60 GHZ PA Design Wireless HDMI/WPAN Application. Demonstrate Complete MMIC ADS Desktop Design Flow

Transcription

1 60 GHz Power Amplifier Design for Wireless HDMI (WPAN) Agilent EEsof EDA Innovative Solutions, Breakthrough Results Michael Thompson US Application Engineer District Manager October 13, 2009 Agilent Technologies, 2009 Agilent EEsof EDA Overview July 2009 Objectives 60 GHZ PA Design Wireless HDMI/WPAN Application Demonstrate Complete MMIC ADS Desktop Design Flow ADS Desktop DRC ADS Desktop LVS Showcase performance of TriQuint s TQP foundry process Process Statistics PDK, with MMIC Layout Toolbar Page 2 1

2 Complete Flow Overview For ADS 2009 Update 1 Page 3 Complete MMIC ADS Desktop Flow Page 4 2

3 Project Timeline And Lesson Reaffirmed The PA design work was completed in about week, including layout This was the first time I had used the TriQuint s TQP13 Process The project once again shows the need for EM, and the advantages of integrated EM Page 5 Presentation Topics Specifications Process Selection Device Stability Device Matching Initial Linear Simulations Non-Linear Simulations EM Simulations WPAN Standard Simulations Page 6 3

4 WPAN Specification WPAN is defined in IEEE P c/D08, "Part 15.3: Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for High Rate Wireless Personal Area Networks (WPANs): Amendment 2: Millimeterwave based Alternative Physical Layer Extension", 2009 A personal operating space is a space about a person or object that typically extends up to 10m in all directions and envelops the person whether stationary or in motion. It s used to convey information over relatively short distances among a relatively few participants. Unlike WLAN, connections effected via WPAN involve little or no infrastructure. This allows small, power efficient, inexpensive solutions to be implemented for a wide range of devices. Page 7 Application Supports streaming uncompressed audio and video at up to 1080p resolution, 24 bit color at 60 Hz refresh rates Delivery compressed A/V streams and data Unlicensed operation at 60 GHz with a typical range of at least 10 m for highest resolution HD A/V Smart antenna technology to enable non line of sight (NLOS) operation Two PHY major modes, both OFDM High rate PHY (HRP): up to 4 Gb/s, beam formed directional link Low rate PHY (LRP): up to 10 Mb/s in omni-directional mode Page 8 4

5 Channel Plan Up to 4 independent HRP channels, depending on the regulatory domain. Channel index HRP frequency plan Start frequency Center frequency Stop frequency GHz GHz GHz GHz GHz GHz GHz GHz GHz GHz GHz GHz LRP frequency plan Page 9 TQP13 Process Overview 0.13um optical self aligned gate depletion phemt process Coupled with high density capacitors, epi resistors, thin film resistors (TFR), and 2 layers of gold interconnect. With a typical Ft of 95GHz, TQP13-N is used for V-band automotive radar, and high frequency point-to-point radio applications. Utilizes optical lithography process in place of traditional e- beam to achieve cost efficiencies um phemt (TQP13-N) Process Cross-section Page 10 5

6 Start By Understanding The Design Medium Microstrip Line Reference A Designer s Guide To Microstrip Line Microwaves, May 1977 I.J. Bahl & D.K. Trivedi Page 11 LineCalc Provides More Accurate Results Page 12 6

7 One Of The Problems With Long Stubs Page 13 Understanding Device Stability Page 14 7

8 Simulation Results In Preconfigured Data Display Page 15 Simulation Results In Preconfigured Data Display Page 16 8

9 Varying The Bias Voltage And Observing Device Behavior Page 17 Review Device Geometry To Better Understand How And Where Feedback Can Be Implemented Page 18 9

10 Considered Series Feedback In The Source Page 19 Removing One Via Has Considerable Effect On Stability Since the device s Source has proven to be a very sensitive area if we were to base our design on adding stability elements to this area the design would also prove sensitive or not very producible. Page 20 10

