ET Envelope Path from digits to PA

Transcription

1 pushing the envelope of PA efficiency ET Envelope Path from digits to PA Gerard Wimpenny Nujira Ltd ARMMS Conference 19 th /2 th November 212

2 Agenda Envelope Processing ET PA Characterisation Isogain shaping CFR shaping Envelope Amplifier Design Requirements Sources of Impairment Integrated Modulator Distributed Modulator 2

3 ET System Anatomy Envelope detection: most accurate if performed in digital domain Envelope shaping: Determines relationship between RF power and PA supply voltage Envelope Amplifier: High BW, Low Noise, High efficiency Amplifier used to generate PA supply voltage Delay Alignment: ET requires accurate (~ns) timing alignment between envelope and RF paths. Most accurate / repeatable if performed in digital domain ET PA: ET can be applied to standard fixed supply PA. Improved performance possible by optimising PA for ET operation 3

4 ET PA System Principles 7 6 In compressed region O/P power is determined by supply voltage RF input power has little influence Efficiency (%) Fixed Supply 4.5V 4.V 3.5V 3.V 2.5V 2.V 1.5V ISO26dB Fixed In transition region O/P power is determined by both supply voltage and RF input power Pout (dbm) In linear region O/P power is determined by RF input power supply voltage has little influence 4

5 Envelope Processing Basics Swing Range Optimise efficiency of combined modulator /PA Prevent gross PA nonlinearity due IV curve knee Envelope Shaping Control envelope bandwidth Optimise efficiency Can be used to linearise PA Timing Alignment Timing error leads to memory effect (AM-PM) Fine adjustment necessary (~1ns)

6 PA Characterisation Methods PA Characteristics must be known to determine Shaping table Test methodology PA current measurement Supply impedance Supply bandwidth requirements ET Efficiency prediction ET Linearity prediction Parameters measured Swept CW testing Bench PSU Low (decoupling Capacitor) Low (Bench PSU) Poor, due to PA die heating Poor, due to PA die heating Gain (AM:AM), Efficiency Pulsed RF /DC testing Instrumentation grade current probe, ~5 us resolution Low (decoupling Capacitor) Low (Bench PSU) Good, if short pulses (~1 us, 1% duty cycle). Fair (if device has low AM/PM) Gain (AM:AM), Efficiency Dynamic supply modulation Challenging high BW with high common mode voltage current sense Requires low impedance dynamic supply (no decoupling) High (~6 MHz BW) V. Good V. Good (if device has low memory effects) Gain (AM:AM), Phase (AM:PM), Efficiency Phase measurement possible in principle but accuracy poor due to heating effects and phase reference wander No phase measurement 6

7 AM/PM Input Surfaces Fixed Supply Voltage Gain contours Fixed Supply Voltage Phase contours Input Gain Surface Input Phase Surface 7

8 Isogain Contours 27dB IsoGain shaping contour 25dB IsoGain shaping contour Phase peak flattening c.f. fixed supply Input Gain Surface Low voltage phase collapse Input Phase Surface 8

9 Isogain Shaping Functions 27dB gain 25dB gain Input Gain Surface Isogain Shaping Functions 9

10 Useful 2D Slices - Efficiency 25/27dB IsoGain contours Fixed Supply contours Output Efficiency Surface Output Efficiency locus 1

11 2D Slices AM/AM, AM/PM AM/AM distortion removed Residual PM distortion Output Gain Output Phase 11

12 Predicted Performance AM/AM ACPR AM/PM Predicted Efficiency = 67.7% Waveform = HSUPA / 5.4dB PAPR Shaping = Isogain 24dB 12

13 Measured Performance AM/AM ACPR AM/PM Predicted Efficiency = 67.7% Measured Efficiency = 67.6% Waveform = HSUPA / 5.4dB PAPR Shaping = Isogain 24dB 13

14 Shaping Table based CFR Envelope Amplifier Max Vout Isogain shaping modified to introduce soft clipping e.g using Rapp function (VCFR) Unmodified Isogain shaping (Viso) V iso Desired PA gain profile V = CFR 1 p æ p Viso 1 V ö ç æ ö ç ç è è pk ø ø p = Smoothness factor + V Limiting voltage pk = 14

15 Increased Pout using CFR 1 Isogain shaping CCDF 8 7 Increased Efficiency Increased Mean Power Probability Efficiency (%) dB 4.9dB Modified Isogain shaping CCDF db above mean Output Signal Statistics 2 1 LTE Power pdf RF Output power (dbm) Controlled use of CFR allows Increased mean power and efficiency for given PA device periphery 15

