Power Amplifier Design Utilizing the NVNA and X-parameters

IMS2011 Power Amplifier Design Utilizing the NVNA and X-parameters Loren Betts 1, Dylan T. Bespalko 2, Slim Boumaiza 2 1 Agilent Technologies, Santa Rosa CA, USA 2 University of Waterloo, Waterloo ON, Canada

Presentation Outline Amplifier Design Considerations using X-parameters Industry Challenges Introduction to X-parameters End-to-End Power Amplifier Design with X-parameters Load-Dependent X-parameters High Power X-parameter Measurement Configuration Final PA Results

Next Generation Communication Systems Power Amplifier Industry Challenges: Research & development of new semiconductor processes for next generation components Develop smaller, higher power, more efficient active device designs Reduce cost and the development time to bring product to market Issues: Components are exhibiting more & more nonlinear behavior (often by design to increase efficiency) Models of newer technologies (such as GaN) may not exist or may not provide an accurate description of all the component behavior.

How do I optimize desired Amplifier Specifications? PAE (accuracy<3%) Vcc Ant Zo=50ohm Matching Network Icc Matching Network 2W max VSWR=2.5 max ACPR (accuracy<1db) PAE= Power Added Efficiency ACPR= Adjacent Channel Power Ratio VSWR= Voltage Standing Wave Ratio

Evolution of the Tools & Measurements Patchwork S-Parameters S-Parameters + Figures of Merit NVNA X-Parameters TOOLS: SS & Oscilloscope Grease pens and Polaroid cameras Slotted line Power meter MEASUREMENTS: Bode plots Gain SWR Scalar network analyzers Y & Z parameters TOOLS: Vector Network Analyzer MEASUREMENTS: Gain Input match Output match Isolation Transconductance Input capacitance TOOLS: NA SA/SS/NFA Power meter Oscilloscope DC Parametric Analyzer MEASUREMENTS: Gain compression, IP3, IMD PAE, ACPR, AM-PM, BER Constellation Diagram, EVM GD, NF, Spectral Re-growth ACLR, Hot S22 Source and Load-Pull

S-parameters: Linear Measurement, Modeling & Simulation Measure with linear VNA: Small amplitude sinusoids Incident S 21 Transmitted Linear Simulation: Matrix Multiplication S-parameters b1 = S11a1 + S12a2 b2 = S21a1 + S22a2 a 1 S b 11 2 Reflected DUT S 22 Port 1 Port 2 Reflected b 1 a 2 Transmitted S 12 Incident Model Parameters: Simple algebra S ij = b i a j ak = 0 k j

S-parameter Measurements xt () yt () 0 a 1 0 b 1 00 e 1 10 e 1 01 e 1 11 e 1 1 a 1 1 b 1 S 21 S11 S22 S 12 1 b 2 1 a 2 11 e 2 01 e 2 10 e 2 00 e 2 0 b 2 0 a 2 b i = k S ik a k b = S a + S a 1 11 1 12 2 b = S a + S a 2 21 1 22 2 b1 S11 S12 a1 b = S S a 2 21 22 2 S-Parameter Definition To solve, VNA s traditionally use a forward and reverse sweep (2 port error correction).

S-parameter Measurements If the S-parameters change versus drive direction then by definition the resulting computation of the S- parameters becomes invalid xt () yt () fwd b 1 rev b 1 fwd b 2 rev b 2 S 11 S 12 S 21 S 22 b1 S11 S12 a1 b = S S a 2 21 22 2 = S 11 S 12 S 21 S 22 = b 1 fwd b 2 fwd rev b 1 rev b 2 fwd a 1 fwd a 2 fwd a 1 fwd a 2 rev a 1 rev a 2 rev a 1 rev a 2 1 This is often why customers are asking for Hot S22 because the match is changing versus input drive power and frequency (Nonlinear phenomena). This still does not provide the complete picture. X-parameters are the solution.

X-parameters X-parameters are the mathematically correct extension of S- parameters to large-signal conditions. Measurement and simulation based, device independent, identifiable from a simple set of automated NVNA measurements or directly from ADS circuit-level designs Fully nonlinear (Magnitude and phase of distortion) Cascadable (correct behavior in even highly mismatched environment) Extremely accurate for high-frequency, distributed nonlinear devices Measure X-parameters -or- Generate X-parameters from circuit-level designs X-parameter Component Simulate using X-parameters ADS, SystemVue & Genesys Design using X-parameters

X-parameters A 1 A 2 B = F ( DC, A, A,..., A, A,...) 1k 1k 11 12 21 22 B = F ( DC, A, A,..., A, A,...) 2k 2k 11 12 21 22 B B 1 2 Harmonic (or carrier) Index Port Index Unifies S-parameters Load-Pull, Time-domain load-pull The X-parameters provide a mathematically correct mapping of the A and B waves at ports, input powers, harmonics, DC bias, etc, etc.