11 Neither Inductive Nor Resistive Shunt Feedback Helped Page 21 Let s Consider Drain To Gate Feedback To get a better understanding of how to model this feedback mechanism we need to consider how it will be implemented Page 22 11

12 Feedback Implementation The gate to drain feedback requires routing around one of the vias. The additions to ADS 2009 PDKs allows easy and quick routing. Trace Routing Trace Routing Extraction Into Microstrip Elements Page 23 Feedback Based On Layout Implementation Page 24 12

13 Quick Monte Carlo Check With Process Statistics Enabled Page 25 ADS Has Routines To Build Matching Networks Page 26 13

14 S-Parameter Application Note Page 27 Using Rollett s Stability Analysis And Conjugate Matching Page 28 14

15 With A Little Tuning Design Was Quickly Centered to WPAN Frequency Range Page 29 Three Stage Amplifier Page 30 15

16 Linear 3-Stage Amplifier Performance Page 31 Non-Linear Amplifier Performance Page 32 16

17 Using ADS Built-In Amplifier Test Benches Page 33 Power Sweeps/Gain/Transisent Response Page 34 17

18 Load-Pull Contours Page 35 Overlay Of Power Contours And PAE Page 36 18

19 Using Design Sync And Some Hand Edits To Create Initial Layout Page 37 Desktop DRC TQP13 PDK MMIC Toolbar Mail DRC PDK Toolbars are implemented for each individual process. Page 38 19

20 ADS Desktop DRC Step 1 Invoke the Desktop DRC Tool Step 2 Select Rule File Step 3 Click on Run Page 39 ADS Desktop DRC Results Viewer Page 40 20

21 ADS Desktop LVS Page 41 Components not in Schematic Page 42 21

22 Final TriQuint Layout With Clean DRC Run Page 43 3D Rendering Of Design Page 44 22

23 Reticle Cell Layout Page 45 Close-up Of Device Feedback Page 46 23

24 Three Stage Amplifier Page 47 Individual Devices Page 48 24

25 Test/Cal Standards Page 49 Final Assembly Ready For Test Page 50 25

26 Cascade Elite 300 Prober Page 51 Positioning Amplifiers Page 52 26

27 Amplifier Under Test Page 53 Measured Vs. Simulated Measurements In theory there is no difference between theory and practice. In practice there is. -Yogi Berra Page 54 27

28 Momentum Simulation For First Stage, Start Of Investigating Differencies Page 55 Schematic Vs. Schematic W/Momentum Results Page 56 28

29 The Additional Couple Within The First Stage Shows A Frequency Shift To The Lower Frequency Range First Stage performance with empirical models First Stage performance with Momentum results Page 57 WPAN Wireless Library Product Features HRP Source Preamble: - Time domain - Frequency domain Header - PHY header - MAC header - CRC, Scrambling - RS coding, Outer interleaver - 1/3 Convolutional coding, bit interleaver, tone interleaver Data - Sub-packet structure - CRC, Scrambling - RS coding, Outer interleaver - Puncturing convolutional coding, bit interleaver, tone interleaver OFDM modulation Measurement Test Benches: Spectrum CCDF Waveform EVM HRP Receiver Frame and frequency synchronization De-framing Cyclic prefix removing Channel estimator Phase tracking Equalizer Demapper Decoder Measurement Test Benches: RawBER and BER on AWGN RX Sensitivity Page 58 29

30 HRP Transmitter Test Benches Page 59 Initial WPAN Simulations Page 60 30

31 Next Steps On the Post Webinar Survey, please tell us if you d like to See a video of full characterization/measurement results Attend a live hands-on-workshop in your area Be invited to future webinars RF/Microwave IC & Package Analysis & Design w/3d Electromagnetic design software Power Amplifier & Mixer design & modeling w/x-parameters Page 61 Q&A If you ask me anything I don't know, I'm not going to answer. - Yogi Berra Page 62 31

32 32