Signal Integrity for PCB Design

Signal Integrity for PCB Design Sep 17, 2014 Lin, Ming Chih

Agenda Customizable in Footer Page 2 What does PCB matters for Signal Integrity? Impedance, Loss and Delay of Transmission Line TDR, TDT and S parameters Impulse Response and ISI Signal Integrity Transmission Line Models SI Simulation

PCB is a key factor for HSD transmission 5Gbps 10Gbps Customizable in Footer Page 3

What characteristics that might degrade SI? PCB Materials Dk and Df Conductivity PCB Layout Stack-up Layout Drills PCB Manufacturing Roughness Etching Customizable in Footer Page 4

Agenda Customizable in Footer Page 5 What do PCB matters for Signal integrity? Impedance, Loss and Delay of Transmission Line TDR, TDT and S parameters Impulse Response and ISI Signal Integrity Transmission Line Models SI Simulation

Measurement Customizable in Footer Page 6

TDR, TDT and S parameters Customizable in Footer Page 7

S Parameters Customizable in Footer Page 8

Dielectric Loss and Conductor Loss Customizable in Footer Page 9

Symbol Response in Time and Frequency Domain Customizable in Footer Page 10

Inter Symbol Interference (ISI) Customizable in Footer Page 11

Non-ideal transmission lines Customizable in Footer Page 12

Open Stub of PTH VIA Customizable in Footer Page 13

Open stub leds to high frequency resonance Customizable in Footer Page 14

Delata-L proposed by Intel Intel s Delta-L Methodology for Electrical Characterization, Rev. 330223-001 Customizable in Footer Page 15

Agenda Customizable in Footer Page 16 What do PCB matters for Signal integrity? Impedance, Loss and Delay of Transmission Line TDR, TDT and S parameters Impulse Response and ISI Signal Integrity Transmission Line Models SI Simulation

SI Design Flow Modeling Simulation Correlation Analyses Customizable in Footer Page 17

Three Types of Channel Modeling Equation Modeling For Pre-simulation Model is equation based Easy to setup Limited structure Optimizable Less accuracy For simple topology EM Modeling Generate S Parameter for Post-Simulation Extracted by EM solver Accurate structure dimension and material parameter are needed EM solver is structure dependent VNA Modeling Generate S Parameter for Post-Simulation Calibration and deembed is necessary Most accurate Full channel can be modeled one time Customizable in Footer Page 18

Transmission Line Model based on Formula Customizable in Footer Page 19

Ideal Transmission Line Models H E Customizable in Footer Page 20

Transmission Line Model based on EM Solvers Customizable in Footer Page 21

Channel Simulation EM Solver S Parameter Simplified PCB Routing Customizable in Footer Page 22

S Parameters Simulation Insertion Loss Far End Crosstalk Near End Crosstalk Customizable in Footer Page 23

Mixed Mode S Parameters Simulation Customizable in Footer Page 24

Design Vias TDR Simualtion Customizable in Footer Page 25

Discontinuity in Current Path Customizable in Footer Page 26

Simulation & Measurement Correlation Simulation Measurement Customizable in Footer Page 27

Summary SI is inevitable in all modern electronics system design. PCB design is an important factor for Signal Integrity Simulation is useful for SI designers to check their PCB design in advance. And it can help designers to debug a system. Correlation is important to improve simulation and measurement. Customizable in Footer Page 28

Challenges to Obtain Measurement Accuracy and Correlation for PCB Impedance Measurements September 17, 2014 Hidekazu Manabe ( 眞鍋秀一 ) Application Expert Marketing Component Test Division-Kobe Keysight Technologies

Agenda PCB Market Overview Challenges to Obtain Measurement Accuracy and Correlation E5063A ENA Series PCB Analyzer Introduction Summary PCB Analyzer Introduction Page 2

Traditional PCB Measurements test coupon L1/50 trace measurement example Test Coupon Example: 15cm (typ) Test Region (typ 30 to 70% of trace length) Parameters = Impedance (SE & DIFF) Required Tolerance = ±10% PCB Analyzer Introduction Page 3

PCB Market Trends and Forecast PCB market expected to grow at 5% per year from 2012 to 2017. Growth drivers are mobile phones (8% annual rate) and IC packages (6% annual rate). PCB Analyzer Introduction Page 4

PCB Market Trends and Forecast by Type Significant growth is expected for flexible PCBs (FPC) (7% annual rate) and IC package substrates (6% annual rate). PCB Analyzer Introduction Page 5

FPC Market Forecast FPC market is expected to grow at an average rate of 7% per year. The growth drivers are multilayer FPC, which account for more than 40% in volume (8% annual rate) and double-sided FPC, accounting for nearly 20% in volume (11% annual rate). Production volume is expected to increase for mobile phone, tablet, and mobile electronics applications. Other applications areas expected to have minimal growth. PCB Analyzer Introduction Page 6

FPC Types and Applications Single-sided FPC FPC Types Double-sided FPC FPC Applications Camera module LCD module Touch sensor panel Antenna Multi-layer FPC Rigid-Flex Measurement Requirements: FPC trace is measured, rather than coupon. Since trace length is short, higher response resolution is required. Due to increase in data rates, tighter impedance control requirements are increasing. (± 10% => ± 5 ~ 8%). S-parameter required for FPC antenna. VSWR (S11) measured in production. In addition, impedance and isolation (S21) are typically measured in QA. PCB Analyzer Introduction Page 7

Agenda PCB Market Overview Challenges to Obtain Measurement Accuracy and Correlation E5063A ENA Series PCB Analyzer Introduction Summary PCB Analyzer Introduction Page 8

Problems with TDR Oscilloscopes 1. Inadequate measurement accuracy 2. Measurement results can differ between channels or instruments PCB Analyzer Introduction Page 9

Measurement Accuracy Verification Impedance Tolerance +/- 10 % +/- 5% PCB Analyzer Introduction Page 10

Measurement Setup TDR Oscilloscope NIST Traceable Standards 50Ω Airline Reference for error compensation Measurement Cables 25Ω Airline Blue Purple Device for measurement result comparison PCB Analyzer Introduction Page 11