Scattering Parameters S-Parameters Linear System Description b i = k S ik a k b = S a + S a 1 11 1 12 2 b = S a + S a 2 21 1 22 2 X-Parameters Linear and Nonlinear System Description X X X ( * A A P + a ) b = ( A ) P + ( ) P a + ( ) ( F) j ( S) j l ( T) j l ij ij 11 ij, kl 11 kl ij, kl 11 kl kl, (1,1) A 11 = Large signal drive to the amplifier input port (port #1) at the fundamental frequency (#1) Definitions i = output port index j = output frequency index k = input port index l = input frequency index For example: T X 21,21 Means: output port = 2 output frequency = 1 (fundamental) input port = 2 input frequency = 1 (fundamental)

NVNA and X-parameters Measure X-parameters with Agilent s NVNA Design and simulate with measured or generated X- parameters in ADS Design with measured X-parameters in ADS

Presentation Outline Amplifier Design Considerations using X-parameters Industry Challenges Introduction to X-parameters End-to-End Power Amplifier Design with X-parameters Load-Dependent X-parameters High Power X-parameter Measurement Configuration Final PA Results

Load-Dependent X-parameters Some components (un-matched transistors) may require input and output tuners because their match is far from 50 ohm. This may require an X-parameter model that also includes dependence an arbitrary fundamental frequency load impedance supplied to the output of the component. This is accomplished by adding an impedance tuner. Depending on the component, and the class of operation, a multi-harmonic tuner may not be required. The source tuner can be fixed at a single impedance that is close to the conjugate match point and a power sweep performed to vary the available power to the component (i.e. No source pull required). X-Parameter definition with port 2 gamma dependence b ij = X ij ( F ) (DC, A 11,Γ 2 )P j + X (S ) (DC, A,Γ )P j l a + X (T ) (DC, A,Γ )P j+l a * ( ij,kl ) 11 2 kl ij,kl 11 2 kl k,l (1,1)

Load-Dependent X-parameter Measurements A 11 is swept through a power sweep Γ 2 is swept through a set of fundamental frequency impedances supplied by tuner. All other harmonic impedances are uncontrolled Γ 1, fixed Γ 2

Power Amplifier Design Goal: Design a power amplifier using X-parameter measurements of a transistor. Desired Design Goals: Frequency = 1.2 GHz Output power > 45 dbm PAE > 60% Class AB X-parameter model measured of a Cree CGH40045F GaN transistor using NVNA. Linearity and other performance parameters were not part of this first phase design.

Power Amplifier Design Total PA design completed in ADS: 1. Simulated impedance contours of output power and PAE at fundamental and harmonic frequencies at input (gate) and output(drain) ports. 2. Simulated impedance contours used to determine appropriate termination impedances at fundamental and harmonic frequencies at input (gate) and output(drain) ports to maximize PAE and output power 3. PCB designed with appropriate matching from (2) 4. Final PA assembled and compared against simulation

NVNA X-parameter System Power Budget (120 W) Port 1-18 dbm to +2 dbm (DT) -18 dbm (ET) Couplers - minicircuits ZGDC10-362HP+ CF ~ -10 db Pmax >+ 53 dbm Maury Microwave Tuner +10 dbm to +32 dbm AR 5S1G4 G = 30 db at lowest gain Pmax = +37 dbm (5 Watts) Frequency =.8 4.2 GHz 2x20 db (< 1 W) R1 A 10 db (< 1 W) 30 db Bias +12 dbm (ET) at +32 dbm input NVNA AR 60S1G4 G = max 38 db min 33 db at 0% Pmax = +48 dbm (60 Watts) Frequency =.8 4.2 GHz -20 dbm -20 dbm -20 dbm -20 dbm Cree CGH40045F GaN HEMT F = 1.2 GHz Gain ~ 12-16 db Pout ~ 46 dbm DUT R3 C 20 db +36 dbm +36 dbm 2x20 db (< 1 W) 40 db (10 W) +46 dbm (max) +26 dbm (ET) External HW Port 2 10 db (100 Watt) 1 dbm (ET) at G=35 db Couplers - minicircuits ZGDC10-362HP+ CF ~ -10 db Pmax >+ 53 dbm Bias Tuner Maury Microwave

High Power X-parameter Measurement System

Model Verification Verification of the measured X-parameter model: 1. Make measurements of component (transistor). These are not X-parameter measurements but instead measurements of other parameters like A and B waves, output power, PAE or another parameter of your choice. 2. Record the input power, impedances (input and output ports, fundamental and harmonic) and bias used during the measurements in (1). 3. Place these impedances in the simulator as terminations of the X-parameter model and sweep power over that used in measurement range.

