Solutions for Power Amplifier Test in PXI PA Reference Solutions. Update on Release 3

Size: px

Start display at page:

Download "Solutions for Power Amplifier Test in PXI PA Reference Solutions. Update on Release 3"

Ashlynn Nichols
5 years ago
Views:

1 Solutions for Power Amplifier Test in PXI PA s Update on Release 3 May 2015

2 Agenda Test Requirements for UE Power Amplifiers Keysight s Power Amplifier Test Solutions PA Architecture in PXI Product Details Demo Solution Deployment Considerations Roadmap Headlines Summary, Q & A Page 2

3 Power Amplifier Test Requirements Page 3

4 Current Market Drivers impacting UE PAs Increasing data rates New modulation formats like LTE have high PAPR Wider signal bandwidths - WLAN BW requirements MHz - CA Uplink Maximizing battery life UEs are becoming more sophisticated and more power hungry Global roaming Greater frequency band coverage (> 14 bands in some cases) Page 4

5 Constraints and Challenges for Test Engineers 1. Shorter PA product development cycle time demands reduced test times Measurement speed and test plan optimization 2. Greater device complexity need to make more measurements Test system automation 3. Test budget limitations Leverage existing code base Maximize utilization of existing and new test assets 4. Test system footprint considerations Mounting location, heat generation, fan noise Page 5

6 The Evolution of a Handset PA Page 6

7 Generalization of PAD architecture IQ Envelope DC Supply DC Modulator Filters & Duplexer Rx PA Mode Switch Filters & Duplexer Antenna Switch RF Transceiver Rx PA Filters & Duplexer Rx MIPI RFFE Device Control Page 7

8 What do we need to test in a PA? Basic Performance P out MAX, Gain, P 3dB, ACPR, PAE - using power servo technique Good Neighbor Harmonic generation, intermod, spectral regrowth, in-band Rx noise Compatibility Return loss, AM/AM, AM/PM, EVM, mismatch tolerance - Passive components of PAD filters, duplexers Measurements as a function of frequency and output power level Performance under ET and DPD? Page 8

9 Test Speed Potential Improvements Primary Factors impacting test throughput 1. Real Time Signal Processing Hardware based measurement acceleration (e.g. FPGA or ASIC) 2. Faster Dataset load Times High speed instrument backplanes 3. Faster State Switching and Tuning Times Digital tuning within the IF bandwidth 4. Faster host PC Embedded or desktop 5. Improved measurement Algorithms and Procedures Test vendor or user IP 6. Parallel Measurements Take advantage of distributed measurement architecture Page 9

10 Production Test Speed Improvement in real terms Example Case Study - ACPR, EVM and Harmonics Conventional approach (2-3 years ago) Pwr Servo ACPR EVM Capture & Calc Harmonics DC Measurements Current Optimized approach Total Test Time Total test time now less than 15% of previous Pwr Servo ACPR EVM Capture DC Measurements EVM Calc Harmonics Page 10

11 Keysight s Power Amplifier Test Solutions Page 11

12 Bench Top and Modular Instruments. Keysight offers the best of both for unbounded test capability in R&D or just enough in manufacturing Extensive specs and Software Integrated Easy setup Front-panel operation Upgradable Portable A shared world-class software portfolio Rapid access to new technology Compact test-sets Multichannel Flexible & reconfigurable System Design & Verification Signal Generation & Waveform Creation Signal Analysis Page 12

Benchtop Solution for R&D/DVT N6705B DC Power

Trigger Timing Alignment RF ETPA X-Series

13 Benchtop Solution for R&D/DVT N6705B DC Power Analyzer Vector Signal Analyzer N7614B Signal Studio s Envelope Tracking/DPD Control Window 33600/33500 series AWG Envelope ETPS Trigger Timing Alignment RF ETPA X-Series Signal Generator DUT X-Series Signal Analyzer Page 13

Hybrid AXIe Wide Bandwidth Solution (DPD only) R&D application 89600 Vector Signal Analyzer N7614B

14 Hybrid AXIe Wide Bandwidth Solution (DPD only) R&D application Vector Signal Analyzer N7614B Signal Studio s Envelope Tracking/DPD Control Window RF PA M8190A + PSG PXA or UXA Signal Analyzer DUT Page 14

