Dispositivi non lineari: una breve introduzione dei ... and Defense Symposium 2010Agilent Seminar Seminario Agilent Dispositivi non lineari: una breve introduzione dei Parametri X

Aerospace and Defense Symposium 2010Agilent Seminar

Seminario Agilent

Dispositivi non lineari:

una breve introduzione dei Parametri X

e del Nonlinear Vector Network Analyzer


Evolution of the Tools & Measurements

TOOLS:

NA

SA/SS/NFA

Power meter

Oscilloscope

DC Parametric Analyzer

MEASUREMENTS:

Gain compression, IP3, IMD

PAE, ACPR, AM-PM, BER

Constellation Diagram, EVM

GD, NF, Spectral Regrowth

ACLR, Hot “S22”

S-Parameters

TOOLS:

Vector Network

Analyzer

MEASUREMENTS:

Gain

Input match

Output match

Isolation

Transconductance

Input capacitance

S-Parameters +

Figures of Merit

TOOLS:

SS & Oscilloscope

Grease pens and

Polaroid cameras

Slotted line

Power meter

MEASUREMENTS:

Bode plots

Gain

SWR

Scalar network analyzers

Y & Z parameters

PatchworkNVNA +

X-Parameters


Agenda

• Steps and Architectures in Network Analysis

• Nonlinear VNA and X-Parameter

3


Classic VNA Block Diagram

4


Network Analysis Step1:

S Parameter (1965)Linear or nonlinear networks operating with

signals sufficiently small to cause the

networks to respond in a linear manner, can

be completely characterized by parameters

measured at the network terminals without

regard to the contents of the networks.

- Relatively easy to obtain at high frequencies (measure voltage

traveling waves with a vector network analyzer don't need shorts/opens

which can cause active devices to oscillate or self-destruct);

- Relate to familiar measurements (gain, loss, reflection coefficient ...);

- Can cascade S-parameters of multiple devices to predict system

performance;

- Can compute H, Y, or Z parameters from S-parameters if desired;

- Can easily import and use S-parameter files in our electronic-

simulation tools.

• K. Kurokawa, ‘Power Waves and the Scattering Matrix,IEEE Transactions on Microwave Theory and Techniques, Vol. MTT-13, No. 2, March, 1965.

5


Modern VNA Block DiagramRF jumpers

Receivers

Mechanical switch

C

R3

Test port 1

R1

Test port 4

R4

A D

rear panel

Pulse generators

1

2

3

4

Source 1

OUT 1 OUT 2

Pulse

modulator

Source 2

OUT 1 OUT 2

Pulse

modulator

Test port 2

R2

B

Noise receivers

10 MHz -

3 GHz

3 -

26.5

GHz

To receivers

LO

+28V (for noise source)

Test port 3

Signal

combiner+

-

Impedance tuner for noise

figure measurements

6


Network Analysis Step2:

X Parameter (2007)Agilent Technologies has

developed X-parameters.

X-parameters represent and

analyze the nonlinear behavior of

RF components in a much more

robust and complete manner. As

an extension of S-parameters

under large-signal operating

conditions, they are driven into

saturation (the real-world

operating environment) and then

measured under these

conditions.

7


Agenda

• Steps and Architecture in Network Analysis

• Nonlinear VNA and X-Parameter

8


Designing a Power Amplifier (1)

The designer wants to drive the amplifier

into the nonlinear region (large signal)

to get the maximum output power as well as

to extract the maximum efficiency

9


Designing a Power Amplifier (2)

Under large-signal conditions there are nonlinear effects:

- distort waveforms (time domain),

- harmonics, inter-modulations… (frequency domain).

S-parameters can only analyze and model the linear behavior.

10


Current Techniques for a Non-linear Device

Linear parameters

S=parameters

Data sheets for amplifiers often have both linear and non-linear information

Non-linear parameters

IMD

Pout vs Pin

P1dB

PAE

11


A [new]-parameter file like a 3-D CAD file

[new] Parameter file3-D file

The “library” contains all the necessary information

[new]-parameter S-parameter

12


X-parameter Concept

13


X-parameters come from the

Poly-Harmonic Distortion (PHD) Framework

Port Index Harmonic (or carrier) Index

14


Harmonic Superposition Principle

• In general, we are working under large-signal, nonlinear operating conditions.The superposition principle (parameters of individual components are sufficient to determine the parameters of any combination of those components) is not valid.

• In many practical cases (power amplifiers stimulated with a narrowband input signal) there is only one dominant large-signal input componentpresent (A11).All other input components (the harmonic frequency components) are relatively small. In that case, we will be able to use the superposition principle for the relatively small input components.

• This is called the harmonic superposition principle.

15


Harmonic Superposition Principle• (to keep the graph simple) Only consider the presence of the A1m and

B2n components (neglect the presence of the A2m and B1n).

