TRL algorithm to de-embed a RF test fixture

Similar documents
Lecture 13. Vector Network Analyzers and Signal Flow Graphs

Unknown Thru Calibration Algorithm

Advancements in mm-wave On-Wafer Measurements: A Commercial Multi-Line TRL Calibration Author: Leonard Hayden Presenter: Gavin Fisher

Application Note. Electrical. Optical

S-Parameter Calibration of Two-Port Setup: Gavin Fisher Cascade Microtech

SPARQ S-Parameter Measurements with Impulse Response Time Limiting

Using the unknown-thru method for waveguide/coax systems with the 8510

Achieving First Pass Success in PCB-Based Filter And Matching Circuit Designs

Calibration Uncertainty Estimation. at the Wafer Level

Elimination of Conductor Foil Roughness Effects in Characterization of Dielectric Properties of Printed Circuit Boards 14 TH1

Broad-band space conservative on wafer network analyzer calibrations with more complex SOLT definitions

Solving Absolute Value Equations Worksheet

De-embedding in High Speed Design

R. W. Erickson. Department of Electrical, Computer, and Energy Engineering University of Colorado, Boulder

An On-Wafer Deembedding Procedure for Devices under Measurement with Error-Networks Containing Arbitrary Line Lengths

Class VIII Chapter 1 Rational Numbers Maths. Exercise 1.1

Advances in Radio Science

WE PRESENT an iterative algorithm, which we first

Update: Reference Correlation for the Viscosity of Ethane

Lab #8 Two Port Networks

S-PARAMETER QUALITY METRICS AND ANALYSIS TO MEASUREMENT CORRELATION

Impact of adapters on LISN's input impedance calibration: How to improve accuracy and reliability of the measurements? Part 2

In-Situ De-embedding (ISD)

Measuring Complex Permittivity, Permeability, or Sheet Impedance of Materials at Microwave Frequencies with a Free-Space Spot Probe System

Accuracy Estimations of S-Parameter Measurements at the Wafer-Level

A CALIBRATION PROCEDURE FOR TWO-PORT VNA WITH THREE MEASUREMENT CHANNELS BASED ON T-MATRIX

2160 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 56, NO. 6, DECEMBER /$ IEEE

Cross Regulation Mechanisms in Multiple-Output Forward and Flyback Converters

An Approximate Approach to Determining the Permittivity and Permeability near λ/2 Resonances in Transmission/Reflection Measurements

In-Situ De-embedding (ISD)

Error Correction in Vector Network Analyzers

Signal integrity simulation strategies for accurate and fast results Correct Material Properties that simulate quickly.

Contents. Transmission Lines The Smith Chart Vector Network Analyser (VNA) ü structure ü calibration ü operation. Measurements

Open-Thru de-embedding for Graphene RF devices

Properties of Real Numbers. Unit 1 Lesson 4

Passive components in MMIC technology

MITOCW watch?v=pqkyqu11eta

Boundary and Excitation Training February 2003

R. W. Erickson. Department of Electrical, Computer, and Energy Engineering University of Colorado, Boulder

Keysight Technologies Measurement Uncertainty of VNA Based TDR/TDT Measurement. Application Note

Calibration Procedure for Complex Permittivity Extraction Using an Open-Ended Coaxial Probe. Arjen Kremers Mai-Han Truong

Characterising The RF Properties of Metamaterials

A simplified method of VNA uncertainty estimation

Two Port Networks. Definition of 2-Port Network A two-port network is an electrical network with two separate ports for input and output

Ripple Method for Evaluating Residual Errors

MFJ 259/259B Operation & Simplified Calibration

Section Gaussian Elimination

Measurements of Electromagnetic Properties of Ferrites above Curie Temperature and impact on kicker performance

Note on Chebyshev Regression

This is our rough sketch of the molecular geometry. Not too important right now.