15 Modular DVT or Production Test Configuration H3353 AWG Envelope Waveform M9210A or 2 ch Scope N6700 SMU M937xA VNA(s) USB Power Sensor(s) User application or M-App Measurement s/w N7614A PA Test s/w or User provided Triggering ETPS Vcc Vbat I/Q Waveform RF In PA RF Out M9381A VSG M9195A Digital Stimulus/Response RFFE V.. DPD Algorithms V.. M939xA 6-27 GHz VSA M9037A Embedded Controller I/Q Waveform Opt. VSA s/w M9018A 18 slot PXI chassis PXI Digital IO M9187A Automated Testing Systems Note: RF switching modules omitted for clarity Page 15

N7614B API Ref Solution SW PXIe VSG / PXIe VSA / AWG / VNA Timing Alignment RF ETPA HW System

16 Modular PXI Solution Automated DVT and Production Test N6700B with N6782A (1 to 4) 4 channel SMU N7614B Signal Studio s Envelope Tracking/DPD Data Envelope ETPS X-Series Measurement Applications N7614B API Ref Solution SW PXIe VSG / PXIe VSA / AWG / VNA Timing Alignment RF ETPA HW System Control (IVI Drivers) Power & Spectrum Measurements Waveform & Envelope Download IQ Data Upload DUT Page 16

17 Power Amplifier Architecture Page 17

18 Power Amplifier Minimize Customer Investment to Evaluate & Deploy Keysight Provided Recipe for Power Amplifier Tests using Keysight PXI Modules with LXI and Third Party Products, as needed Provides Demonstration of Capabilities, Measurement Ranges and Test Times Configurable to Test Customer Devices without Code Changes for First Level Evaluation Source Code to Minimize Porting Instruments into Customer s Test Environment for Detailed Evaluation and Deployment Focus on PA design/validation/test and production test including support for ET and DPD The bottom line Evaluate quicker, leverage Keysight s measurement knowledge, deploy with lower risk Page 18

Close_Instruments setupvsgvsafixedpin measstdacpr Load Waveforms servoinputpower measlteacpr measspecharms X-Series

19 Evaluation Software Environment Software RF PA Test Evaluation GUI Test Program: Test_WCDMA Test_LTE5MHz Test_LTE_10MHz Test_GSM Test_EVDO Test_WLAN Customer Test Executive Power Amp Test Library: Init_Instruments setupvsgvsa measpout Close_Instruments setupvsgvsafixedpin measstdacpr Load Waveforms servoinputpower measlteacpr measspecharms X-Series Measurement Applications M938xA VSG IVI-COM Driver M9391A/93A VSA IVI-COM Driver PXI Modules M937xAVNA IVI-COM Driver Signadyne AWG IVI-COM Driver Page 19

20 Characterization & Test ET/ DPD Evaluation Software Environment CFR & ET Test Conditions DPD Test Conditions Page 20

21 Characterization & Test Optimized for insight Code Example Extract LUT Values and Memory Model Polynomial Coefficients Page 21

Current Block Diagram (Release 2) N7614B Signal Studio For PA Test N7614B API Software SCPI X-Series Measurement Applications Modulation Analysis Data HW

Studio for LTE PXI VSG M9381A PA ETPS Vcc Duplexer PXI VSA M9391/3A RF measurements With ET/DPD RFFE USB DIO (PA Control) PXI VNA M937xA S- Parameter

22 Current Block Diagram (Release 2) N7614B Signal Studio For PA Test N7614B API Software SCPI X-Series Measurement Applications Modulation Analysis Data HW System Control (IVI Drivers) Power & Spectrum Measurements Waveform & Envelope Download IQ Data Upload PXI AWG SD AWG-H3353 SMU N6700B/N6782A N7624/25B Signal Studio for LTE PXI VSG M9381A PA ETPS Vcc Duplexer PXI VSA M9391/3A RF measurements With ET/DPD RFFE USB DIO (PA Control) PXI VNA M937xA S- Parameter S-parameter Measurements M9187A DIO ** (Switch/Handler Control) Page 22 ** M9187A module not directly supported in reference solution but often used in customer deployments