Measurement Accuracy with TDR Oscilloscopes 1. Offset Error Measurement 2. Device Measurement (25 Ω Airline) PCB Analyzer Introduction Page 12

Measurement Setup Channel 1 50 Ω airline PCB Analyzer Introduction Page 13

Offset Error Measurement at Channel 1 60 Ω 50 Ω (airline) 51.8 Ω Offset error is +1.8 Ω 40Ω PCB Analyzer Introduction Page 14

Offset Error Measurement at Channel 2 60 Ω 50 Ω (airline) 50.2 Ω Offset error is +0.2 Ω 40Ω PCB Analyzer Introduction Page 15

Measurement Accuracy with TDR Oscilloscopes 1. Offset Error Measurement 2. Device Measurement (25 Ω Airline) PCB Analyzer Introduction Page 16

Measurement Accuracy with TDR Oscilloscopes Considerations for Offset Compensation PCB Analyzer Introduction Page 17

Agenda PCB Market Overview Challenges to Obtain Measurement Accuracy and Correlation E5063A ENA Series PCB Analyzer Introduction Summary PCB Analyzer Introduction Page 18

What is E5063A ENA Series PCB Analyzer? E5063A ENA Series PCB Analyzer The Best Solution for PCB Manufacturing Test More Accuracy and R&R* More Languages Supported More ESD Robustness also the lowest cost solution in the industry. * Repeatability & Reproducibility www.keysight.com/find/ena-pcb PCB Analyzer Introduction Page 19

What is E5063A ENA Series PCB Analyzer? www.keysight.com/find/ena-pcb PCB Analyzer Introduction Page 20

Dedicated GUI for PCB Manufacturing Test Similar look-and-feel to traditional solutions PCB Analyzer Introduction Page 21

Modes of Operation Edit Test Mode Integrated test file editor allows test engineers to create files with just a few mouse clicks Test engineer determines measurement requirements. Test File (Instrument configuration, test list, limits, etc ) Execute Test Mode Simple and intuitive GUI for non-technical operators Operator recalls test setup file and executes test. PASS / FAIL PCB Analyzer Introduction Page 22

Edit Test Mode Intuitive setup flow allows for simple and intuitive operation Setup and Error Correction Wizards Intuitive and error free setup, error correction, and measurements. Dedicated controls for common adjustments The DUT Length Wizard automatically measures the length of the DUT. Test limits can be defined as a percentage of the DUT length. PCB Analyzer Introduction Page 23

Execute Test Mode Operator Instructions Customize operator prompts as appropriate for your application. Saving Test Results All test results are displayed on screen in waveform and statistical format for real-time analysis. Results can also be saved to file for later inspection. PCB Analyzer Introduction Page 24

More Accuracy and R&R (Repeatability & Reproducibility) (Source: IPC-TM-650 Number 1.9 Measurement Precision Estimation for Variables Data) PCB Analyzer Introduction Page 25

Accuracy Verification (Measurement Setup) PCB Analyzer Introduction Page 26

Accuracy Verification (Performing Full Calibration) Calibration Wizard Electronic Calibration module (ECal) PCB Analyzer Introduction Page 27

Accuracy Verification (Performing Measurements) Channel 1 25 Ω airline PCB Analyzer Introduction Page 28

25 Ω Airline after Full Calibration (Channel 1) 25.3 Ω PCB Analyzer Introduction Page 29

25 Ω Airline after Full Calibration (Channel 2) 25.3 Ω PCB Analyzer Introduction Page 30

Measurement Accuracy after Full Calibration Standards of Different Impedance in Series Connection 25 Ω 50 Ω 75 Ω ケーブル Z = 75.2 Ω Z = 50.5 Ω Z = 25.2 Ω PCB Analyzer Introduction Page 31

Accuracy Verification using a NIST Traceable Standard Measurement results are within 0.1 ohm of 25 ohm airline standard. DUT: 25 ohm airline (85052B Verification Kit) The verification kit includes measurement data and uncertainties which are traceable to National Institute of Standards and Technology (NIST). PCB Analyzer Introduction Page 32

R&R Evaluation (Average and Range Method) Single-ended Differential R&R 0.010 ohm 0.020 ohm DUT: R&R Evaluation Board (unused ports left open) DUT Length = 22 cm Single-ended 50 ohm trace Differential 100 ohm trace Test Conditions 3 operators 7 DUTs 3 measurements on each DUT Average impedance within 30-70% of the DUT length used for R&R calculations PCB Analyzer Introduction Page 33

Why Error Correction? Measure your device, not your measurement system. delay loss DIFF probe USB footswitch SE probe(s) Cables, probes, switches, and fixtures are no longer ideal at today s data rates. To get the most accurate information about the device under test, you must account for errors introduced by your measurement system, such as delay, loss, and mismatch. mismatch PCB Analyzer Introduction Page 34

Error Correction Method Comparison Two common types of error correction methods: Deskew Commonly used in time domain instruments Simple to perform Only corrects for delay delay loss mismatch Full calibration (ECal) Commonly used in frequency domain instruments Requires more standards Accounts for all major sources of error delay loss mismatch PCB Analyzer Introduction Page 35

Accuracy Considerations Same DUT with different test cable lengths, results in very different impedance values. test cable E5063A PCB Analyzer 85053B NIST Traceable 25Ω Airline 150 cm test cable 30 cm test cable Z(avg) = 27.9 Ω Deskew Z(avg) = 25.9 Ω Deskew PCB Analyzer Introduction Page 36

Accuracy Considerations Cable loss affects measurement results. E5063A PCB Analyzer test cable Rise Time = 35 ps 150 cm test cable 30 cm test cable Rise Time Insertion Loss Rise Time Insertion Loss Actual rise time:58.5 ps Insertion Loss: -4.4 db Actual rise time:40.6 ps Insertion Loss: -1.9 db PCB Analyzer Introduction Page 37

