wwwwildrivertechcom Alfred P Neves Alwildrivertechcom phone 503 679 2429 A VNA Manifesto A Primer for Practical Mastery Day 4 Application Topics of SParameters Day 4 Deembedding with Tmatrix approach ID: 529684
Download Presentation The PPT/PDF document "Wild River Technology LLC" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Wild River Technology LLCwww.wildrivertech.comAlfred P. NevesAl@wildrivertech.comphone 503 679 2429
A VNA Manifesto:
A Primer for Practical MasteryDay 4: Application Topics of S-Parameters Slide2
Day 4De-embedding with T-matrix approachTRL calibrationPassivity and Causality – practical testsS-parameter work flowRational Compact ModelingTo fix or not to fix a bungled S-parameterSlide3
De-Embedding ApproachesT-matrixCreate S-parameter from scaled T-linePort ExtensionSlide4
© H. Heck 2008
T-matrix, Chaining ABCD matrixSlide5
VNA Approach to De-embeddingMeasure a de-embedding structure and get s-parameterModel something and get S-parameterUse VNA firmware to de-embed with file from active measurement using T-matrix de-embeddingMatlab is another option for T-matrix approachSlide6
TRL Calibration, On-BoardLines 1,2,3
THRU
OpenSlide7
TRL Calibration is finicky Launch must be decent (low S11, no resonance) – good launch designLaunch Connector Repeatability from SMA to SMA – TDR must be very goodLine lengths accurate – layout, etchImpedance variation across board low – etch, fiber- weave, etc., can wreak havocWe measure the LINE standards group delay, then use those group delays in Cal Kit, also verify impedanceSlide8
Establish a Concerted Calibration Verification Ideal TRL THRU S11=S22=0, S12=S21=1Slide9
Verification of Group Delay of THRU includes noise and moving average, perfect group delay=0psecSlide10
TRL- tipWithout delving into all the details, I don’t use TRL calibration for general signal integrity work.There are better calibrations to get reference plane near DUT, especially for multi-port 4,6,.. 12 port structuresSlide11
Advanced Cal VerificationVerification includesTHRU or Insertion Response using insertable adapterreturn loss using wideband terminators
symmetry, S11=S22Reciprocity, S21=S12
Group DelaySlide12
Again, use validation structure features to your advantageSlide13
Calibration Verification 2-Port using KF-KF AdapterSlide14
Start with Simple Cal Verification:
Simple adapter THRU for non-insertable and flush THRU with insertable Cal. KF-KF adapter has approximately 0.1dB insertion loss and 50psec delay.Slide15
Verify Low-Frequency Calibration. Passivity issues!Slide16
Another example using on-board THRU
Example of Simple
THRU
for
TRL:
Simple check of obvious Passivity Violations.
The check of |S21|>1 is NOT sufficient however!Slide17
Analysis of Calibration Causality Using Polar Plot, Causal S-parameters should only rotate clockwise
Quick Polar view of Insertion locates non-causal behaviorSlide18
Cal verification using precision airlinesFirst, examine transmission aberrations, and return loss
I may check this S21 with no calibration
enabled Slide19
Secondly, use Resonant structure like Beatty Standard.Why?Slide20
Simple Matlab
routine to analyze for Beatty - Symmetry issues Slide21
For 4 Ports:
After Cal Verification, Assess S-parameter of DUT
Overall return loss Symmetry
Group Delay Distortion
Reciprocity
Insertion Loss variation
Quick Time Domain TransformSlide22
Rational Compact Modeling of S-parametersSlide23
RCMBe definition it fixes passivity and causalityIs very accurate, low residual MSE errorAddress DC operating pointNo interpolation issues, continuous functionsWorks in Spice engines directlyRequired intermediate step for Time Domain SimulationSlide24
1st Import S-parameters into Simbeor and evaluate quality, then RCM structureSlide25
Look Closely to S-parameter measured versus RCM simulation, CLOSE FIT!
Stripline
Resonator, CMP-28Slide26
S-parameter Work FlowValidate calibration based on DUT Make measurement of external NIST standard, such as stepped impedance or 50ohm airlineTry to predict what DUT measurement will look like based on length, resonances, pathologies, etc.,Make measurement on on board standard, such as Channel Modeling PlatformAnalyze measurementImport measurement into SI Tool, obtain quality metrics – quality, passivity, causalityRCM modelSlide27
What to do with problem S-parameterDon’t fix S-parameters, the tools on the market don’t work very well.If a tool reports problems, check calibrationSometimes you have to live with and manage issues when using T-matrix, or partial calibration, or fancy de-embedding. See DesignCon2012 tutorial: High-confidence S-parameter Measurement Methodologies for 15-28 Gbps, it is available WRT website.Slide28
Thank YouQuestions, discussion?Alfred P. Nevesal@wildrivertech.com503 679 2429