Mathematics Curriculum Document for Middle School Algebra 1

Chapter 3. Steady-State Equivalent Circuit Modeling, Losses, and Efficiency

Experiment 06 - Extraction of Transmission Line Parameters

George T. Flowers Dean Graduate School

Waves on Lines. Contents. ! Transmission Lines! The Smith Chart! Vector Network Analyser (VNA) ! Measurements

Estimation of Complex and Linear Uncertainities in S-Parameter Measurements for Metrology Applications

Networks and Systems Prof. V. G. K. Murti Department of Electrical Engineering Indian Institution of Technology, Madras

TDR, S-Parameters & Differential Measurements. TDR, S-Parameters & Differential Measurements

R. W. Erickson. Department of Electrical, Computer, and Energy Engineering University of Colorado, Boulder

ECE 497 JS Lecture -07 Planar Transmission Lines

Shielding effectiveness measurements and uncertainty estimation for textiles by a VNA-based free space transmission method

24 = 5x. 4.8 = x. December 13, 2017

MP112D GHz GaAs MMIC Attenuator. Datasheet

Contents. Transmission Lines The Smith Chart Vector Network Analyser (VNA) ü structure ü calibration ü operation. Measurements

Physics for Scientists and Engineers 4th Edition 2017

Solving Equations by Multiplying or Dividing

A New Coaxial Flow Calorimeter for Accurate RF Power Measurements up to 100 Watts and 1 GHz Andrew S. Brush Jefferson D. Lexa

Accounting for High Frequency Transmission Line Loss Effects in HFSS. Andrew Byers Tektronix

Data Reconciliation Techniques

A Union of Euclidean Spaces is Euclidean. Konstantin Makarychev, Northwestern Yury Makarychev, TTIC

Calibrating On-Wafer Probes to the Probe Tips

Complex fraction: - a fraction which has rational expressions in the numerator and/or denominator

4/2/ Hunting4Gold.com - DoradoVista, Inc. Page 1 of 20

The Pennsylvania State University. The Graduate School. Science, Engineering, and Technology

A Comparison of Measurement Uncertainty in Vector Network Analyzers and Time Domain Reflectometers. White Paper

Analytic Solutions for Periodically Loaded Transmission Line Modeling

Alg 1B Chapter 7 Final Exam Review

The trick is to multiply the numerator and denominator of the big fraction by the least common denominator of every little fraction.

Probe Tip Characterization Using Self-Calibration. Lab. RF Modeling

Insertion Loss. Yantel Corporation

Grade 8 Rational Numbers

Networks and Systems Prof. V. G. K. Murti Department of Electrical Engineering Indian Institute of Technology, Madras

Microwave Impedance Measurement for Nanoelectronics

Using Full Wave Solvers for Practical Analysis of Capacitor Mounting Structures

Technique for the electric and magnetic parameter measurement of powdered materials

Reaction rates and rate law expressions Page 1

Chemical Thermodynamics I

MULTILAYER CHIP BEADS EMI SUPPRESSORS

Mechanics, Heat, Oscillations and Waves Prof. V. Balakrishnan Department of Physics Indian Institute of Technology, Madras

Fractions. Review R.7. Dr. Doug Ensley. January 7, Dr. Doug Ensley Review R.7

Notes for course EE1.1 Circuit Analysis TOPIC 10 2-PORT CIRCUITS

264 CHAPTER 4. FRACTIONS cm in cm cm ft pounds

2 EXPRESSIONS AND SEQUENCES

Quality metrics for S-parameter models

Formative Assessment Opportunity: Performing operations with integers is an essential skill in algebraic mathematics. This lesson is an excellent

15. LECTURE 15. I can calculate the dot product of two vectors and interpret its meaning. I can find the projection of one vector onto another one.

6.1. Rational Expressions and Functions; Multiplying and Dividing. Copyright 2016, 2012, 2008 Pearson Education, Inc. 1

Chemical Thermodynamics I

Announcements. Unit 6: Simple Linear Regression Lecture : Introduction to SLR. Poverty vs. HS graduate rate. Modeling numerical variables

Math-3. Lesson 3-1 Finding Zeroes of NOT nice 3rd Degree Polynomials

Transcription:

TRL algorithm to de-embed a RF test fixture T. Reveyrand University of Colorado - Boulder ECEE department 425 UCB Boulder, Colorado 80309 USA July 2013

1 TRL Standards THRU and LINE Measurements T OUT parameters : T12 and T IN parameters : The THRU equality : T 11 T 12 and T11 and T21 T 22 T 22 T21 T 22 The REFLECT equality : Extracting 2 Reciprocity 3 Scilab Code Presentation Example Insights

Motivations for this talk De-embedding of a test-fixture measured in the coaxial reference planes ; Perform a TRL calibration when the VNA provides some ill conditioned solutions "Reflect" checking not correct ; Offer a open, complete and ready-to-use source code for educational purpose.