23 Characterization & Test Block Diagram Current Capabilities (Release 2) GUI Libraries Host PC based Drivers LXI SMU Signal Studio PXI AWG Signadyne ETPS User Characterization Application ET/DPD Measurements X-Apps PXI VSG PXI VSA RF PA Duplexer EVM PXI VNA S-Parameter Measurements USB DIO RFFE Application Software Hardware Device Under Test Page 23

Using N7614B to Evaluate a Physical Device Easy exploration of multiple PA performance improvement techniques 1: RF waveform selection 3: Shaping table selection, time

24 Using N7614B to Evaluate a Physical Device Easy exploration of multiple PA performance improvement techniques 1: RF waveform selection 3: Shaping table selection, time alignment 5: ETPS configuration 7: Standard RF analysis 2: CFR technique & parameter selection 4: DPD algorithms & coefficient export 6: Output power servo control Page 24

25 DPD Test Flow Create and Load Initial Waveforms Capture IQ Data Measure Initial AM/AM, AM/PM, DEVM Significant performance benefits are gained from an FPGA-based deployment Extract and Apply DPD Model Create and Load New Waveforms Capture IQ Data Measure AM/AM, AM/PM, DEVM More Iterations Other RF and DC Measurements Complete Page 25

PA Mode Switch Filters & Duplexer Antenna Switch

26 PAD Passive Components DC Supply IQ Envelope DC Modulator Filters & Duplexer Rx RF Transceiver Rx PA Mode Switch Filters & Duplexer Antenna Switch PA Filters & Duplexer Rx MIPI RFFE Device Control Page 26

27 Impact of Duplex Filtering An FDD radio has to isolate the transmit & receive paths This is typically implemented using a pair of SAW or FBAR band pass filters A VNA measurement shows how reflections from a filter with adequate rejection may have significant group delay ripple in S S11 return loss S11 group delay S21 pass-band E E E E E E+08 Page 27

28 S-parameter measurements for PADs Consider impact of the passive components (filters, switches, duplexer) on PA performance Isolation, insertion loss, filter pass-band, mismatch tolerance Test times are impacted and test plans need optimization Swept or Stepped coverage What test probe points are available on the PAD? Switched VNA measurements versus multiple (or multiport) VNA considerations Test time budget Number ports needed Cost comparison Page 28

29 Test Set & DUT Mode VSG/VSA measurement for Mode 1 VSG VSA Tx1 Mode 1 Tx2 VIO SCLK SDATA MIPI RFFE Out ANT1 VNA ANT2 Rx1 Rx2 Power Meter Rx3 DUT External Test Set Page 29

30 Test Set & DUT Mode VSG/VSA measurement for Mode 2 VSG Tx1 Mode 2 VIO SCLK SDATA VSA Tx2 MIPI RFFE VNA Out ANT1 ANT2 DUT mode is changed via MIPI EFFE. Rx1 Rx2 Power Meter Rx3 DUT External Test Set External test set is controlled to select DUT ports. Page 30

31 Test Set & DUT Mode VNA measurement for Mode 2 VSG Tx1 Mode 2 VIO SCLK SDATA VSA Tx2 MIPI RFFE Out ANT1 VNA ANT2 Rx1 Rx2 Power Meter Rx3 DUT External Test Set External test set is controlled to select measurement types. Page 31

32 Test Set & DUT Mode VNA measurement for Mode 3 VSG Tx1 Mode 3 VIO SCLK SDATA VSA Tx2 MIPI RFFE Out ANT1 VNA ANT2 Rx1 Rx2 Power Meter Rx3 DUT External Test Set DUT has several modes. Test set should be controlled for each mode/measurement. Page 32

33 Full Multiport Capability Increase productivity and throughput Full N-port correction capability Add up to 32-ports in a single chassis Page 33

34 Release 3 Highlights Measurement Acceleration in DVT & PT Focus on automated DVT test applications measurement acceleration Significant improvement in DPD loop measurement time Better than 10x test time reduction - Release 2 loop time approx ms - Release 3: <100 ms for full DPD loop measurement Typically ms today LUT method today - Memory Polynomial & Volterra Series in future release gateware of M9451A N7614B GUI is not supported in this release - Make use of API only Digital stimulus/response capabilities Page 34

