Welcome. Steven Baker Founder & Director OpenET Alliance. Andy Howard Senior Application Specialist Agilent EEsof EDA Agilent Technologies, Inc.

Transcription

1 Welcome Steven Baker Founder & Director OpenET Alliance Andy Howard Senior Application Specialist Agilent EEsof EDA 1

2 Outline Steven Baker, OpenET Alliance What problem are we trying to solve? What is Envelope Tracking? Envelope Tracking Market Status Andy Howard, Agilent EEsof EDA Simulating Envelope Tracking What type of amplifier model? Characterizing the power amplifier Modulation signals Preliminary simulations Investigating non-idealities 2

3 Envelope Tracking An Introduction Steven Baker, OpenET Alliance

4 Introduction What is the OpenET Alliance? A non-profit corporation that operates as an open membership industry association With a mission to drive the development and adoption of envelope tracking techniques, components and products in the wireless communications industry, to deliver more efficient RF transmitters We do this by explaining the benefits of envelope tracking and enabling the ecosystem through standardisation, tools, research and networking Members include baseband & transceiver vendors, PA vendors, power modulator vendors, OEMs and leading academics Page 4 Copyright 2012 OpenET Alliance

5 Introduction Outline What s the problem? What s envelope tracking and how does it help? Market status Page 5 Copyright 2012 OpenET Alliance

6 Introduction What s the Problem? Increasing Peak-to-Average Power Ratio (PAPR) Increasing band fragmentation Standard Launched Spectral Efficiency (bpshz -1 ) PAPR (db) Fixed Supply Efficiency (typ. %) GSM WCDMA HSUPA LTE As PAPR goes up PA efficiency goes down Band fragmentation broadband operation even lower efficiency Page 6 Copyright 2012 OpenET Alliance

7 Introduction Envelope Tracking at 50,000 Feet Constant supply Variable envelope Variable supply Variable envelope Page 7 Copyright 2012 OpenET Alliance

8 Device Efficiency (%) Introduction Instantaneous Efficiency V 1.5 V 2.0 V 2.5 V 3.0 V 3.5 V 4.0 V 4.5 V 5.0V Fixed supply efficiency locus Envelope Tracking efficiency locus 30 Signal PDF 20 Average Peak 10 PAPR ~8dB Pout (dbm) Page 8 Copyright 2012 OpenET Alliance

9 Introduction Envelope Tracking at 5,000 Feet V batt I I-DAC Duplexer Q Q-DAC Power Amplifier To Rxr Page 9 Copyright 2012 OpenET Alliance

10 Introduction Envelope Tracking at 5,000 Feet V batt Preenvelope gain Envelope shaping Postenvelope gain & offset Env- DAC Envelope Tracking Modulator Magnitude calculator B A I I-DAC Duplexer Delay Q Q-DAC Power Amplifier To Rxr Page 10 Copyright 2012 OpenET Alliance

11 Introduction Market Status LTE Infrastructure Successfully deployed in fully commercial 40W remote radio heads LTE Terminals All major silicon platform vendors are working on ET solutions today If you re working on LTE terminals or LTE terminal technologies you ll be working with envelope tracking in the next year to 18 months LTE Small Cells Growing interest in ET as a solution for small cells power consumption, thermal management, broadbanding Page 11 Copyright 2012 OpenET Alliance

12 Outline Steven Baker, OpenET Alliance What problem are we trying to solve? What is Envelope Tracking? Envelope Tracking Market Status Andy Howard, Agilent EEsof EDA Simulating Envelope Tracking What type of amplifier model? Characterizing the power amplifier Modulation signals Preliminary simulations Investigating non-idealities 12

13 Simulating Envelope Tracking Modulate drain voltage in response to input signal envelope Shaping curve attempts to maintain constant gain or gain compression Sample Input Signal Power Apply Shaping Modulate Drain Voltage Modulated Input Signal Load 13

14 Outline Simulating Envelope Tracking What type of amplifier model? Characterizing the power amplifier Modulation signals Preliminary simulations Investigating non-idealities 14

