6.

776 High Speed Communication Circuits Lecture 11 Noise Figure, Impact of Amplifier Nonlinearities
Michael Perrott Massachusetts Institute of Technology March 10, 2005
Copyright 2005 by Michael H. Perrott

Noise Factor and Noise Figure (From Lec 10)

enRs
Equivalent output referred current noise (assumed to be independent of Zout and ZL)

Rs

vin
Zin

vx

Linear,Time Invariant Circuit (Noiseless)

inout
Zout

iout
ZL

Definitions

Calculation of SNRin and SNRout

H.-S. Lee & M.H. Perrott

MIT OCW

Alternative Noise Factor Expression

enRs
Equivalent output referred current noise (assumed to be independent of Zout and ZL)

Rs

vin
Zin

vx

Linear,Time Invariant Circuit (Noiseless)

inout
Zout

iout
ZL

From previous slide

Calculation of Noise Factor

H.-S. Lee & M.H. Perrott

MIT OCW

Thevenin Computation Model For Noise (from Lec 10)

iout

D
Iout

G D S
Zg Zgs vgs Cgs gmvgs indg

G
Zg

Zdeg

S
Zdeg

H.-S. Lee & M.H. Perrott

MIT OCW

Thevenin Computation Model For Noise (from Lec 10)

iout

D
Iout

G D S
Zg Zgs vgs Cgs gmvgs indg

G
Zg

Zdeg

S
Zdeg

We analyzed a non-degenerated CMOS amplifier in Lec 10

- Broadband amplifiers: gate noise not significant - Narrowband amplifiers:

Assumed in analysis that we operated at resonance Q turned out to be key design parameter Can we simultaneously design amplifiers for optimal noise match and optimal power match?
H.-S. Lee & M.H. Perrott MIT OCW

Input Referred Noise Model

Equivalent output referred current noise (assumed to be independent of Zout and ZL)

Rs

enRs

vin Zin

vx

Linear,Time Invariant Circuit (Noiseless)

inout Zout ZL

iout

en

iin

is

Ys

in Zin

vx

Linear,Time Invariant Circuit (Noiseless)

iout

en

Can remove the signal source since Noise Factor can be expressed as the ratio of total output noise to input noise

is

Ys

in

iin,sc=

iout

H.-S. Lee & M.H. Perrott

MIT OCW

Input-Referred Noise Figure Expression

en

is

Ys

in

iin,sc=

iout

We know that

Lets express the above in terms of input short circuit current

H.-S. Lee & M.H. Perrott

MIT OCW

Calculation of Noise Factor

en

is

Ys

in

iout iin,sc=

By inspection of above figure

In general, en and in will be correlated

-Y

is called the correlation admittance

H.-S. Lee & M.H. Perrott

MIT OCW

Noise Factor Expressed in Terms of Admittances

en

is

Ys

Yc

iu

iout iin,sc=

We can replace voltage and current noise currents with impedances and admittances

H.-S. Lee & M.H. Perrott

MIT OCW

Optimal Source Admittance for Minimum Noise Factor

Express admittances as the sum of conductance, G, and susceptance, B

Take the derivative with respect to source admittance and set to zero (to find minimum F), which yields

Plug these values into expression above to obtain

H.-S. Lee & M.H. Perrott

MIT OCW

Optimal Source Admittance for Minimum Noise Factor

After much algebra (see Appendix L of Gonzalez* book for derivation), we can derive

- Contours of constant noise factor are circles centered about (G ,B ) in the admittance plane - They are also circles on a Smith Chart (see pp 299-302
opt opt

of Gonzalez for derivation and examples)

How does (Gopt,Bopt) compare to admittance achieving maximum power transfer?