Accuracy Considerations Cable loss is removed by calibration. E5063A PCB Analyzer test cable Rise Time = 35 ps 150 cm test cable 30 cm test cable Rise Time Insertion Loss Rise Time Insertion Loss Actual rise time:35.1 ps Insertion Loss: 0 db Actual rise time:35.1 ps Insertion Loss: 0 db PCB Analyzer Introduction Page 38

Accuracy Considerations Error correction is essential to measure the true performance of the device. test cable E5063A PCB Analyzer 85053B NIST Traceable 25Ω Airline 150 cm test cable 30 cm test cable Z(avg) = 25.2 Ω ECal Z(avg) = 25.2 Ω ECal PCB Analyzer Introduction Page 39

More Languages Supported An analyzer that speaks your language PCB Analyzer Introduction Page 40

More ESD Robustness TDR Scopes Difficult to implement protection circuits inside the instrument without sacrificing performance. In addition, protection diodes cannot be placed in front of the sampling bridge as this would limit the bandwidth. This reduces the safe input voltage for a sampling oscilloscope to about 3 V, as compared to 500 V available on other oscilloscopes. Tektronix ApNote XYZ of Oscilloscopes, p17 (02/09, 03W-8605-3) External ESD protection module (80A02) available, but rise time is degraded. Rise time degradation from 28ps to 37ps with 80E04 TDR module. Single-channel protection, but only four slots are available. Additional cost of $4K/module. PCB Analyzer Introduction Page 41

More ESD Robustness E5063A PCB Analyzer Higher robustness against ESD, because protection circuits are implemented inside the instrument for all ports, while maintaining excellent RF performance. ESD protection circuits inside the instrument To ensure high robustness against ESD, E5063A PCB Analyzer is tested for ESD survival according to IEC801-2 Human Body Model (150 pf, 330Ω). RF Output Center pins tested to 3000 V, 10 cycles. Proprietary ESD protection chip significantly increase ESD robustness, while at the same time maintaining excellent RF performance (24.8 ps rise time for 18 GHz models). PCB Analyzer Introduction Page 42

Configuration E5063A PCB Analyzer = E5063A + Option 011 E5063A => frequency domain Option 011 => time domain and PCB GUI Model/Option Description E5063A ENA Series Network Analyzer Test set options (choose one): E5063A-245 2-port test set, 100 khz to 4.5 GHz E5063A-285 2-port test set, 100 khz to 8.5 GHz E5063A-2H5 2-port test set, 100 khz to 18 GHz Software option (mandatory): E5063A-011 Time Domain Analysis / Test Wizard Note: Option 011 is a superset of Option 010. Option 010 is not available separately. PCB Analyzer Introduction Page 43

Typical Configuration ENA Mainframe E5063A-245: 100 khz to 4.5 GHz, 2P E5063A-285: 100 khz to 8.5 GHz, 2P E5063A-2H5: 100 khz to 18 GHz, 2P Time Domain / Test Wizard Option (E5063A-011) USB footswitch U1810B USB Coaxial Switch, DC to 18 GHz, SPDT DIFF probe SE probe(s) ECal Module N4431B for E5063A-245/285 N4433A for E5063A-2H5 Third Party Solutions TDR Passive Probes (*1) USB Footswitch USB Barcode Reader (*1) Any TDR passive probe can be used with the PCB Analyzer. PCB Analyzer Introduction Page 44

Agenda PCB Market Overview Challenges to Obtain Measurement Accuracy and Correlation E5063A ENA Series PCB Analyzer Introduction Summary PCB Analyzer Introduction Page 45

Summary TDR Oscilloscopes with Offset Compensation Method Users need to make sure measurement devices and measurement standards for compensation have the same impedance value If the above condition is not satisfied, 1. measurement errors cannot be properly compensated resulting measurement inaccuracy 2. measurement results between channels or TDR instruments can differ from each other Keysight E5063A PCB Analyzer Measurement errors can be completely removed with full calibration for 1. accurate measurements 2. measurement correlation between different channels and instruments PCB Analyzer Introduction Page 47

PCB Analyzer Introduction Page 48

Breakthrough Developments in TDR/TDT Measurement Technology TDR/TDT 信號完整性量測新進展蔡怡杏 Flory Tsai 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載

Agenda Introduction TDR/TDT and S-parameters Hot TDR Measurements Intel Delta-L Q&A 5 min 15 min 10 min 5 min 5 min 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 2

Introduction As data rates increase: Design margins decrease: inter-symbol interference (ISI) increases due to channel/interconnect losses and reflections Trend toward parallel/multi-lane architectures increases crosstalk concerns Characterize signal integrity issues using Scope and Vector Network Analyzer (VNA) based TDR/TDT Solutions 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 3

Measurement needs Target applications Demanding Applications for TDR/TDT Measurements SI Lab (PCBs, interconnects, connectors ) Production Test of cables, connectors, interconnects S-Parameter generation for de-embedding Active Device Characterization Research Fast edge and high BW to isolate impedance issues Easy to use. Many channels Low capital cost High accuracy Fast calibration Easy to generate S- parameter files Hot TDR measurements High accuracy (fast step, high BW) Flexibility Fast to program Compliance Applications 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 4

VNAs vs TDR Oscilloscopes ENA-TDR is a great solution for <20 GHz Best dynamic range compared to TDR faster acquisition time ESD robustness Ideal for Hot TDR Keysight can address the needs of all your customers (offer both NA and TDR solutions) PNA provides the highest performance S-Parameter measurements PNA is the Gold Standard for S-Parameter measurements superior dynamic range vs TDR (from any vendor) (especially important for low levels of crosstalk, but ultra-low sensitivity is often not required for digital designs) preferred by RF/Microwave engineers Complement each other (not compete). TDR/TDT very intuitive and easy-to-use, preferred by many digital designers useful for quick troubleshooting, fault location analysis measures closer to DC (vs 10 MHz) less expensive than NA having similar BW As a bonus it is a scope too! 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Most SI labs have both solutions. Page 5

Agenda Introduction TDR/TDT and S-parameters Hot TDR Measurements Intel Delta-L Q&A 5 min 15 min 10 min 5 min 5 min 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 6