PA Test Release 3 (May 2015) Target Market -

M9451A Measurement Accelerator - M9195A DSR

Reference Solution Libraries Now FPGA based

Customer Test Application Signal Studio for

PXI VSA PXI VNA M937xA RFFE RF PA Duplexer

35 PA Test Release 3 (May 2015) Target Market - RF PA production test - RF PA DVT with ET/DPD test Description - Measurement accelerated DPD calculations New features - M9451A Measurement Accelerator - M9195A DSR capabilities Reference Solution GUI Reference Solution Libraries Now FPGA based Drivers LXI SMU PXI AWG Signadyne H3353 ETPS Customer Test Application Signal Studio for RF PA + LTE X-Apps PXI DSR M9195A PXI VSG PXI VSA PXI VNA M937xA RFFE RF PA Duplexer PXI M9451A Application Software Hardware Device Under Test Page 35

36 DPD FPGA Speed Enhancement Architecture Shaping Table Envelope Shaper V Envelope AWG D A C ETPS I Q Reference LUT (coefficients) LUT Predistorter I Q PA In VSG D A C R F 2. Fine Sync Filter Coef 3. Extractor AM-AM, AM-PM P A Sample Offset, Gain 1. Coarse Sync and Power Estimate. I Q PA Out VSA ADC R F M9451A Measurement Accelerator Page 36

Block Diagram (new Release 3) N7614B Software

System Control (IVI Drivers) Power & Spectrum

Data Upload Measurement Accelerator M9451A SMU

Modulation Analysis VSG M9381A PA RFFE ETPS Vcc

ET/DPD Digital Stimulus Response M9195A RF PA

37 Block Diagram (new Release 3) N7614B Software for PA Test Software SCPI AWG SD H3353 Data HW System Control (IVI Drivers) Power & Spectrum Measurements Waveform & Envelope Download IQ Data Upload Measurement Accelerator M9451A SMU N6700B/N6782A X-Series Measurement Applications Modulation Analysis VSG M9381A PA RFFE ETPS Vcc Duplexer VSA M9391/3A RF measurements With ET/DPD Digital Stimulus Response M9195A RF PA Control S- Parameter VNA M937xA S-parameter Measurements PXI Digital IO M9187A Ext. Device Control Package Handler Wafer Prober Switch Matrix Page 37

38 Product Details Page 38

39 M9381A VSG Performance characteristics Frequency coverage from 1 MHz to 3 GHz or 6 GHz 10 µs switching speed with fast-tune, an exclusive baseband tuning technology innovation RF modulation bandwidth up to 160 MHz (± 0.3 to 0.5 db flatness) ± 0.15 to 1.0 db absolute amplitude accuracy Software Enhancements in 1.2 release: Shared Frequency Reference PXI backplane triggers New PLL Mode (Best Wide Offset) for WLAN EVM and GSM/EDGE ORFS optimization Improved sync out trigger for Envelope Tracking Software Enhancements in 1.3 release: Add Time Synchronous MIMO capability, channel-to-channel deviation <20 ns Page 39

Meet the M9393A PXIe Performance VSA The world s fastest, most

derived from PXA FieldFox calibrator for excellent accuracy

cutting-edge technology Novel solid-state switches for speed &

size & cost Frequency range Analysis BW Amplitude accy Tuning

40 Meet the M9393A PXIe Performance VSA The world s fastest, most accurate μw PXI VSA Leverages proven designs Downconverter derived from PXA FieldFox calibrator for excellent accuracy M9391A PXI VSA digitizer & reference modules Deploys cutting-edge technology Novel solid-state switches for speed & reliability Unique Keysight Labs production process to lower size & cost Frequency range Analysis BW Amplitude accy Tuning speed Size 9 khz to 8.4/ 14/ 18/ 27 GHz 40/100/160 MHz +/-.15 db 150 us 5 slots (4 + ref) Page 40

41 Reference Solution Page 41

M9451A Measurement Accelerator - Dedicated FPGA module - Uses same carrier card as M9195A and future AWG - Initial release supports ET/DPD measurement acceleration

42 M9451A Measurement Accelerator - Dedicated FPGA module - Uses same carrier card as M9195A and future AWG - Initial release supports ET/DPD measurement acceleration only - Uses same Keysight DPD algorithms utilized in SystemVue and N7614B PA test software - Future release will support the inclusion of user IP in the FPGA Page 42