16 Software Defined PA RF Spectrum Soft clipping Hard clipping 75% Clipping level (2.5dB CFR) 85% Clipping level (1.4dB CFR) PSD (db) PSD (db) x x x Reference -2 Spectrum 2 4 (No clipping) Offset from Carrier (Hz) x 1 7 Offset from Carrier (Hz) x x x p=4 p=6 p=1 p=1 Shaping Table based CFR allows dynamic configuration of PA s Power / ACPR / Efficiency characteristics x 1 7 x

17 Agenda Envelope Processing ET PA Characterisation Isogain shaping CFR shaping Envelope Amplifier Design Requirements Sources of Impairment Integrated Modulator Distributed Modulator 17

18 Envelope Amplifier Requirements High Bandwidth Low Noise Power High Efficiency High Bandwidth (e.g 4ch WCDMA, 2MHz LTE, 2x 1MHz WiMAX) Envelope Bandwidth ~3x RF Bandwidth Cannot be achieved with switcher only architecture Low Noise / Distortion Required to meet ACPR specifications Many factors to consider Requires high Tracking Accuracy High Efficiency Must consider combined PA / modulator efficiency Linear supply would be pointless Power Must maintain BW and Noise at increased power levels 18

19 ET Impairment Categories System (Env & RF paths) RF/Env Delay match RF/Env Gain match PA AM/AM and AM/PM RF Path Noise Thermal Quantisation Linearity PA Memory effects Bias Thermal Envelope Path 19

20 Envelope Path Impairments Shaping Accuracy Tracking Accuracy Noise DAC Quantisation Env Amp Thermal Switcher breakthrough Linear Amp PSRR Frequency Response Amplitude Group Delay flatness Env Amp Distortion Harmonic Crossover Env Amp to PA Interaction Env Amp Output Impedance PA Interconnect Impedance PA Non Linear Load Impedance 2

21 Tracking Accuracy Explained The difference between ideal and measured supply waveform after removal of DC offset, gain and timing errors Analogous to EVM for modulated signals Tracking error analysis is useful diagnostic tool: RMS, Peak, Spectrum Residual Gross Gross Small modulator gain timing and timing tracking error error error V 3 V Ideal and measured waveforms Tracking Error 21

22 Supply Noise RF Conversion PA in compression Supply Noise & Distortion modulates RF carrier PA can be considered as mixer O/P spectrum is convolution of Supply and PA input Spectra Conversion factor (Supply Sensitivity) for noise on supply to RF sidebands is similar to ideal AM modulator (mixer) MHz Supply Spectrum RF Spectrum 4MHz Start: Hz Stop: 2. MHz -1 Start: 1.85 GHz Stop: 2.5 GHz 4MHz test tone added to Envelope Amplifier O/P (whilst amplifying 5MHz WCDMA signal) Corresponding RF sidebands 22

23 Measured Supply Sensitivity An ideal AM modulator is described by: where modulation index h = This can be re-expressed in terms of carrier and LSB and USB components M A y(t ) = [ A+ M cos( w m t)]sin( w c t) y(t ) = Asin( w c t ) + R[sin(( w c +w m ) t ) + sin(( w c -w m ) t )] where for an ideal AM modulator R = Average DC drain voltage 2.62V Measured 4MHz injected tone level 17.3mV rms Calculated RF sideband level for ideal AM modulator -49.6dBC Measured RF sideband level -51dBC PA Supply Sensitivity (db) -1.4dB PA Supply Sensitivity (%) 85% M 2 DV V env env DV V rf rf DV V DV V rf rf env env 23

24 Integrated Modulator Example Coolteq.L Boost and Buck capable Battery depletion resilience Increased PA peak Power Slow switching Buck converter provides LF power Fast switching multilevel converter provides HF power Error Amplifier cleans up output 24

25 Distributed Modulator Example - Coolteq.u Exciter Modulation Envelope signal generation Digital Pre Distortion RF Upconvert DC input Envelope input 4 x Coolteq.u High Accuracy Tracking module (HAT ) Coolteq.u Power Supply Module (PSM) RF SPLITTER HAT HAT HAT HAT PA PA PA RF COMBINER Scalable O/P Power Allows multiple PAs per Power Supply Module (PSM) Allows Envelope path Linear Amplifier to be placed close to PA PA supply impedance minimised RF out db PAPR RF in RF Driver PA 25

26 Conclusions Understanding of PA characteristics key to achieving good ET performance. Careful selection of shaping table contents allows optimisation key ET system performance metrics ET is a simple concept, but attention must be paid to multiple potential sources of impairment to realise full potential 26

27 pushing the envelope of PA efficiency