• No results found

LARGE-SIGNAL NETWORK ANALYZER MEASUREMENTS AND THEIR USE IN DEVICE MODELLING

N/A
N/A
Protected

Academic year: 2021

Share "LARGE-SIGNAL NETWORK ANALYZER MEASUREMENTS AND THEIR USE IN DEVICE MODELLING"

Copied!
24
0
0

Loading.... (view fulltext now)

Full text

(1)

Ewout Vandamme (Agilent Technologies, NMDG),

Wladek Grabinski (Motorola, Geneva),

Dominique Schreurs (K.U.Leuven), and

Thomas Gneiting (ADMOS)

LARGE-SIGNAL NETWORK ANALYZER MEASUREMENTS AND THEIR USE IN

(2)

Mixdes 2002 2

Outline

Large-Signal Network Analyzer (LSNA) technology

Advantages of using LSNA for device modelling engineers

LSNA measurements

• de-embedding

• implementation in CAE tool (iccap) • measurement and simulation results

• tuning of model parameter to LSNA measurements

(3)

Mixdes 2002 3

Agilent’s Large-Signal Network Analyzer technology

Cal Kit,

e.g. LOS, LRRM, etc.

Power Std Phase Std Calibration Standards: a1 b1 ab22 i1 v1 or, equivalently, i2 v2 (a) • RF bandwidth: 600 MHz - 20 GHz • max RF power: 10 Watt

• IF bandwidth: 8 MHz

(4)

Mixdes 2002 4

2-port Device-Under-Test (DUT) under periodic excitation

• e.g. transistor excited by a 2.4 GHz tone with an arbitrary output

termination

All current and voltage waveforms are represented by a fundamental and harmonics

Spectral components Xh = complex Fourier Series coefficients of the waveforms

CW class of signals measured with LSNA

Freq. (f0=2.4 GHz) 1*f0 2*f0 3*f0 4*f0

(5)

Mixdes 2002 5

LSNA measurements: time domain, frequency

domain or combination of both (e.g. envelope in modulation)

      =

= H h t f h j h e X t x 0 2 Re ) ( π

− − = 1 0 2 ) ( 2 f t f h j h f x t e dt X π frequency l fundamenta period f = /1 =

(6)

Mixdes 2002 6

Advantages of using the LSNA in device modelling

Measure the following characteristics of your DUT making a single

connection, using one measurement setup (the LSNA)

• DC,

• Small-signal (Scattering parameters), and • Large-signal behaviour

Verify the model accuracy of your device under realistic operation conditions

• power amplification • high-speed switching

Identify modelling problems at a single glance

• LSNA measurements, e.g., immediately reveal weaknesses in capacitance and

(7)

Mixdes 2002 7

Use of LSNA measurements in CAE tool (iccap)

⇒ ⇒⇒

⇒ model verification, optimisation (and extraction)

ICCAP specific input

ADS netlist. Used, a.o., to impose the measured impedance to the output of the transistor in simulation

(8)

Mixdes 2002 8

Use of LSNA measurements for simulation (1/2)

Measurements RF de-embedding V1m,dc I1m,dc V2m,dc I2m,dc a1c b1c a2c b2c v1c i1c v2c i2c calibrated V1,dc I1,dc V2,dc I2,dc

LSNA accounts for cable resistances v1c i1c v2c i2c v1d i1d v2d i2d

@ f0, 2*f0, …

Reference planes before and afterde-embedding

(9)

Mixdes 2002 9 De-embedding intermezzo (1/2) 0 0.5 1 1.5 2 -3 -2 -1 0 1 2 before after de-embedding Time/period Ga te cu rren t / m A

Equivalent circuit of the RF test-structure,

including the DUT and layout parasitics

(10)

Mixdes 2002 10

De-embedding intermezzo (2/2)

Detailed view on the layout of the RF MOSFET for minimum influence of pad parasitics

(11)

Mixdes 2002 11

Use of LSNA measurements for simulation (2/2)

RF de-embedding Simulations v1c i1c v2c i2c v1d i1d v2d i2d ⇒ v1d i1d v2d i2d Compare measurements: with simulations: vv1s i1s 2s i2s

Rde1and Rde2are de-embedding resistances (in dc path) The load impedance ZLat f=n*f0equals 50 Ω if a2n<-50 dBm

Reference planes before and afterde-embedding

(12)

Mixdes 2002 12

Input capacitance behaviour

Vgs,dc=0.9 V

Vds,dc=0.3 V Vds,dc=1.8 V

(13)

Mixdes 2002 13

Dynamic loadline & transfer characteristic

Vgs,dc=0.3 V

(14)