X-parameter Model Verification - Circuit Template

X-parameter Model Verification - Results 46 Measured Verification of X-parameter Model 2.0 Delivered Power (dbm) 44 42 40 Simulated Drain Current (Amps) 38 20 22 24 26 28 30 Incident Power (dbm) 0.7

Simulated Fundamental Pout and PAE Contours Maximum Power and contour levels, dbm: 46.05 Output Power > 45 dbm Power_contours PAE_contours E P 46.00 45.50 45.00 44.50 44.00 43.50 43.00 42.50 Maximum PAE and contour levels, %: 70.812 PAE > 70 % E level=70.711610, number=65 P level=46.029871, number=52 68.000 64.000 60.000 56.000 52.000 48.000 44.000 40.000

Simulated Parameters ADS Amplifier DesignGuide 50 Output Spectrum, dbm 19 m1 Transducer Power Gain, db Gain Compression between markers, db 80 PAE, % 1.8 High Supply Current 0-50 -100-150 0.000 500.M 1.00G 1.50G 2.00G 2.50G 3.00G 3.50G 4.00G 18 17 16 15 38 39 40 41 42 43 44 45 46 Fund. Output Power, dbm m2 3.220 Output Power at Marker m2, dbm 45.70 70 60 50 40 30 38 39 40 41 42 43 44 45 46 Fund. Output Power, dbm 1.6 1.4 1.2 1.0 0.8 0.6 38 39 40 41 42 43 44 45 46 Fund. Output Power, dbm 20.0 21.0 22.0 RF Power Selector m3 27.0 26.0 25.0 24.0 23.0 RFpower[0,::] 28.0 29.0 30.0 Fundamental Frequency 1.200 GHz Available Source Power dbm 20.000 21.000 22.000 23.000 24.000 25.000 26.000 27.000 28.000 29.000 30.000 Fundamental Output Power dbm 38.916 39.826 40.717 41.596 42.460 43.318 44.122 44.775 45.180 45.472 45.696 Transducer Power Gain 18.916 18.826 18.717 18.596 18.460 18.318 18.122 17.775 17.180 16.472 15.696 Power- Added Efficiency, % 37.953 42.019 46.465 51.376 56.108 61.767 67.226 71.322 72.659 73.308 74.317 DC Power Consumpt. Watts 20.281 22.583 25.069 27.747 30.984 34.283 37.877 41.453 44.579 47.124 48.783 High Supply Current 0.724 0.807 0.895 0.991 1.107 1.225 1.353 1.481 1.592 1.683 1.743 Thermal Dissipation Watts 12.547 13.047 13.361 13.415 13.500 12.977 12.242 11.653 11.865 12.142 11.988 40 Input and Output Voltage Waveforms 50 Fundamental and Third Harm., dbm 30 45 20 40 ts(vload), V ts(vinput), V 10 0-10 Spectrum[2] Spectrum[3] Spectrum[1] 35 30 25-20 20-30 15-40 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 time, nsec 10 20.0 20.5 21.0 21.5 22.0 22.5 23.0 23.5 24.0 24.5 25.0 25.5 26.0 26.5 27.0 27.5 28.0 28.5 29.0 29.5 30.0 RFpower

Final Power Amplifier PAE could be improved based on a better physical termination impedance at the 2 nd harmonic at the input. However, there is a tradeoff between the bandwidth of the impedance and PAE. Courtesy University of Waterloo

X-parameter technology is expanding rapidly X-parameter breakthroughs: Load-dependent X-parameters 50 GHz Agilent NVNA High-Power X-parameters X-parameter generator in ADS Simulation of XnP component in ADS, Genesys & SystemVue Two-tone measured X-parameters Three-port measured X-parameters (mixers/converters) Memory: Dynamic X-parameters Education, training, app. Notes

Thank You to Supporting Partners AR RF/Microwave Instrumentation Cree Maury Microwave Mini-Circuits University of Waterloo Thank You

Appendix

X-parameter Measurements Rules-of-Thumb Leave enough pre-amplifier linear gain for extraction and drive tones If the pre-amplifier is saturated with the drive signal then adding the extraction signal will degrade X-parameters. Generally seen during a power sweep where there is divergence between simulated and measured results at the higher end of the power sweep (watch receiver compression though). Simulation termination impedance When comparing simulated a s and b s from X-parameters against measured a s and b s it is critical that the terminations in the simulation match that used during measurement. Ensure proper calibration and de-embedding techniques where applicable. Calibration procedure using 8 term error model, tuners and preamplifiers Often pre-amplifiers are removed behind the couplers during calibration and then placed back in-line after the calibration procedure is complete. This may effect the tuner characterization and therefore the source and load impedances behind the tuners should be determined from the NVNA and accounted for to ensure proper applied impedance to the component by the tuners.