Quick Review TDR Time Domain Reflectometry (TDR) Impedance measurements Locate the position and nature of each discontinuity Propagation/Time delay Excess Reactance (Capacitance or Inductance) Effective dielectric constant TDR (Impedance Profile) 2 1 3 4 5 6 FFT S-parameters (Return Loss) 1. Reference Plane 2. Connector Launch 3. Uncoupled TX Line 4. Coupled Diff TX Line 5. Connector 6. Open Circuit 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 7

Quick Review TDT Time Domain Transmission (TDT) Step Response Propagation/Time delay Propagation velocity Rise time degradation Near-end crosstalk (NEXT) Far-end crosstalk (FEXT) Skew FFT TDT (Step Response) S-parameters (Insertion Loss) 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 8

S-Parameters and TDR/TDT Port 1 Port 2 Port 3 Port 4 Four-port single-ended device Return Loss or TDR Insertion Loss or TDT Near End Crosstalk (NEXT) Far End Crosstalk (FEXT) Frequency Domain Parameters FFT IFFT Time Domain Parameters 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 9

Together, TDR/TDT and S-Parameters provide tremendous insight TDR Z profile TDT Step Response S11 Return Loss S21 Insertion Loss TDR and S21 are most intuitive, insightful. 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 10

What TDR edge speed should I use? Edge speed determines two important parameters: 1. TDR Resolution: The faster the edge, the closer two impedance discontinuities can be identified as separate events on the TDR trace. D min = Ɛ = dielectric constant of the transmission system c = speed of light in a vacuum. For Ɛ = 4 and system rise time of 8 ps, D min < 1mm. D D 2. Max S-parameter frequency A step with a fast edge has higher frequency content and enables S- parameter testing to a higher frequency. 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 11

What TDR edge speed should I use? Select a solution based on your application: Too fast: you ll see impedance discontinuities that will not affect the real signals in your design (you ll waste time fixing things that do not matter) Too slow: discontinuities are masked Choose your TDR edge speed: 1. Full Characterization Rule of Thumb : use TDR edge speeds that are minimum 2x faster than the rise times of your design 2. Compliance Test: use 20%-80% TDR edge speed specified by Standard 500ps 200ps 100ps 25ps 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 12

TDR Two-Event Resolution (Spatial-Resolution ) To increase the two-event resolution of the TDR system, three items are considered: 1. Increase the speed of the step generator 2. Increase the bandwidth of the oscilloscope 3. Minimize the bandwidth-limiting effects of the test system - minimize use of adapters, cabling - use good quality fixturing - compensate for losses using TDR calibration (de-embedding) 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 13

86100D DCA-X with N1055A 35/50 GHz TDR/TDT Modules A fully integrated TDR/TDT/S-parameter measurement system that provides calibrated impedance and S- parameter analysis on up to 16 channels in real-time. Fastest TDR Edge Speed Up to 8 ps (typ) 50 GHz, yields highest TDR resolution. Single-ended and Differential Device Testing (True-Mode Stimulus) N1055 A Options: N1055A Bandwidth 35 GHz (2.92mm)* 50 GHz (1.85mm) N1055A Channel Count 2 Channels per module* 4 channels per module N1055A Connector Type Male Female * upgradeable Calibrated impedance and S-parameter results displayed in real-time Electronic Calibration (ECal) DC-67 GHz module support Calibration Made Easy using ECal modules or mechanical SOLT standards. Built-in Electrostatic Discharge (ESD) Protection Adjustable TDR Edge Speed 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Ultra-thin remote heads - minimize adapters/cables Highest Channel Count - Up to 16 channels per mainframe. High-Bandwidth Oscilloscope N1055A operates in receiver-only mode. Page 14

86100D DCA-X with N1055A 35/50 GHz TDR/TDT Modules Performance & Accuracy: Compare the Steps: Keysight Raw Step, risetime = 9.5 ps Keysight Calibrated Step, risetime = 8.4 ps Keysight Calibrated Step, risetime = 11.6 ps 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 15

TDR Resolution < 1mm 25 GHz Bandpass Filter 1 Impedance Profile 3 Zoom 1 2 4 5 6 7 1.1 mm S-parameters All features visible! 2 3 4 5 6 7 TDR resolution < 1mm Point 3 = taper Points 4, 5, 6 = 3 capacitors spaced by < 1mm 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Bandpass Filter Page 16

86100D DCA-X New Software FlexDCA Demo 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 17

Agenda Introduction TDR/TDT and S-parameters Hot TDR Measurements Intel Delta-L Q&A 5 min 15 min 10 min 5 min 5 min 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 18

db ohm Why Measure Hot TDR? Hot TDR measurement is the impedance analysis of active devices under actual operation conditions. Typically, impedance of the device in the OFF state and ON state (Hot TDR) is significantly different. Impedance may vary with the data rate as well. TDR(Time Domain) OFF Return Loss (Freq Domain) 1333Mbps (active) 334Mbps (active) 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 1G 2G 3G 4G 5G 6G 666M freq(hz) Page 19

Multiple Reflections 1. Signal transmitted from Tx Eye Degradation Tx Channel Rx 3. Re-reflection from Tx due to impedance mismatches... 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 2. Partial reflection from Rx due to impedance mismatches Page 20

Source Termination Effects Source Impedance NOT Matched Source Impedance Matched 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 21

Many Standards Require Hot TDR Measurements 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 22

Serial ATA 7.4.13 Return Loss and Impedance Balance Transmitters (Tx): When measuring output impedance of transmitters the operating condition shall be during transmission of MFTP. This is to assure the measurement is performed during a mode of operation that represents normal operation. MFTP from Tx Measurement Challenge How to avoid effects of the transmitter signal on measurements? 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 23

Minimizing Errors from the Transmitter Signal TDR Scopes VNA Tx t t wideband receiver captures all of the signal energy from the transmitter t Tx t t narrowband receiver minimizes the effects of the data signal from the transmitter t fc freq fc freq time Extensive averaging is necessary to obtain a stable waveform. 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 time In many cases, averaging is not necessary to obtain a stable waveform. Page 24