43 PXIe Vector Network Analyzer, M937XA Driving down the size of test Full two-port network analyzer in just one slot Widest available frequency range: 300 khz to 4, 6.5, 9, 14, 20, 26.5 GHz Best PXI VNA performance in four key areas: Speed: 18 ms across 401 points Dynamic range: 114 db (9 GHz), 110 db (20 GHz) Trace noise: < db specified, < db typical Stability: ±0.005 db/ C at 4 GHz, ±0.020 db/ C at 26.5 GHz Page 43

M9195A PXIe Digital Stimulus/Response Module DSR Module Key Features Front Panel IO: 16 bidirectional channels with each channel capable of: High speed digital stimulus/response Parametric

44 M9195A PXIe Digital Stimulus/Response Module DSR Module Key Features Front Panel IO: 16 bidirectional channels with each channel capable of: High speed digital stimulus/response Parametric Measurements (PPMU) with remote sense Static digital RFFE DUT control 4 high voltage IO for flash/polyfuse tests 4 open drain ports for fixture relay control 2 GPIO for future use Advanced Features Single and multi-site configurations (max 4 per module) Pattern cyclizer for flexible per bit timing control for waveform generation with edge placement resolution of 1ns On-the-fly pattern editing without recompiling and downloading the test Channel delay adjustment to compensate for cable a fixture propagation delays Page 44

45 Demo Page 45

46 Summary of Key Points Keysight s PXI for Power Amplifier Testing 1. Proven solution in DVT and manufacturing by top tier PA vendors. 2. Enables software reuse for your solution that reduces deployment effort, time and cost. 3. Provides support for emerging UE technologies including ET, DPD and uplink CA. Market leading measurement precision and performance 4. As our test partner, you have access to Keysight s worldwide team of RF measurement experts. Page 46

47 Page 47

48 Backup Slides Page 48

49 Full Configuration - Hardware Caution: Pricing based on April 2015 US List Prices (HW) LP $ Qty List Price TTL Discount % Quote M9018A PXIe Chassis: 18-slot, 3U, 8GB/s $8,816 1 $8, $8,816 PXI Main Frame Y1212A chassis airflow slot blockers $89 2 $ $178 Y1213A EMC Filler panels $152 $ $304 Infrastructure 2 M9037A PXIe embedded PC controller, Intel I7 $6,963 1 $6, $6,963 PC VSG/VSA M9037A-WE6 Windows Embedded Standard 7 operating system (64 bit) $0 $0 0.0 $0 1 M9037A-M16 Memory upgrade from 4GB RAM to 16GB $800 1 $ $800 Y1206A Keyboard and optical mouse $0 1 $0 0.0 $0 M9381A PXIe Vector Signal Generator $0 1 $0 0.0 $0 M9381A-F06 Frequency range, 1 MHz to 6GHz $25,490 1 $25, $25,490 M9381A-B16 RF modulation bandwidth, 160 MHz $10,038 1 $10, $10,038 M9381A-M05 Memory, 512 MSa $5,019 1 $5, $5,019 M9381A-1EA High output power $3,513 1 $3, $3,513 M9381A-UNZ Fast switching $5,019 1 $5, $5,019 M9381A-300 PXIe Frequency Reference: 10 MHz and 100 MHz $4,118 $4, $4,118 RF SA/SS 1 M9393A PXIe Microwave Vector Signal Analyzer $0 1 $0 0.0 $0 M9393A-F14 Frequency range, 9 khz to 14 GHz $48,093 1 $48, $48,093 M9393A-B16 RF modulation bandwidth, 160 MHz $20,039 1 $20, $20,039 M9393A-M01 Memory,128MSa(512MB) $0 1 $0 0.0 $0 M9393A-UNZ Fast Switching $7,515 1 $7, $7,515 M9393A-300 PXIe Frequency Reference: 10 MHz and 100 MHz $0 0 $0 0.0 $0 DSR M9195A Digital Stimulus/Response Module w 16 ch PPMU $10,500 $10, $10,500 DUT control 1 M9195A-M12 Memory 125 Mb/s $1,500 1 $1, $1,500 M9195A-SR2 Max Clock Rate 250 MHz $2,000 1 $2, $2,000 M9195A-S04 Multi-site Configuration $1,000 1 $1, $1,000 Y1253A SMA Break Out Cable 2m $350 1 $ $350 Digital IO M9187A PXI Digital IO: 32 ch, 0.3 to 50V $1,679 $1, switch matrix $1,679 control 1 AWG AWG-H3353 Signadyne AWG (3rd Party) $7,000 1 $7, $7,000 DSP M9451A $12,000 $12, $12,000 ET and DPD Measurement Accelerator 1 M9451A DPD DPD and ET measurement $8,000 1 $8, $8,000 Power N6710B N6700B Custom configured Modular Power System 400 W, GPIB, LAN, USB, LXI $2,685 1 $2, $2,685 Pwr Sup Supply $5,852 $11, $11,704 N6781A ATO Source/Measure Unit,2 Quadrant, 20V/1A or 6V/3A, 20W 2 M9371A PXI VNA to 6.5GHz $15,495 1 $15, $15,495 VNA M9371A-102 Adv Features $5,175 $5, $5,175 S parameters 1 M9371A-551 N-port calibration $10,350 1 $10, $10,350 $235,343 $235,343 Page 49