For More Information X-parameters www.agilent.com/find/x-parameters Nonlinear Vector Network Analyzer www.agilent.com/find/nvna ADS MMIC design seminar (click on X-parameters link) www.agilent.com/find/eesof-mmic-seminar X-Parameters Aid MMIC Design http://www.mwrf.com/articles/index.cfm?ad=1&article ID=22811 X-Parameter YouTube Videos http://www.youtube.com/user/agilenteesof Trademark Usage, Open Documentation & Partnerships http://www.agilent.com/find/x-parameters-info

Selected References and Links 1. D. E. Root, J. Horn, L. Betts, C. Gillease, J. Verspecht, X-parameters: The new paradigm for measurement, modeling, and design of nonlinear RF and microwave components, Microwave Engineering Europe, December 2008 pp 16-21. http://www.nxtbook.com/nxtbooks/cmp/mwee1208/#/16 2. D. E. Root, X-parameters: Commercial implementations of the latest technology enable mainstream applications Microwave Journal, Sept. 2009, http://www.mwjournal.com/search/expertadvice.asp?hh_id=res_200&searchword=root 3. J. Verspecht and D. E. Root, Poly-Harmonic Distortion Modeling, in IEEE Microwave Theory and Techniques Microwave Magazine, June, 2006. 4. D. E. Root, J. Verspecht, D. Sharrit, J. Wood, and A. Cognata, Broad-Band, Poly-Harmonic Distortion (PHD) Behavioral Models from Fast Automated Simulations and Large-Signal Vectorial Network Measurements, IEEE Transactions on Microwave Theory and Techniques Vol. 53. No. 11, November, 2005 pp. 3656-3664 5. J. Verspecht, J. Horn, L. Betts, D. Gunyan, R. Pollard, C. Gillease, D. E. Root, Extension of X-parameters to include long-term dynamic memory effects, IEEE MTT-S International Microwave Symposium Digest, 2009. pp 741-744, June, 2009 6. J. Verspecht, J. Horn, D. E. Root A Simplified Extension of X-parameters to Describe Memory Effects for Wideband Modulated Signals, Proceedings of the 75 th IEEE MTT-S ARFTG Conference, May, 2010 7. J. Xu, J. Horn, M. Iwamoto, D. E. Root, Large-signal FET Model with Multiple Time Scale Dynamics from Nonlinear Vector Network Analyzer Data, IEEE MTT-S International Microwave Symposium Digest, May, 2010. 8. J. Horn, S. Woodington, R. Saini, J. Benedikt, P. J. Tasker, and D. E. Root; Harmonic Load-Tuning Predictions from X- parameters, IEEE PA Symposium, San Diego, Sept. 2009 9. D. Gunyan, J. Horn, J. Xu, and D. E. Root, Nonlinear Validation of Arbitrary Load X-parameter and Measurement-Based Device Models, IEEE MTT-S ARFTG Conference, Boston, MA, June 2009 10.G. Simpson, J. Horn, D. Gunyan, and D. E. Root, Load-Pull + NVNA = Enhanced X-Parameters for PA Designs with High Mismatch and Technology-Independent Large-Signal Device Models, IEEE ARFTG Conference, Portland, OR December 2008. 11.J. Horn, J. Verspecht, D. Gunyan, L. Betts, D. E. Root, and J. Eriksson, X-Parameter Measurement and Simulation of a GSM Handset Amplifier, 2008 European Microwave Conference Digest Amsterdam, October, 2008 12.J. Verspecht, D. Gunyan, J. Horn, J. Xu, A. Cognata, and D.E. Root, Multi-tone, Multi-Port, and Dynamic Memory Enhancements to PHD Nonlinear Behavioral Models from Large-Signal Measurements and Simulations, 2007 IEEE MTT-S Int. Microwave Symposium Digest, Honolulu, HI, USA, June 2007. 13.D. E. Root, J. Xu, J. Horn, M. Iwamoto, and G. Simpson, Device Modeling with NVNAs and X-parameters, 2010 IEEE MTT-S INMMiC Conference, Gοtenborg, Sweden, April 26, 2010 14.J. Horn, G. Simpson, D. E. Root, GaN Device Modeling with X-parameters, Accepted for publication 2010IEEE CSICS, Oct. 2010