Avoid Spurious Feature TDR(Time Domain) Return Loss (Freq Domain) Fluctuations due to Tx signal Spurs due to Tx signal 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 25

Avoid Spurious Feature TDR(Time Domain) Return Loss (Freq Domain) 1-click Operation From the data rate (user input), spurious frequencies are determined and automatically avoided during the sweep. 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 26

What is ENA Option TDR? The ENA Option TDR is an application software embedded on the ENA, which provides an one-box solution for high speed serial interconnect analysis. Time Domain Frequency Domain 3 Breakthroughs for Signal Integrity Design and Verification Eye Diagram Simple and Intuitive Operation Fast and Accurate Measurements ESD protection inside High ESD Robustness www.keysight.com/find/ena-tdr 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 27

Summary of Hot TDR Multiple reflections due to impedance mismatches significantly impact signal integrity. Typically, impedance of the device in the OFF state and ON state (Hot TDR) is significantly different. Impedance may vary with the data rate as well. Therefore it is essential to characterize the device under actual operating conditions. www.keysight.com/find/ena-tdr 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 28

Agenda Introduction TDR/TDT and S-parameters Hot TDR Measurements Intel Delta-L Q&A 5 min 15 min 10 min 5 min 5 min 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 29

Intel Delta-L A new VNA-based solution to effectively move the Via effect for PCB diff. trace insertion loss characterization Issue & Methodology How to De-embedding the Via? Total Solution ❶ 4-port E5071C ❷ Intel Test Coupon Delta-L Loss Characterization ❸ Probes & S/W Intel Web: http://www.intel.com/content/www/us/en/processors/xeon/delta-l-methodology-for-electrical-characterization-guide.html Video On-line: http://youtu.be/rhqwqi_-5kk 台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 30

台灣是德科技股份有限公司以是為本以德致遠專注量測 75 載 Page 31

High-Frequency Material Measurement Techniques Sep. 17, 2014 Kenny Liao 廖康佑資深專案經理是德科技 Confidential 1

Agenda Page 2 What are we measuring What are the applications What are our solutions PCB Materials Test

Permittivity and Permeability Definitions Permittivity (Dielectric Constant) Permeability 0 r interaction of a material in the presence of an external electric field. ' r j " r 0 ' " r j r interaction of a material in the presence of an external magnetic field. Dk Page

Electromagnetic Field Interaction Electric Fields STORAGE LOSS Magnetic Fields Permittivity MUT Permeability r ' r j " r STORAGE r ' r j " r LOSS Page

Loss Tangent tan " r ' r '' r r ' r tan D 1 Q Energy Lost per Cycle Energy Stored per Cycle Df D Dissipation Factor Q Quality Factor Page

Relaxation Constant t t = Time required for 1/e of an aligned system to return to equilibrium or random state, in seconds. 100 ' r Water at 20 o C 10 t 1 1 2 c f c " r 1 1 most energy is lost at 1/t 10 100 f, GHz Debye equation : ( ) s 1 j t Page

Applications for Materials Measurement Industry Electronics Aerospace/Defense Industrial materials Food & Agriculture Forestry & Mining Pharmaceutical & Medical Applications / Products Capacitor, substrates, PCB, antenna, ferrites, magnetic recording heads, absorbers, SAR phantom materials, sensor Stealth, RAM (Radiation Absorbing Materials), radomes Ceramics and composites: IC package, aerospace and automotive components, cement, coatings, bio-implants Polymers and plastics: fibers, substrates, films, insulation materials Hydrogel: disposable diaper, soft contact lens Liquid crystal: displays Rubber, semiconductors and superconductors Other products containing these materials: tires, paint, adhesives, etc. Food preservation (spoilage) research, food development for microwave, packaging, moisture measurements Moisture measurements in wood or paper, oil content analysis Drug research and manufacturing, bio-implants, human tissue characterization, biomass, chemical concentration, fermentation Page 7

Probe Kit / Fixture Portfolio Material types Materials measurement software 85071E Liquid 16452A Gel Liquid test fixture 85070E Dielectric probe Semi-solids (Powder) Solid Substrate Toroidal core 16453A 16451B Dielectric test fixture 16454A 85072A 85071E-Exx Split post dielectric resonators (SPDR) Magnetic material test fixture 10 GHz split cylinder resonator DC 1 khz 1 MHz 1 GHz 10 GHz 20 GHz 50 GHz 100 GHz Frequency Page 9

Lots of methods. Which is Best? It Depends on: Frequency of interest Form of material (i.e., liquid, powder, solid, sheet) Expected value of ε r and μ r Required measurement accuracy Material properties (i.e., homogeneous, isotropic) Sample size restrictions Destructive or non-destructive Contacting or non-contacting Temperature Page

LCR and Impedance Analyzer Solutions High Loss Liquid 16452A Liquid Fixture (Parallel Plate) 85070E Dielectric Probe Kit Solid 16454A Magnetic Test Fixture Low Loss 16451B Parallel Plate 16453A Parallel Plate 1KHz 1 MHz 10 MHz 1Ghz 3 Ghz Page

Parallel Plate Solutions Page E4990A 16451B E4991B with option 002 16453A 16452A Liquid Test Fixture

Inductance Solutions Page

Network Analyzer Solutions High Loss Liquid 85070E Dielectric Probe Kit 85071E Transmission Line Solid 85071E-100 Free Space Low Loss 85071E-300 Resonant Cavity Open Resonator (KeyCom) 200MHz --1 GHz 20GHz 50Ghz 110GHz 1.1 THz Page

Coaxial Probe System Computer (Optional for PNA or ENA) Network Analyzer (or E4991A Impedance Analyzer) GP-IB, LAN or USB 85070E Software (included in kit) 85070E Dielectric Probe Calibration is required Page

Coaxial Probe Material assumptions: effectively infinite thickness Page Reflection (S 11 ) non-magnetic isotropic homogeneous no air gaps or bubbles r