• Case A11 <> 0. A1m and B2n are indicated by black arrows. Note harmonic components for the B2n components.

• Leave the A11 the same and add a small A12 component (second harmonic at the input). This will result in a deviation of the output spectrum B2The same holds for a third harmonicand a fourth harmonic.

• The harmonic superposition principle holds when the overall deviation of the output spectrum B2 is the superposition of all individual deviations

16


The [new] X-parameter

Classic S-parameters: only linear behavior and ignore nonlinear

behavior (harmonic, intermodulation, higher order mixing effects).

X-parameters capture linear behavior and linearize

nonlinear behavior about a large signal operating point (LSOP).

NVNA measures X-parameters stimulating the DUT

with a single large tone at port 1 and, in the same time,

injecting additional small tones at both ports 1 and 2

at all harmonics of interest. At least two phase-offset small tones must be injected at each port/frequency of interest in

order to extract the corresponding X-parameters.

17


Test port 3

C

R3

Test port 1

R1

Source

1

OUT 1OUT 2

Source 2 (standard)

OUT 1 OUT 2

Test port 4

R4

Test port 2

R2

35 dB

65 dB

35 dB

65 dB 65 dB 35

dB

65 dB

Rear panel

A

35 dB

D B

SW1 SW3 SW4 SW2

J11 J10 J9 J8 J7 J4 J3 J2 J1

Calibration phase reference

Measurement phase reference

Injecting a large tone at port 1

To port 1 or 3 for calibration

10 MHz

Internal

Reference

18


Test port 3

C

R3

Test port 1

R1

Source 1

OUT 1OUT 2

Source 2 (standard)

OUT 1 OUT 2

Test port 4

R4

Test port 2

R2

35 dB

65 dB

35 dB

65 dB 65 dB 35

dB

65 dB

Rear panel

A

35 dB

D B

SW1 SW3 SW4 SW2

J11 J10 J9 J8 J7 J4 J3 J2 J1

To port 1 or 3

Calibration

Phase

Reference

Ext Source in

Measurement

Phase

Reference

Injecting a drive and probe tone at port 1

19


Test port 3

C

R3

Test port 1

R1

Source 1

OUT 1OUT 2

Source 2 (standard)

OUT 1 OUT 2

Test port 4

R4

Test port 2

R2

35 dB

65 dB

35 dB

65 dB 65 dB 35

dB

65 dB

Rear panel

A

35 dB

D B

SW1 SW3 SW4 SW2

J11 J10 J9 J8 J7 J4 J3 J2 J1

To port 1 or 3

Calibration

Phase

Reference

Ext Source in

Measurement

Phase

Reference

Injecting a drive tone at port 1 and probe tone at port 2

20


DC f0 2f0 3f0 4f0 5f0-f0-2f0-3f0

21

aj2aj2*

Measuring B2,3 which is the

Port2 output value at 3rd harmonic (3f0).X(F)

2,3 … + X(S)23,gh … + X(T)

23,gh …


On-frequency Upper sideband Lower sideband

22


Understanding X-parameter display

X-parameters: XFp, XSpq and XTpq

Both the p and q terms are described by a Port and Harmonic combination.

For example: XS21,13

Parameters for each possible (port, harmonic) selection of p For example: for p=(2,1) , and assuming 3 harmonics of interest:

XF21

(XS21,11),(XT21,11)

(XS21,12), (XT21,12)

(XS21,13), (XT21,13)

(XS21,21), (XT21,21)

(XS21,22), (XT21,22)

(XS21,23), (XT21,23)

For a 2-port device, with 3 harmonics of interest at each port:

- there are 6 possible (port, harm) combinations for p.

- each combination yields 13 parameters,

Total of 6*13

78 X-param

23


X-parameters reduce to S-parameters

11

( )

21 11 21| | 0

[ ( )] F

A

X A s

11

( )

21,21 11| | 0

( ) 0T

AX A

-60-25 -20 -15 -10 -5 0 5 10

|A11| (dBm)

-40

40

20

0

-20

dB

( )

21,21

SX

( )

21,21

TX

( )

21 11/ | |FX A

Reduces to (linear)

S-parameters in the

appropriate limit

11

( )

21,21 11 22| | 0

( )S

AX A s

11| |A

24


25

[X]p,f,Vg,Vd,z3

[X]p,f,Vg,V,z2

[X]p,f,Vg,Vd3

[X]p,f,Vg,Vd2

[X]p,f,Vg3[X]p,f3

[X]p,f2

[X]p3

[X]p2

X-parameters is actually a large data library

-- X-param is measured with many variables --

Various

Power

X-param one bookshelf

one library

Measurement takes from tens of

minutes to a several hours depending

on the size of the “library”

[X]p1 [X]p,f1

[X]p,f,Vg2

[X]p,f,Vg,Vd1

[X]p,f,Vg,Vd,z

one page one chapter

[X]p,f,Vg1

one book

one series

of books

[X]p,f,Vg,Vd,z1

Various

Frequencies

Various

gate voltages

Various

drain voltages

one X-param file

Various

load Z

=

26


Example of an X-Par .xnp file

27


Calibration

Three step

calibration process:

1) SOLT or TRL style calibration

2) Absolute receiver calibration using a power meter/sensor

3) Receiver phase calibration using Agilent’s new comb generator.