Some definitions : [S parameters a 1 a 2 A b 1 b 2 b1 b 2 [ S11 S = 12 S 21 S 22 a1 a 2 1

Some definitions : [T parameters a 1 a 2 A b 1 b 2 a1 b 1 [ T11 T = 12 T 22 b2 a 2 2

Some definitions : [T parameters a 1 a 2 A B C b 1 b 2 [T Total = [T A [T B [T C 3

Conversions between [S and [T [S to [T [T = [ 1 S 21 S 11 S 22 S 21 S 21 S 12 S 11 S 22 S 21 S 21 4 [T to [S [S = [ T 22 T 12 1 T 12 5

Standards for the TRL algorithm THRU : totally known [ TTHRU Std = [ 1 0 0 1 6 REFLECT : unknown Sgn { } R Γ Std REFLECT = ±1 7 LINE : partially known [ TLINE Std = [ e γ l 0 0 e +γ l 8

Measuring the THRU [ [ TTHRU Meas = [T IN TTHRU Std [T OUT 9 [ [ [T IN 1 TTHRU Meas = TTHRU Std [T OUT 10 [ [T IN 1 TTHRU Meas = [T OUT 11 [T IN 1 = [ T IN [ TIN [T THRU Meas = [T OUT 12

Measuring the LINE [ T Meas LINE [ [T IN 1 [ = [T IN TLINE Std [T OUT 13 T Meas LINE = [ [ TIN [T LINE Meas e γ l 0 = 0 e +γ l [ TLINE Std [T OUT 14 [T OUT 15

OUTPUT : Defining the [M matrix Equation 15 is : [ T 12 T 22 12 in 16 give : [ T11 T 12 or T 22 [ T11 T 12 [ T Meas LINE [ 1 [ TTHRU Meas T 22 [ T11 e = γ l T 12 e γ l e +γ l T 22 e +γ l T Meas LINE 16 [ T11 e = γ l T 12 e γ l e +γ l T 22 e +γ l [ T11 e [M = γ l T 12 e γ l e +γ l T 22 e +γ l 17 18 with [M = [ 1 [ TTHRU Meas T Meas LINE

OUTPUT : TRL Equations Equations given by 18 are : M 11 + T 12 M 21 = e γ l 19 M 12 + T 12 M 22 = T 12 e γ l 20 M 11 + T 22 M 21 = e +γ l 21 M 12 + T 22 M 22 = T 22 e +γ l 22

OUTPUT : Solving T 12 20 gives : 23 in 19 gives : e γ l = T11 T 12 M 11 + T 12 M 21 = T12 M 11 + T12 2 M 21 + M 21 = T12 M 12 + M 22 23 [ T11 T 12 T11 T 12 M 12 + M 22 24 M 12 + M 22 25 M 11 M 22 M 12 = 0 26

OUTPUT : Solving T 22 21 gives : 27 in 22 gives : e +γ l = M 11 + M 12 + T 22 M 22 = T 22 T22 M 22 + T21 2 M 21 + M 12 = T 22 T22 T22 [ T22 T22 M 21 27 M 21 + M 11 28 M 21 + M 11 29 M 11 M 22 M 12 = 0 30

OUTPUT : T 12 and T 22 X 2 M 21 + X [M 11 M22 M 12 31 This polynom has 2 solutions : T12 and T22 T 21 Usually, T 12 < T 22 If we consider the following polynom : X 2 M 12 + X [M 22 M11 M 21 32 Then the 2 solutions are T11 T 12 and T21 T 22

INPUT : Defining the [N matrix Equation 15 is : [ T 12 T 22 [ 12 in 33 gives : [ T 12 or T 22 [ [ T 12 T 22 T Meas LINE T Meas LINE = = [ e γ l 0 0 e +γ l [ e γ l 0 0 e +γ l [ T11 T 12 T 22 [ T 11 T 12 T 22 [ T [N = 11 e γ l T 12 e γ l e +γ l T 22 e +γ l 33 [ TTHRU Meas 34 35 with [ [N = T Meas LINE [ 1 TTHRU Meas