Mixdes 2002 14

Dynamic loadline & transfer characteristic

Vds,dc=0.9 V Vgs,dc=0.9 V

DC operating point if RF

(15)

Mixdes 2002 15

Dynamic loadline & transfer characteristic

Vgs,dc=1.8 V

Cable resistance + Rde2loss

(16)

Mixdes 2002 16

Intermezzo (1/2): extrapolation example SiGe HBT

Model parameters extracted using DC measurements up to 1 V

100 200 300 400 500 600 700 800 0 900 -0.002 -0.001 0.000 0.001 -0.003 0.002 time, psec i1s ts i1m ts _d e 100 200 300 400 500 600 700 800 0 900 0.6 0.7 0.8 0.9 1.0 1.1 0.5 1.2 time, psec v1s ts v1m ts _d e 100 200 300 400 500 600 700 800 0 900 1.3 1.4 1.5 1.6 1.2 1.7 time, psec v2s ts v2m ts _d e 100 200 300 400 500 600 700 800 0 900 0.000 0.002 0.004 0.006 -0.002 0.008 time, psec i2s ts i2m ts _d e SiGe HBT Vbe= 0.9 V; Vce=1.5 V; Pin= - 6 dBm; f0= 2.4 GHz simul. meas.

(17)

Mixdes 2002 17

Intermezzo (2/2): extrapolation example SiGe HBT

Measured and simulated DC characteristics

Measurement Simulation SiGe HBT - DC characteristics 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 1.6 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020 -0.015 0.025 VbDC DCm eas 1..I ce 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 1.6 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020 -0.015 0.025 VbDC i2.i

Alcatel Microelectronics and the Alcatel SEL

Stuttgart Research Center teams are acknowledged for providing these data.

(18)

Mixdes 2002 18

AM to AM (gain) and AM to PM versus input power

Vds,dc=Vgs,dc=1.2 V 1 dB

compressi on

(19)

Mixdes 2002 19

Drain current & gate voltage time domain waveforms

Vgs,dc=0.3 V

“Class C” Class AB Class A

Vgs,dc=1.2 V Vgs,dc=0.9 V Vds,dc=0.9 V vin vin vin

(20)

Mixdes 2002 20

Effect of operating regime on dissipated power in the DUT, load, and DC power supply — class AB

Vds,dc=0.9 V, Vgs,dc=0.9 V RFoutput power at f=f0 Instantaneous power dissipated in DUT Power delivered by DC supply PAE=37 %

(21)

Mixdes 2002 21

Effect of operating regime on dissipated power in the DUT, load, and DC power supply — “class A”

Vds,dc=0.9 V, Vgs,dc=1.8 V RF output power at f=f0 Instantaneous power dissipated in DUT Power delivered by DC supply PAE=11 %

(22)

Mixdes 2002 22

Tuning of model parameters to LSNA measurements

before

(23)

Mixdes 2002 23

LSNA measurements in device modelling Conclusions:

Unique tool for complete large-signal model accuracy assessment under realistic RF or microwave signals

• information on amplitude and phase

Reduce number of design cycles and reduce manufacturing costs through better device models, thus more optimal designs

Optimize model parameters to LSNA measurements

Benchmark various device models, e.g.,

• BSIM, MM11, EKV, …

• Gummel-Poon, VBIC, MEXTRAM, HICUM, ...

(24)

Mixdes 2002 24

Contact

For info on LSNA technology, visit

http://www.agilent.com/find/lsna

Soon, a measurement and consulting service related to Large-Signal Network Analyzer Technology will be available through the ‘NMDG’ group in Belgium. For info, you need to contact NMDG directly at email: Marcus_Vandenbossche@agilent.com, or

References

Related documents

• Velocity saturation occurs at minimum channel length and can be modeled by including a source degeneration resistor with the simple large

Use the large amplitude waveform generator (available at the top of the function generator menu), to produce a larger signal.. The large amplitude waveform generator is output

The copies of the Examination time table of all the classes are enclosed here in for your kind perusal There fore please kindly approve the dates.

 Most Internet traffic will continue to originate or terminate in a data center, however it will shift to become cloud traffic (expected to be around 69% of data center traffic

In this study, we use the RF reflectometry technique to readout charge transition in a physically defined PMOS silicon QD device. Large gate capacitance.. prevents probing signal

These frequencies include desired IF frequencies, but conversion losses of harmonics of RF signal will be so large that result in the decrease of the level of IF signal to

The standard Model 845-XX includes amplitude modulation (AM), DC-coupled, low distortion wideband frequency modulation (FM), PM, FSK and PSK, frequency chirp, and fast pulse

How to verify the interoperability functionalities of IPv6 capable network management equipments using SNMPv2C to check the implementation of IP-MIB(RFC 4293)