50 Full Configuration - Software Caution: Pricing based on April 2015 US List Prices (SW) List Price Qty List Price TTL Quote N7614B Sig Studio for PA Test software $0 1 $0 0.0 $0ET/DPD SignalStudio N7614B-EFP Env Tracking Fixed Perp Lic $2,500 1 $2, $2,500 N7614B-FFP DPD Fixed Perp lic $2,500 1 $2, $2,500 N7624B Signal Studio for LTE/LTE Advanced FDD $0 1 $0 0.0 $0 LTE FDD N7624B 9FP Connect to M9381A and M9252A, fixed perpetual license $0 1 $0 0.0 $0 N7624B HFP Basic LTE FDD R9, fixed perpetual license $6,086 1 $6, $6,086 N7625B Signal Studio for LTE/LTE Advanced TDD $0 1 $0 0.0 $0 LTE TDD N7625B 9FP Connect to M9381A and M9252A, fixed perpetual license $0 1 $0 0.0 $0 N7625B EFP Basic LTE TDD R9, fixed perpetual license $4,682 1 $4, $4,682 N7600B Signal Studio for W CDMA/HSPA+ $0 1 $0 0.0 $0 W CDMA N7600B 9FP Connect to M9381A and M9252A, fixed perpetual license $0 1 $0 0.0 $0 N7600B EFP Basic W CDMA / HSPA R7, fixed perpetual license $4,682 1 $4, $4,682 N7602B Signal Studio for GSM/EDGE/Evo $0 1 $0 0.0 $0 GSM/EDGE N7602B 9FP Connect to M9381A, fixed perpetual license $0 1 $0 0.0 $0 N7602B EFP Basic GSM/EDGE, fixed perpetual license $4,682 1 $4, $4,682 N7612B Signal Studio for TD SCDMA $0 1 $0 0.0 $0 TD SCDMA N7612B 9FP Connect to M9381A, fixed perpetual license $0 1 $0 0.0 $0 N7612B EFP TD SCDMA, fixed perpetual license $5,819 1 $5, $5,819 N7601B Signal Studio for cdma2000/1xev DO $0 1 $0 0.0 $0 C2K/1xEVDO N7601B 9FP Connect to M9381A, fixed perpetual license $0 1 $0 0.0 $0 N7601B EFP Basic cdma2000, fixed perpetual license $3,646 1 $3, $3,646 N7601B FFP Basic 1xEV DO, fixed perpetual license $3,646 1 $3, $3,646 M9071A 2TP GSM/EDGE measurement application $5,354 1 $5, $5,354 M9071A XTP GSM/EDGE combined measurement app $2,053 1 $2, $2,053 M9072A 2TP CDMA2000 Measurement Application $6,508 1 $6, $6,508 M9073A 1TP W-CDMA measurement application $8,248 1 $8, $8,248 M9073 2TP HSDPA/HSUPA measurement app $6,353 1 $6, $6,353 M9073 3TP HSPA+ measurement app $2,151 1 $2, $2,151 X-Apps M9076A 1TP 1xEV-DO measurement application $8,901 1 $8, $8,901 M9079A 2TP HSPA measurement application $3,344 1 $3, $3,344 M9079A 1TP TD-SCDMA measurement application $7,801 1 $7, $7,801 M9080B 1TP LTE FDD measurement application $13,426 1 $13, $13,426 M9080B 2TP LTE-A FDD $4,500 1 $4, $4,500 M9082B 2TP LTE-A TDD $4,500 1 $4, $4,500 M9082B 1TP LTE TDD measurement application $13,426 1 $13, $13,426 Ref Soln s/w Y1299A-004 PA Reference solution startup kit $25 1 $ $25 $124,833 $124,833 $360,176 Total List Price Page 50