15 What type of amplifier model? Do you have a transistor-level amplifier model? If you just have a physical amplifier, can you measure its X-parameters? 15

16 Outline Simulating Envelope Tracking What type of amplifier model? Characterizing the power amplifier Modulation signals Preliminary simulations Investigating non-idealities 16

17 Characterizing the power amplifier Checking for memory effects and bandwidth Obtaining data for constant-gain shaping Obtaining data for constant-gain-compression shaping 17

18 Testing for memory effects Inject two tones; sweep their spacing Inequality between lower and upper intermodulation distortion tone amplitudes => memory effects 18

19 Testing for memory effects (a different amplifier) Memory effects significant for tones > ~100 khz spacing Amplifier may still work well for narrower band signals 19

20 An alternative way of testing for memory effects using Envelope simulator 2 tones, spaced 150 khz Vload phasor versus time 2 tones, spaced 1.5 MHz 20

21 Obtaining data for constant-gain shaping table Simulate amplifier with HB; sweep both Pavs (available source power) and drain bias Specify desired gain: If available source power = 10.5 dbm, set drain bias to 3.5 V to keep gain = 11 db 21

22 Obtain new shaping table for different gain, instantly If available source power = dbm, set drain bias to 3.0 V to keep gain = 10 db 22

23 Drain Bias Voltage Obtaining data for constant-gain-compression shaping table Desired gain compression set to 1.5 db: Drain Bias Voltage Drain bias versus source power to maintain constant gain compression Available Source Power, dbm 23

24 Drain Bias Voltage Obtain new shaping table for different amount of gain compression, instantly Desired gain compression set to 2.0 db: Drain Bias Voltage Drain bias versus source power to maintain constant gain compression Available Source Power, dbm 24

25 Outline Simulating Envelope Tracking What type of amplifier model? Characterizing the power amplifier Modulation signals Preliminary simulations Investigating non-idealities 25

26 Modulated RF signal considerations Complexity of amplifier model (behavioral or transistor-level?) Longer simulation time Shorter simulation time Signal Complexity (primarily length) and measurement (PAE, AM-to-AM, AM-to-PM, or specification-compliant EVM?) Early in design process: use short signals, obtain quick results, make changes Late in design process: use longer signals, obtain specification-compliant results 26

27 Obtaining modulated RF signals Numerous sources exist in Ptolemy Wireless Libraries I and Q baseband data may be modulated onto RF carrier Obtain from Agilent Signal Studio software Example of LTE signal and its statistics 27

28 Outline Simulating Envelope Tracking What type of amplifier model? Characterizing the power amplifier Modulation signals Preliminary simulations Investigating non-idealities 28

29 Short simulation with fixed bias 29

30 Fixed bias simulation results Power delivered to load, dbm 30

31 Short simulation with envelope tracking bias Ideal, behavioral components detect input power, read corresponding drain bias from shaping table in file 31

32 Simulation results with envelope tracking bias Power delivered to load, dbm 32

33 Things to consider Power amplifier gate bias and minimum drain bias Amplitude of input modulation signal Type of shaping table constant gain or gain compression Performance sensitivity to external source and load impedances 33

34 Ptolemy co-simulation with a Vector Signal Analysis sink 34

35 Outline Simulating Envelope Tracking What type of amplifier model? Characterizing the power amplifier Modulation signals Preliminary simulations Investigating non-idealities 35

36 Modeling a finite slew rate in the bias modulator Slew rate = 40 V/microsecond Slew rate = 10 V/microsecond 36

37 Modeling a time delay difference between RF and bias modulation signals Time delay delta = 0 nanoseconds Time delay delta = 20 nanoseconds 37

38 Summary Agilent ADS well suited for: Modeling power amplifiers Investigating power amplifier performance Generating and simulating modulated signals Investigating envelope tracking schemes Modeling various non-idealities 38

39 For more information Download these examples from the Agilent EEsof Knowledge Center: to+improve+efficiency Application note on envelope tracking simulation: On envelope tracking: Or contact me directly: 39

40 You are invited Dr. Peter H. Aaen RF Modeling and Measurement Technology Team Freescale Semiconductor You can find more webcasts