28


Amplitude Calibration

Vector Calibration

Phase Calibration

Phase Reference

X-Parameter extraction/multi-tone source

Standard PNA-X Network Analyzer

NVNA System Configuration

29


Measurement Display

30


1. Nonlinear Component Characterization

* Absolute amplitude and cross frequency relative phase of measured spectra traceable to standards lab

* Data displayed in frequency, time and power domains:- ‘a’ and ‘b’ waves versus sweep domain- V and I versus sweep domain

- V versus I, I versus V

Frequency Domain

31






Time Domain

32




* Data displayed in frequency, time andpower domains:- ‘a’ and ‘b’ waves versus sweep domain- V and I versus sweep domain


Power Domain

33






Customize Domain: V/I

34


2. X-parameter measurementFrequency Domain

35


2. X-parameter measurementPower Domain

36


Three-port X-parameter

Two-tone X-parameter

Pulse envelope domain

37


Three-port X-parameter measurements

38


Two-tone X-parameter measurements

39


Nonlinear pulse envelope domain

Configuration, Cal and Meas (Waves and Harmonics): menu Utility > Measurement Domain… > Envelope Domain

40


NVNA and ADS

41


Don’t you have this problem? device model model takes time and is not accurate

My simulation gave

me a great answer

but the reality is…

I want to start

my design now!

-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0-2.5 2.5

-70

-60

-50

-40

-30

-20

-10

-80

0

freq, MHz

Spectr

um

_out

Transmitted Spectrum

measured

ACPR-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0-2.5 2.5

-100

-80

-60

-40

-20

-120

0

freq, MHz

Sp

ectr

um

_o

ut

Transmitted Spectrum

simulated

ACPR

Modeling: I need a

couple of months

42


X-parameter enables 3 activities to run in parallel

process dev. amplifier desing module desing

before

after

amp designers can start design

without waiting for the model

modeling

process dev.

amplifier design

module design

[X]

[X]

[X]Sim Meas

module designers can start design

with the simulated X-param then with

the measured X-param

Meas

New technologies like GaN

43


NVNA and ADS

44


S ParameterNA

S ParameterADS

45


Component Char.NVNA

Component Char. ADS

____ -20dBm Input

____ 0dBm input

46


X ParameterNVNA

X ParameterADS

47


ADSS-parameter vs X-parameter

S-parameter

X-parameter

S-par not valid

____ -20dBm Input

____ 0dBm input

48


ADSS-parameter vs X-parameter

____ -20dBm Input

____ 0dBm input

S-par not valid

S-par

X-par

Transmission Reflection Output Voltage

49


Power Amplifier Design

50


X-Parameter Example on GaN amplifier

• 32 Watt 1.2 GHz GaN (Gallium Nitride) Amplifier,

- PAE (Pout-Pin)/(Vdc*Idc) on datasheet: 55%

1. Measured with PNA-X/NVNA this amplifier and obtained X-parameters.

2.Loaded the measured X-parameters into ADS

3.Performed simulation in ADS:

- ADS predicted an PAE of 75%,

by using optimum output impedance matching circuitry on 1s, 2nd and 3rd harmonic.

4.Realized PCB for this amplifier,

following the ADS suggestion for optimization

5.Measured the PCB (with amplifier mounted on it) with PNA-X/NVNA

- Actual measured PAE was 75%

- Matches ADS prediction

- By using this X-par approach we could achieve 20% better PAE.

51


NVNA X-parameter System – Power budget (120W)

Application Note 1408-19: High Power Amplifier Measurements using Agilent’s Nonlinear Vector Network Analyzer

52


High Power X-parameter Measurement System

53


Agilent is using X-parameters

for the internal design and

also selling ICs to the external

customers with X-parameter

files.

USACanadaUKGermanyFranceNetherlnandSwedenTurkeyPolandRomaniaJapanTaiwanChinaKorea

Component manufacturers start to make

X-Parameters available to their customers

54


Summary

The NVNA nonlinear measurements

with full match correction and accurate amplitude

and cross frequency relative phase information

provide a new standard in accuracy

and insight of the behaviors of nonlinear

components

55

Dispositivi non lineari: una breve introduzione dei ... and Defense Symposium 2010Agilent Seminar Seminario Agilent Dispositivi non lineari: una breve introduzione dei Parametri X

Documents