*Guillermo Gonzalez, Microwave Transistor Amplifiers: Analysis and Design, Prentice Hall, 1996
H.-S. Lee & M.H. Perrott MIT OCW

Optimizing For Noise Figure versus Power Transfer

Signal Source Source conductance susceptance
en

iin

is

Gs

Bs

in Zin

vx

Linear,Time Invariant Circuit (Noiseless)

iout

Source noise produced by source conductance

Bs

Example source admittance for maximum Bopt power transfer Bmax Gmax Gopt

Circles of constant Noise Factor (Fmin at the center) Gs

One cannot generally achieve minimum noise figure if maximum power transfer is desired
H.-S. Lee & M.H. Perrott MIT OCW

Optimal Noise Factor for MOS Transistor Amp

Consider the common source MOS amp (no degeneration) considered in Lecture 10

- In Tom Lees book (pp. 272-276 (1

ed.), pp. 364-369 (2nd ed.)), the noise impedances are derived as

st

- The optimal source admittance values to minimize noise

factor are therefore

H.-S. Lee & M.H. Perrott

MIT OCW

Optimal Noise Factor for MOS Transistor Amp (Cont.)

Optimal admittance consists of a resistor and inductor (wrong frequency behavior broadband match fundamentally difficult)

- If there is zero correlation, inductor value should be set

to resonate with Cgs at frequency of operation

Minimum noise figure

- Exact if one defines w = g /C

t m

gs

H.-S. Lee & M.H. Perrott

MIT OCW

Recall Noise Factor Comparison Plot From Lecture 10

Noise Factor Scaling Coefficient Versus Q for 0.18 NMOS Device
8 7

Noise Factor Scaling Coefficient

6 5 4 3 2 1 0

c = -j0 c = -j0.55 c = -j1

Achievable values as a function of Q under the constraint that 1 = wo LgCgs c = -j0 c = -j0.55 Minimum across all values of Q and 1 LgCgs
4 5 6 7

Note: curves meet if we approximate Q2+1 Q2

c = -j1

10

H.-S. Lee & M.H. Perrott

MIT OCW

Example: Noise Factor Calculation for Resistor Load

Source Rs Source Rs enRs enRL

vin

RL

vout
RL

vnout

Total output noise

Total output noise due to source

Noise Factor

H.-S. Lee & M.H. Perrott

MIT OCW

Comparison of Noise Figure and Power Match

Source Rs Source Rs enRs enRL

vin

RL

vout
RL

vnout

To achieve minimum Noise Factor

To achieve maximum power transfer

H.-S. Lee & M.H. Perrott

MIT OCW

Example: Noise Factor Calculation for Capacitor Load

Source Rs Source Rs enRs

vin

CL vout

CL

vnout

Total output noise

Total output noise due to source

Noise Factor
H.-S. Lee & M.H. Perrott MIT OCW

Example: Noise Factor with Zero Source Resistance

Source RL RL enRL

vin

CL vout

CL

vnout

Total output noise

Total output noise due to source

Noise Factor
H.-S. Lee & M.H. Perrott MIT OCW

Example: Noise Factor Calculation for RC Load

Source Rs Source Rs enRs

enRL

vin

CL

RL vout

CL RL

vnout

Total output noise

Total output noise due to source

Noise Factor

H.-S. Lee & M.H. Perrott

MIT OCW

Example: Resistive Load with Source Transformer

Source RS

Vs Rin= 1 N
2

Vx RL Rout=N Rs
2

RL

Vout=NVx

1:N
RS enRs

Source

Rin=

1 N
2

Vx RL Rout=N2Rs

enRL Vnout=NVx RL

1:N

For maximum power transfer (as derived in Lecture 3)

H.-S. Lee & M.H. Perrott

MIT OCW

Noise Factor with Transformer Set for Max Power Transfer

Source RS enRs

Vx Rin= Rs Rout=RL RL Rs

enRL Vnout= RL

RL V Rs x

1:N=

Total output noise

Total output noise due to source

Noise Factor

H.-S. Lee & M.H. Perrott

MIT OCW

Observations
Source RS enRs

Vx Rin= Rs Rout=RL RL Rs

enRL Vnout= RL

RL V Rs x

1:N=

If you need to power match to a resistive load, you must pay a 3 dB penalty in Noise Figure

- A transformer does not alleviate this issue

What value does a transformer provide?