51 Page 51

52 AWG/VSG Synchronization System Initialization M9300A Reference Module in PXI Chassis Slot 10 Configure PXI Chassis to route PXI_Trigger2 from trigger segment including the M9311A module to the entire chassis Typically 1 ->2 -> 3 Configure M9381A VSG: Send internal 10 MHz to chassis backplane - Lock PXI chassis 10 MHz reference to M9300A reference which is used for VSG and VSA Route Synchronization Output Trigger to M9311A front panel Trig2 and PXI_Trigger2 - Backplane trigger will trigger VSA and AWG - Front panel available to trigger other instruments, scope Configure Synchronization Output Trigger Signal: us Pulse Width - Synchronization output type to Per ARB Configure Start ARB Playback on 10 MHz Edge to true - All VSG waveform playback starts on rising edge of 10 MHz signal Page 52

53 AWG/VSG Synchronization Start Waveform Procedure Timing Page 53

54 Envelope Tracking Signals Real Time Timing Adjustment +/- 250 ns Range 1 ps resolution Adjust Timing without restarting or reloading waveform Page 54

Example Shmoo Plot Example application: Vary multiple pin V or I on the FEM by generation waveform patterns and evaluate the impact on other device parameters.

55 Example Shmoo Plot Example application: Vary multiple pin V or I on the FEM by generation waveform patterns and evaluate the impact on other device parameters. The M9195A can generate waveform patterns with superior timing precision for better characterization insight. Example from Page 55

56 Optimizing for Repeatability and Test Time Immediate Trigger with Full Waveform Power Meas Power Meas Power Meas Power Meas Power Meas Power Meas Fastest Measurement Time: Waveform looping continuously Measure at Any Time No Delay to Wait for External Trigger Power Servo measurement iterations Measurement Repeatability May be Bad: Variations in Power Level during Waveform will Add To Measurement Uncertainty in Power and ACPR Measurements Can be Improved by increasing measurement duration Page 56 56

57 Optimizing for Repeatability and Test Time External Trigger with Full Waveform: VSG Sync Output as External Trigger to VSA Power Meas Slowest Measurement Time: Measure only at One Point in Waveform Delay to Wait for External Trigger will be ½ Total Time of Waveform, on Average No Measurements During Most of Waveform if Measurement Time is Fast Measurement Repeatability Improved: Always measuring at Same Time within Waveform No Variation in Modulation Signal During Measurement Page 57 57

58 Optimizing for Repeatability and Test Time External Trigger with Short Waveform Power Meas Power Meas Power Meas Power Meas Power Meas Power Meas Power Meas Power Meas Faster Measurement Time: Still Measure only at One Point in Waveform Waveform Time Reduced to be Slightly Longer than Longest Measurement Time Delay to Wait for External Trigger will be ½ Total Time of Short Waveform, on Average Much Less than with Full Waveform Measurement Repeatability Improved: Always measuring at Same Time within Waveform No Variation in Modulation Signal During Measurement Not Valid for Burst Waveforms such as GSM or LTE-TDD Page 58 58

59 Optimizing for Short Waveform Length Waveform Length Too Short, adds to Test Time Power Acquisition C al c VSG Set Power Acquisition C al c VSG Set Waveform Time too short, VSA does not see second trigger, must wait for next trigger Ideal Waveform Length Power Acquisition C al c VSG Set Power Acquisition C al c VSG Set 500 us Waveform Time slightly longer than complete Servo Cycle Page 59 59

60 Three Methods to improve PA performance Page 60

61 The Problems 1. Wideband signals with high peak / average power 2. Lots of frequency bands 3. Complexity of system integration & modelling Page 61

The SC-FDMA LTE Uplink is designed to have a lower PAPR than OFDM 3. Without clipping, it varies from 5.8 db (6RB) to 9 db (100RB).