INPUT : T 12 and T 22 Equation 35 is similar to 18. Thus we can consider : X 2 N 21 + X [N 11 N 22 N 12 36 This polynom has 2 solutions : T 12 and T 22 Usually, T 12 T < T 22 11 If we consider the following polynom : X 2 N 12 + X [N 22 N 11 N 21 37 Then the 2 solutions are and T 12 T 22

The THRU equality a 1 a 2 a 3 T IN T OUT b 1 b 2 b 3 b2 a 2 [ T = 11 T 12 T 22 a1 b 1 b2 a 2 [ T11 T = 12 T 22 b3 a 3

Forward mode : b 2 equality to extract The b 2 equality leads us to : a 1 + T 12 b 1 = b 3 + T 12 a 3 38 And by definition, about the THRU measurement, we know : S21 Meas = b 3 a3 a 1 and S Meas 11 = b 1 a3 =0 a 1 =0 Thus, a 1 + T 12 b1 1 + T11 T 12 = a 1 S11 Meas 1 + T 12 S Meas 21 = b 3 39 = S Meas 21 40 S11 Meas 41

Reverse mode : a 2 equality to extract The a 1 equality leads us to : a 1 + T 22 b 1 = b 3 + T 22 a 3 42 For the THRU measurement, we know : S12 Meas = b 1 a1 a 3 and S Meas 22 = b 3 a1 =0 a 3 =0 Thus, leads us to : b 1 T 22 = b 3 + a 3 43 T21 T 22 = S12 Meas S Meas 22 + T22 44

REFLECT Measurement a 1 a 2 Γ Std REF LECT Γ Std REF LECT b 1 b 2 Γ Std REFLECT = b 1 = + T 22 S11 Meas a 1 + T 12 S Meas Γ Std REFLECT = b 2 = T 12 + S22 Meas a 2 T 22 + S Meas 11 22 45 46

REFLECT Equality + T 22 S Meas 11 + T 12 S Meas 11 = T 12 + S22 Meas T 22 + S22 Meas 47 1 + T 22 1 + T 12 2 T21 T 22 S Meas 11 S Meas 11 T 22 = 2 = S Meas T11 22 + T12 S Meas 22 + T22 48 S22 Meas T12 + T 11 S22 Meas T22 + T 21 1+ T 22 S Meas 21 T 1+ 12 S Meas 11 49

REFLECT Equality = ± S Meas T11 T21 T 22 T 22 T12 22 + T 11 S22 Meas T22 + T 21 1+ T 22 S Meas 21 T 1+ 12 S Meas 11 50 There are 2 solutions. We select the good one thanks to the knowledge of Sgn R { Γ Std REFLECT } = ±1 in 45 : T Γ Std REFLECT = 21 1 + T 22 1 + T 12 S Meas 11 S Meas 11 51

TRL Completed The TRL algorithm is completed. We got 7 parameters from : The [M matrix : The [N matrix : The THRU equality : The REFLECT equality : T12 and T22 T 12 and T 22 T 21 T11 from 31 ; from 36 ; from 41 and from 50 ; T21 from 44 ; T 22 It is suffisent for [S parameters de-embedding but not for power measurement. We need to normalize correctly the system of equation finding the absolute value of. For that purpose we will consider a reciprocity assumption : S21 IN = S 12 IN.

Reciprocity assumption We should have : [ T IN = [ S21 S 12 S 11 S 22 S 12 S 22 S 12 S 11 S 12 1 S 12 Thus the reciprocity assumption S 21 = S 12 leds to : 52 T 22 T 12 = 1 53

Reciprocity assumption We can obtain from TRL the complete [ T IN and [TOUT matrix from an arbitrary value of. Those matrix has to be multiplied by K in order to fullfill equation 53 such as : K 2 1 = T 22 T 12 54 K = ± 1 T 22 T 12 55 There are 2 solutions. The good one is selected such as the extrapolated phase of S 21 on DC is as close as possible of zero.

This code is now available in Scilab http://www.microwave.fr/uw.html

Example with an OPEN reflect S2P Measurements : Thru black, Line blue and Open red.

Example with an OPEN reflect S2P Extracted : Port 1 black, Port 2 blue and De-embedded open red.

Example with an SHORT reflect S2P Measurements : Thru black, Line blue and Short red.

Example with an SHORT reflect S2P Extracted : Port 1 black, Port 2 blue and De-embedded short red.

Source Code : uw_trl_calc.sci

Source Code : uw_trl_calc.sci