- Almost-true answer: maximizes voltage gain given the power match constraint, thereby reducing effect of noise of following amplifiers Accurate answer: we need to wait until we talk about cascaded noise factor calculations
MIT OCW

H.-S. Lee & M.H. Perrott

Nonlinearities in Amplifiers
We can generally break up an amplifier into the cascade of a memoryless nonlinearity and an input and/or output transfer function
Vdd
RL Memoryless Nonlinearity

Vout Vin
Id

Lowpass Filter

Id

-RL 1+sRLCL

Vout

Vin

M1

CL

Impact of nonlinearities with sine wave input

- Causes harmonic distortion (i.e., creation of harmonics) - Causes harmonic distortion for each input AND
intermodulation products

Impact of nonlinearities with several sine wave inputs

H.-S. Lee & M.H. Perrott

MIT OCW

Analysis of Amplifier Nonlinearities

Focus on memoryless nonlinearity block

- The impact of filtering can be added later

Memoryless Nonlinearity

Model nonlinearity as a Taylor series expansion up to its third order term (assumes small signal variation)

- For harmonic distortion, consider - For intermodulation, consider

H.-S. Lee & M.H. Perrott MIT OCW

Harmonic Distortion

Substitute x(t) into polynomial expression

Fundamental

Harmonics

- M.H. Perrott for small A, very large for large A Very small H.-S. Lee &

Notice that each harmonic term, cos(nwt), has an amplitude that grows in proportion to An
MIT OCW

Frequency Domain View of Harmonic Distortion

3c A Afund = c1A + 3 4
3

A 0 w

Memoryless Nonlinearity

2w 3w

Harmonics cause noise

- Their impact depends highly on application

LNA typically not of consequence Power amp can degrade spectral mask Audio amp depends on your listening preference!

Gain for fundamental component depends on input amplitude!

H.-S. Lee & M.H. Perrott MIT OCW

1 dB Compression Point
Memoryless Nonlinearity w 3c A Afund = c1A + 3 4
3

A 0

2w 3w

Definition: input signal level such that the small-signal gain drops by 1 dB

20log(Afund)
1 dB

- Input signal level is high!

A1-dB

20log(A)

Typically calculated from simulation or measurement rather than analytically

series to be accurate in this context
H.-S. Lee & M.H. Perrott

- Analytical model must include many more terms in Taylor

MIT OCW

Harmonic Products with An Input of Two Sine Waves

DC and fundamental components

Second and third harmonic terms

Similar result as having an input with one sine wave

- But, we havent yet considered cross terms!

H.-S. Lee & M.H. Perrott

MIT OCW

Intermodulation Products

Second-order intermodulation (IM2) products

Third-order intermodulation (IM3) products

- These are the troublesome ones for narrowband

systems
H.-S. Lee & M.H. Perrott MIT OCW

Corruption of Narrowband Signals by Interferers

Memoryless Nonlinearity

X(w)

Interferers

Desired Narrowband Signal

W

w1 w 2

Y(w)

Corruption of desired signal

w1 w2 2w1 2w2 3w1 3w2 0 w2-w1 2w1-w2 2w2-w1 w1+w2 2w1+w2 2w2+w1

Wireless receivers must select a desired signal that is accompanied by interferers that are often much larger

- LNA nonlinearity causes the creation of harmonic and intermodulation products - Must remove interference and its products to retrieve
desired signal

H.-S. Lee & M.H. Perrott

MIT OCW

Use Filtering to Remove Undesired Interference

Memoryless Nonlinearity

X(w)

Interferers

Desired Narrowband Signal

W

Bandpass Filter

w1 w 2

Y(w)

Corruption of desired signal

w1 w2 2w1 2w2 3w1 3w2 0 w2-w1 2w1-w2 2w2-w1 w1+w2 2w1+w2 2w2+w1

Z(w)

Corruption of desired signal

2w1 2w2 3w1 3w2 w1 w2 0 w2-w1 2w1-w2 2w2-w1 w1+w2 2w1+w2 2w2+w1

Ineffective for IM3 term that falls in the desired signal frequency band
H.-S. Lee & M.H. Perrott MIT OCW