Three Probe Designs High Temperature Probe 0.200 20GHz (low end 0.01GHz with impedance analyzer) Withstands -40 to 200 degrees C Flanged design allows measuring flat surfaced solids Slim Form Probe 0.500 50GHz Low cost consumable design Fits in tight spaces, smaller sample sizes For liquids and soft semisolids only Performance Probe 0.500 50GHz Withstands -40 to 200 degrees C Hermetically sealed on both ends, OK for autoclave Food grade stainless steel Page

Coaxial Probe Example Data Page

Pred Cal Martini Meter! 100 95 100 real 98.5 real 99.5 real 99.0 real 97.6 98.0 real real 97.1 real 96.6 real 95.796.2 real real 95.2 real 90.9 real 90 87.0 real 85 83.3 real 80 80.0 real 80 85 90 95 100 Measured Y Infometrix, Inc. 5 Page

Transmission Line System Computer (Optional for PNA or ENA) Network Analyzer GP-IB, LAN or USB 85071E Materials Measurement Software Calibration is required Sample holder connected between coax cables Page

Transmission Line Sample Holders Coaxial Waveguide Page

Transmission Line Material assumptions: sample fills fixture cross section Page l no air gaps at fixture walls flat faces, perpendicular to long axis Known thickness > 20/360 λ Reflection (S ) 11 Transmission (S ) 21 r and r

Transmission Example Data Page

Transmission models in the 85071E Software Algorithm Measured S-parameters Output Nicolson-Ross S11, S21, S12, S22 ε r and μ r NIST Precision S11, S21, S22 ε r Fast Transmission S21, S12 ε r Poly Fit 1 S11, S21, S12, S22 ε r and μ r Poly Fit 2 S12, S21 ε r Stack Two S21, S12 (2 samples) ε r and μ r Page

Reflection models in the 85071E Software Algorithm Measured S-parameters Output Short Backed S11 ε r Arbitrary Backed S11 ε r Single Double Thickness S11 (2 samples) ε r and μ r Page

Transmission Free-Space System Computer (Optional for PNA or ENA) Network Analyzer GP-IB, LAN or USB 85071E Materials Measurement Software Calibration is required Sample holder fixtured between two antennae Page

Non-Contacting method for High or Low Temperature Tests. Free Space with Furnace Page

Transmission Free-Space Material assumptions: Flat parallel faced samples Sample in non-reactive region l Beam spot is contained in sample Known thickness > 20/360 λ Reflection (S11 ) Transmission (S21 ) r and r Page

Free Space Example Data Page

Resonant Cavity System Computer (Optional for PNA or ENA) Network Analyzer GP-IB or LAN Resonant Cavity Software No calibration required Resonant Cavity with sample connected between ports. Page

Resonant Cavity Fixtures Agilent Split Cylinder Resonator IPC TM-650-2.5.5.5.13 ASTM 2520 Waveguide Resonators Split Post Dielectric Resonators from QWED Page

Resonant Cavity Technique fc = Resonant Frequency of Empty Cavity fs = Resonant Frequency of Filled Cavity empty cavity Q c Qc = Q of Empty Cavity Qs = Q of Filled Cavity Vs = Volume of Empty Cavity Vc = Volume of Sample S21 r r 1 Vc 4V s V c fc 2V 1 Q s ASTM 2520 s f f f c s f s 1 Q c 2.3 0.0 Page

Resonant Cavity Technique r r 1 Vc 4V s V c fc 2V 1 Q s s empty cavity fc = Resonant Frequency of Empty Cavity sample inserted fs = Resonant Frequency of Filled Cavity Q s Q c Qc = Q of Empty Cavity S21 Qs = Q of Filled Cavity f s f c f Vs = Volume of Empty Cavity Vc = Volume of Sample f ASTM 2520 s f s 1 Q c 2.303 0.00 Page

Resonant Cavity Technique r r 1 Vc 4V s V c fc 2V 1 Q empty cavity fc = Resonant Frequency of Empty Cavity sample inserted fs = Resonant Frequency of Filled Cavity Q s Q c Qc = Q of Empty Cavity S21 Qs = Q of Filled Cavity f s f c f Vs = Volume of Empty Cavity Vc = Volume of Sample s ASTM 2520 s f s f s 1 Q c 2.3 0. Page

Resonant Cavity Example Data Page

Resonant vs. Broadband Transmission Methods Low Loss materials Thin Films and Sheets Resonant Yes e r resolution 10-4 Yes 10GHz sample thickness <1mm Broadband No e r resolution 10-2 No 10GHz optimum thickness ~ 5-10mm Calibration Required No Yes Measurement Frequency Coverage Single Frequency Broadband or Banded Page

Microwave Materials Measurement Products Model Number Description 85070E 020 030 050 400 Dielectric Probe Kit High Temperature Probe Slim Form Probe Performance Probe Advanced Functionality 85071E 100 200 300 400 Materials Measurement Software Free Space Calibration Arch Reflectivity Software Resonant Cavity Software Advanced Functionality 85072A 10GHz Split Cylinder Resonant Cavity Page

Materials Ordering Convenience Specials Model Number Description 85071E E19 E03 E04 E15 E07 Split Post Dielectric Resonators from QWED 1.1GHz 2.5GHz 5GHz 15GHz 22GHz 85071E E02 E01 E22 E18 E24 Quasi-optical products from Thomas Keating Ltd. 60-90GHz Quasi-optical Table 75-110GHz Quasi-optical Table 90-140GHz Additional set of horns for above tables 220-326GHz Additional set of horns for above tables 325-500GHz Additional set of horns for above tables Page

