65 2.7 2.75 2.8 2.85 2.9

S21

S22

S11

S12

Frequency (GHz)

S-pa

ram

eter

s (d

B)

-95

-75

-55

-35

-15

5

25

-30 -25 -20 -15 -10 -5 0 5 10 15 20

IIP3=14.5dBm

Input power (dBm)

Out

put

pow

er (

dBm

)

mW8.9PddB2.1NF

dBm14.5IIP3dB6.4Gain

GHz2.7Frequency

Active area: 390um x 290um

18

Extend to a Differential Version

678

910111213

141516

0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7678

910111213

141516

Auxilliary gate bias, Vaux (V)

IIP2IIP3

Vmain=0.74V

IIP3

(dB

m)

IIP2

(dB

m)

Single-ended suffers from small IIP2

Out-of-band termination

( ) ( ) ( )⎥⎦⎤

⎢⎣

⎡+

+Δ+

=Δωω

ωω2

12322, 2

2 gggggg

mmoB

The 3rd order term of MOS and BJT differential pair has the same sign.

BJT is more non-linear than MOS

Less current for BJT to present same non-linearity as MOS

Cross-couple MOS and BJT differential pair will help

( )( ) ( ) ( )ωωεωω

ωω2,6

123 31

12Δ⋅⋅⋅

=−AHR

IIPs

( ) ( )ωωωωε 2,2, 3 Δ−=Δ oBgg

19

M1 M2

Q1 Q2

re re

VDD

VOUT+

VIN-

LD LD

VIN+

IEE ISS

VOUT-

MB1 MB2 MB3 MB4

-1.60

-1.20

-0.80

-0.40

0.00

0.40

0.80

1.20

-300 -200 -100 0 100 200 300

Input voltage (mV)

IM3

(A/V

^3)

BJT

BJT+NMOS

NMOS

-120

-80

-40

0

40

80

120

-300 -220 -140 -60 20 100 180 260

Input voltage (mV)

IM2

(mA/

V^2)

NMOS

BJT+NMOS

Extended flatnessfor small IM2

BJT

Extend to a Differential Version (Cont’d)

BJT pair contributes 15% of noiseLarger noise figure: 3.4 dBLarger current dissipation: 10mABetter reverse isolation: 25 dBNo need out-of-band termination

20

Experimental Results of the Proposed Differential LNA

-100

-80

-60

-40

-20

0

20

40

-30 -25 -20 -15 -10 -5 0 5 10 15

IIP3 with BJT enabled: 12.3 dBm

IIP3 with BJT disabled: 7.3 dBm

12.5 dB

Input power (dBm)

Out

put

pow

er (

dBm

)

-40

-35

-30

-25

-20

-15

-10

-5

0

5

10

15

2.3 2.35 2.4 2.45 2.5 2.55 2.6 2.65 2.7

S21

S22S11

S12

Frequency (GHz)

S- p

aram

eter

s (d

B)

mW19.8PddB3.4NF

dBm12.5IIP3dB10Gain

GHz2.5Frequency

21

IM3 Cancellation Demo

Measurement video clip shows the IM3 cancellation effect of BJT differential pair in the differential LNA.

SMIQ03

Test board

SMIQ03

FSEB30

Powercombiner

Measurement setup

22

Comparison Table

7.219.812.33.4102.5Differential[proposed]

676.6121.29.53BiCMOS

[Simulated]

4.945182.850.9Differential

[2]

22.88.914.52.16.42.7Single-ended [proposed]

18.521.115.62.85100.9Single-ended [1]

mWdBmdBdBGHz

FOMPdIIP3NFGainFrequency

( ) DPFIIPGFOM1

3−⋅

=

C. Xin, E. Sanchez-Sinencio, "A Linearization Technique for RF Low Noise Amplifier", IEEE International Symposium on Circuits and Systems, May, 2004.ISCAS 2004 , pp. 313 -316, 2004

23

Conclusions

Hybrid: Bipolar linearizes MOST

Differential structure: no degradation on IIP2

Better trade-offs between design parameters

Good figure of merit