Characterization of Intermodulation
Magnitude of third order products is set by c3 and input signal amplitude (for small A)

Magnitude of first order term is set by c1 and A (for small A)

Relative impact of intermodulation products can be calculated once we know A and the ratio of c3 to c1

- Problem: its often hard to extract the polynomial

coefficients through direct DC measurements Need an indirect way to measure the ratio of c3 to c1
H.-S. Lee & M.H. Perrott MIT OCW

Two Tone Test

Input the sum of two equal amplitude sine waves into the amplifier (assume Zin of amplifier = Rs of source)
vin(w)
Equal Amplitude Sine Waves

2A
w 1 w2

Note: v (w) in vx(w) = 2

W

Amplifier Rs

Vx

Vout

vin Vbias
Zin=Rs

first-order output

Vout(w)

third-order IM term

Vout=co+c1Vx+c2Vx+c3Vx
w1 w2 2w1 2w2 3w1 3w2 0 w2-w1 2w1-w2 2w2-w1 w1+w2 2w1+w2 2w2+w1
W

- First order term will increase linearly -& Third order IM term will increase as the cube of A H.-S. Lee M.H. Perrott

On a spectrum analyzer, measure first order and third order terms as A is varied (A must remain small)

MIT OCW

Input-Referred Third Order Intercept Point (IIP3)

Plot the results of the two-tone test over a range of A (where A remains small) on a log scale (i.e., dB)

- Extrapolate the results to find the intersection of the

first and third order terms
20log(Afund)

1 dB First-order output = c1A Third-order 3 3 = 4 c3 A IM term

- IIP3 defined as the input power at which the

extrapolated lines intersect (higher value is better) Note that IIP3 is a small signal parameter based on extrapolation, in contrast to the 1-dB compression point
H.-S. Lee & M.H. Perrott MIT OCW

A1-dB

Aiip3

20log(A)

Relationship between IIP3, c1 and c3

Intersection point
20log(Afund)

Solve for A (gives Aiip3)

1 dB First-order output = c1A Third-order 3 3 = 4 c3A IM term

A1-dB

Aiip3

20log(A)

Note that A corresponds to the peak value of the two cosine waves coming into the amplifier input node (Vx)

- Would like to instead like to express IIP3 in terms of power

MIT OCW

H.-S. Lee & M.H. Perrott

IIP3 Expressed in Terms of Power at Source

IIP3 referenced to Vx (peak voltage)
vin(w)
Equal Amplitude Sine Waves

2A
w 1 w2
W

Note: v (w) in vx(w) = 2

Rs

Amplifier

Vx

Vout

vin

IIP3 referenced to Vx (rms voltage)

Vbias

Zin=Rs

Vout=co+c1Vx+c2Vx+c3Vx

Power across Zin = Rs

Note: Power from vin

H.-S. Lee & M.H. Perrott

MIT OCW

IIP3 as a Benchmark Specification

Since IIP3 is a convenient parameter to describe the level of third order nonlinearity in an amplifier, it is often quoted as a benchmark spec Measurement of IIP3 on a discrete amplifier would be done using the two-tone method described earlier

- This is rarely done on integrated amplifiers due to poor access to the key nodes - Instead, for a radio receiver for instance, one would simply

put in interferers and see how the receiver does Note: performance in the presence of interferers is not just a function of the amplifier nonlinearity

Calculation of IIP3 is most easily done using a simulator such as Hspice or Spectre

- Two-tone method is theoretically not necessary simply

curve fit to a third order polynomial

H.-S. Lee & M.H. Perrott

MIT OCW

Impact of Differential Amplifiers on Nonlinearity

I1 vid 2 M1 vx 2Ibias I2 M2 -vid 2 Memoryless Nonlinearity

vid

Idiff = I2-I1

Assume vx is approximately incremental ground

Second order term removed and IIP3 increased!

H.-S. Lee & M.H. Perrott MIT OCW