Summary of methods Page

Resources Websites Materials Test Equipment webpsite Application notes Basics of Measuring the Dielectric Properties of Materials (5989-2589EN, Aug 2013) Solutions for Measuring Permittivity and Permeability with LCR Meters and Impedance Analyzers (5980-2862EN, Sep 2013) Split Post Dielectric Resonators for Dielectric Measurements of Substrates (5989-5384EN, Jul 2006) Tech overviews, brochures and selection guides Measuring Dielectric Properties using Agilent s Materials Measurement Solutions - Brochure (5991-2171EN, Apr 2013) Choosing the Optimal Method for Testing Dielectric Properties of Materials - Application brief (5991-2599EN, Jun 2013) 85070E Dielectric Probe Kit (5989-0222EN, Jun 2012) 85071E Materials Measurement Software (5988-9472EN, Jun 2012) 85072A 10-GHz Split Cylinder Resonator (5989-6182EN, May 2012) LCR Meters, Impedance Analyzers and Test Fixtures Selection Guide (5952-1430E, Apr 2012) Videos PNA Demo videos > Materials Measurements YouTube (external Agilent) 85070E dielectric probe with FieldFox (2009) 85072A 10 GHz Split Cylinder Resonator (2009) Customer viewable training slides Microwave Dielectric Spectroscopy Workshop (2004) Free Space Materials Measurement Seminar (Jun 2005) 44 Page

PCB Materials Test

Available Measurement Methods PCB Dk/Df Characterization Methods Parallel plate Transmission line Resonant cavity Typical frequency range Up to 1 GHz 100 MHz to 110 GHz 1 GHz to 40 GHz Parallel plate Transmission line Resonant cavity Dk: Dielectric constant (Relative permittivity) Df: Dissipation factor (Loss tangent) www.keysight.com/find/materials Page 46

Resonant Cavity Method with SPDR Fixture ENA/PNA Network Analyzer with Materials Measurement Software E5063A SPDR fixture Sample Key Features Superior accuracy to measure lower loss materials Convenient and fast measurement of substrates, PCB and thin films Non-destructive and noncontacting measurement (Just insert your MUT into the slit of the fixture) Enable to measure multi-layer PCB SPDR: Split Post Dielectric Resonator (SPDR supplier: QWED) Page 47

Cross-Section of SPDR fixture Sample l Dielectric Resonator Dielectric Resonator h h G Coupling Loop z Metal Enclosure SPDR Performance (QWED) Nominal freq. of the basic line of SPDRs: 1.1/1.9/2.45/3.2/5/10/15 GHz (Customization is possible) Uncertainty of the real permittivity (typical): ±1 % (Providing that the average thickness of MUT is measured with accuracy ±0.7 % or better; Δε/ε = ±(0.15+Δh/h) %) Uncertainty of tan δ (typical): ±3 % Resolution of tan δ: 2 x 10-5 Operational temperature range: -200 ~ +110 Page 48

Parallel Plate Method Permittivity Evaluation E4990A E4991B with option 002 16453A 16451B 16451B LF/HF Solution (20 Hz to 30 MHz) RF Solution (1 MHz to 1 GHz) Page 49

Appendix Page 50

SPDR (for 10 GHz) Page 51

SPDR (Split Post Dielectric Resonator) SPDR Resonators for different frequencies Page 52

Page Cp G Equivalent Circuit Thickness = t Electrodes (Area=A) Co : Air Capacitance 0 0 0 C G j C C C j C j G Y p p 0 0 * C G j C C p r.tan 0 0 r r p r A G t A C t Parallel Plate Method 53

Materials Measurement Methods 85070E 16454A Inductance Coaxial Probe 16451B/53A Parallel Plate (Capacitance) Resonant Cavity Transmission Line Free Space 85072A Page 54

Thank You Confidential 55

References R N Clarke (Ed.), A Guide to the Characterisation of Dielectric Materials at RF and Microwave Frequencies, Published by The Institute of Measurement & Control (UK) & NPL, 2003 J. Baker-Jarvis, M.D. Janezic, R.F. Riddle, R.T. Johnk, P. Kabos, C. Holloway, R.G. Geyer, C.A. Grosvenor, Measuring the Permittivity and Permeability of Lossy Materials: Solids, Liquids, Metals, Building Materials, and Negative-Index Materials, NIST Technical Note 15362005 Test methods for complex permittivity (Dielectric Constant) of solid electrical insulating materials at microwave frequencies and temperatures to 1650, ASTM Standard D2520, American Society for Testing and Materials Janezic M. and Baker-Jarvis J., Full-wave Analysis of a Split-Cylinder Resonator for Nondestructive Permittivity Measurements, IEEE Transactions on Microwave Theory and Techniques vol. 47, no. 10, Oct 1999, pg. 2014-2020 J. Krupka, A.P. Gregory, O.C. Rochard, R.N. Clarke, B. Riddle, J. Baker-Jarvis, Uncertainty of Complex Permittivity Measurement by Split-Post Dielectric Resonator Techniques, Journal of the European Ceramic Society No. 10, 2001, pg. 2673-2676 Basics of Measureing the Dielectric Properties of Materials. Agilent application note. 5989-2589EN AM. Nicolson and G. F. Ross, "Measurement of the intrinsic properties of materials by time domain techniques," IEEE Trans. Instrum. Meas., IM-19(4), pp. 377-382, 1970. Improved Technique for Determining Complex Permittivity with the Transmission/Reflection Method, James Baker-Jarvis et al, IEEE transactions on microwave Theory and Techniques vol 38, No. 8 August 1990 P. G. Bartley, and S. B. Begley, A New Technique for the Determination of the Complex Permittivity and Permeability of Materials Proc. IEEE Instrument Meas. Technol. Conf., pp. 54-57, 2010. Page

LiTek s PCB Delta-L Measurement Solution Presented by 厘科科技 (LiTek Technologies) Leo Lee ( 李文福 ) By 2014

Background of PCB Delta-L Measurement. Triggered by Intel. Remove the via effect that not included in our measurement. Use a easy way to implement (Delta of Loss). Suitable for any layers in the PCB. Real Differential Loss Measurement Page 2

Page 3

Page 4

LiTek s PCB Delta-L Solution E5071C, 20G TpNA software LiTek s PCB Probe Agilent E-Cal PCB (DUT) Page 5