62 The PAPR of an LTE Uplink The PAPR depends on the measurement interval 1. If the signal content changes over time, the PAPR is probably changing too 2. The SC-FDMA LTE Uplink is designed to have a lower PAPR than OFDM 3. Without clipping, it varies from 5.8 db (6RB) to 9 db (100RB). With waveform shaping it s typically 6 db 4. With PUCCH+PUSCH, PAPR depends on the relative power levels. ~ 8 db measured over 500 us 5. Intra-band Carrier Aggregation can increase PAPR to 12 db ( RB) UL PUSCH PAPR 7.2 db UL Ref.Signal PAPR 3.25 db Page 62

63 Techniques to Mitigate PAPR Related Issues Know your alphabet Amplifier sizing Back-off Crest Factor Reduction Distort RF PA input signal to reduce distortion produced by PA compression Choose the method and compression level to match the PA characteristics Digital pre-distortion Distort the RF PA input to cancel distortion produced by the PA Choose the method to suit the baseband capability and the PA characteristics Envelope tracking Drive the RF PA with a fully modulated RF signal and synchronously modulate the PA supply voltage Adjust the RF PA operating point by controlling the RF / Envelope magnitude ratio, called the shaping table, or de-troughing function Page 63

64 Making LTE Uplink Signal Measurements 89601B VSA application CCDF A single IQ data capture gives multiple measurements 3GPP EVM Channel frequency response AM/AM or AM/PM 25 RB PAPR = 7.2 db Measurement made at signal generator output ACPR Waveform envelopes & difference (delta EVM) Gain compression Demodulation EVM using RS only Page 64

65 CFR - The How & Why Reduce the PAPR to limit distortion produced in the PA CFR is not a linearization technique Page 65

66 CFR Techniques Peak Windowing Peak / average set to 5 db Adds out-of-band distortion, ACPR degraded EVM degraded only around amplitude peaks Channel response flat Page 66

67 CFR Techniques Clipping & Filtering Peak / average set to 5 db Generates inband distortion only (EVM increases) Modifies channel frequency response Page 67

68 Making ACP-EVM Performance Trade-off Using CFR CFR using clipping & filtering CFR gives a few db improvement, enough to provide some design margin Non-linearity is NOT changed Page 68

69 DPD - The How & Why Improve linearity of PA output signal by modifying the input Page 69

70 DPD When a Look Up Table is Sufficient Check for dispersion (spreading) in AM/AM, AM/PM graphs In-band frequency un-flatness causes dispersion in both gain & AM/PM graphs LTE channel frequency response quantifies the signal error 400 ns -26 db channel added using LTE Signal Studio Page 70

the spectrum of the input signal gets much wider when DPD is

71 DPD: Look Up Table PA in Average Power Tracking Mode Little spreading in AM/AM indicates a good fit for LUT DPD Notice how the spectrum of the input signal gets much wider when DPD is applied ACPR reduction of 5 db constrained by power supply clipping Page 71

72 DPD: Volterra When the distortion depends on frequency Reflections from duplexing filter at the PA output introduce a channel frequency response LUT DPD cannot remove effect of frequency by itself. Volterra DPD algorithm introduces frequency dependent non-linear corrections Memory Polynomial is a simplified implementation Page 72

73 ET - The How & Why Continuously adjust the supply voltage to change the PA s operating point Improve battery life and RF amplifier performance Reduce heat dissipation Allow use of CMOS with more demanding signals Page 73

74 Using ET to Reduce PA Non-linearity The ET bonus Page 74

75 Shaping Table Examples Design of the shaping table determines the linearity of the PA A designer can make trade offs between efficiency, linearity and compression level Page 75

Testing RFIC Power Amplifiers with Envelope Tracking. April 2014

Testing RFIC Power Amplifiers with Envelope Tracking April 2014 1 Agenda Key Test Challenges Addressing Test Challenges New emerging technologies such as envelope tracking and DPD and their implications