Guide Pin Information Page 6

Why TpNA for PCB Delta-L Measurement? A PC based GUI Software, friendly control interface. Help to finish all steps including: >>Perform E-Cal >>Setup all features for SDD21 measurement (optional including Time- Domain Differential Impedance) >>Recording all layers, Reference and Longer Length PCB s insertion loss in the TpNA. >>Auto Calculating Delta-L result in the TpNA >>Easily export a result by just one button press Who can finish these jobs One who can use a PC and without the knowledge to use ENA. Page 7

How to do the measurement 當 TpNA 開啟後, 建立一個 PCB for Delta-L 的專案. 使用者將 E-Cal 分別與 ENA 的 4 個埠連接後, 按下 TpNA 的執行 E-Cal 按鈕,TpNA 就會完成所有的 E-Cal 校正 Page 8

How to do the measurement Page 9

How to do the measurement Page 10

How to do the measurement 當 Reference Length 及 DUT Length 的 Insertion Loss 量測完畢,TpNA 就會自動的計算 Delta-L Page 11

How to do the measurement 在報表頁中, 使用者按下輸出報表的按鈕,TpNA 就會完成一份 Excel 格式的報表 Page 12

E5063A Application 1

Agenda E5063A Introduction Option 011(Start up Time Domain) Option PIC TDR /SET2DIL Software 2

外觀介紹 LCD 觸碰瑩幕硬體按鍵 USB 接孔測試孔 (N-Type) 設備電源開關接地端口瑩幕輸入孔 USB488 外部控制 SSD GPIB 24 Bit I/O Port 外部觸發輸入端子參考訊號端子 USB 接孔網路接孔設備電源開關設備電源接孔設備序號標籤設備校正標籤 3

Agenda E5063A Introduction Option 011(Start up Time Domain) Option PIC TDR /SET2DIL Software 4

Main Frame 支持五國語言 : 英文繁體中文簡體中文韓文日文 5

Prepared to Start 開始設定 6

步驟一, 儀器組態設定點選進階模式 7

步驟一, 儀器組態設定外接二個 U1810B, 故選擇此項選擇單端與差分量測 8

步驟二, 誤差校正 Port1 &2 ; Port3 皆選擇 Deskew 校正並按下執行按鈕進行校正程序 Open Thru Load Load Thru Open 9

步驟三, 測試流程設定 10

步驟四, 測試清單列表 1 2 3 步驟 1: 可無限制增加量測的參數步驟 2: 選擇編輯的項目步驟 3: 點選編輯來進行細項的設定, 記得每一組項目都需執行編輯中的測試待測物長度 11

步驟五, 編輯量測開路定義 : 差分線長測試開路定義 : 單動線長測試 12

步驟六, 編輯限制值 13

步驟七, 開始量測及結果儲存 14

Agenda E5063A Introduction Option 011(Start up Time Domain) Option PIC TDR /SET2DIL Software 15

步驟一, 儀器組態設定由外部控制 ( 儀測軟件 ) 控制選擇差分量測 16

步驟二, 誤差校正選擇 Deskew 與振幅 ; 損失校正並按下執行按鈕進行校正程序 Open Thru Load 17

步驟三, 開啟儀測 TDR 軟件 1. 點選桌面 TDR2.2.5.0 的軟件 2. 執行 TDR Automation Test 程式 18

步驟三, 開啟儀測 TDR 軟件 19

功能說明 -File 刪除量測資料 20

功能說明 -System 選擇軟體的語系 ( 中 / 英文 ) 選擇測量的功能 ( 阻抗 / 損失 ) 查看軟件資訊 21

功能說明 - 測試參數 1. 設定產品資訊 3. 設定單分 / 差動測試 4. 設定量測的誤差值設定說明 : 1. 設定產品量測資料, 以便於產生報告時可以使用 2. 設定量測存資訊, 可在量測時同步記錄曲線資料及圖形 3. 設定量測類別, 開路位置及量測範圍 4. 設定量測結果的允收值 ( 多數以百分比 ) 5. 設定板材 ( 目前並無使用 ) 6. 設定量測結果的判定依據 ( 多數以平均值為標準 ) 2. 設定量測儲存資訊 5. 設定板材 6. 結果判定依據 22

功能說明 - 操作設定 1. 設定取樣資料 2. 設定觸發方式 3. 設定測試程序 4. 設定顯示單位 5. 操作畫面設定 7. 開路定義程序設定說明 : 1. 設定取樣資料, 無需設定 2. 設定觸發方式, 可設定鍵盤及腳踏開開觸發量測 3. 設定測試程序, 即設定若多量數據量測時, 切換下一個量測程序的判斷條件 : a. 過關 : 量測值若符合設定值即繼續量測下一點 b. 手動 : 由人工切換至下個量測點 c. 下一步 : 無論量測數據為何, 量測後繼續量測下一點 d. 過關記錄 : 量測合格後才記錄 4. 設定量顯示單位 : ( 多數以英吋表示 ) 5. 操作畫面設定, 無需設定 6. 設定待測物序號 : 此設定目的為多組阻抗量測時, 可自動於報告上顯示時, 產生這些數據資料 7. 開路定義程序 : 請選用依自動判定 6. 設定待測物序號 23

步驟四 - 測試參數 1. 開新檔 2. 編輯 3. 存檔 4. 確認 5. 離開 24

步驟五 - 操作設定 25

步驟六 - 開路定義探頭不要連接任何待測物此範例為兩筆量測數據 : 第一組為單端阻抗 ; 第二組為差分阻抗第一筆量測 : 連接探頭至待測物第二筆量測 : 連接探頭至待測物 26

步驟七 - 開始量測 27

SET2DIL measure mode 請選擇 SET2DIL 模式 28

步驟一, 設定測試參數點選測試參數, 以進入設定畫面軟件主頁面 1 4 3 2 離開 Step1: 開啟第一組測試條件 Step2: 新增欲測試的項目 Step3: 點選測試項目並按編輯修改如下參數 Step4: 設定完畢後, 請記得儲存後再離開這個畫面內外層量測選擇 Intel Defination 29

步驟二, 執行校正 30

步驟三, 開始測量 31