Workshop WSF Power Detection and Control for Handset Power Amplifiers David Ripley Skyworks Solutions Incorporated 5110 North River Blvd Cedar Rapids, Iowa, 52402 1 Outline • Bipolar Power Amplifier Biasing Techniques • Control Characteristics of Bias Techniques • Power Control Methods • Indirect Power Control Characteristics • Direct Power Control Characteristics • Transient Considerations 2 Complex Product Requirements Switching Spectrum Power Vs. Time Noise Spectrum -75 +20 MHz Amplitude (dBm) (dBm) -80 -85 -90 -95 -100 -105 -110 0 5 10 15 20 25 30 35 Pout (dBm) (dBm) Dynamic Range 3 2.5 Total Radiated Power Load Invariance 30 29 1.5 1 1 28 0.8 0.6 Delivered Pout (dBm) Delta POUT_MEAS (dBm) 2 0.5 0 -0.5 -1 -1.5 0.4 0.2 27 0 -0.2 -0.4 26 -0.6 -0.8 -1 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 25 24 -2 -2.5 -3 -5 0 5 10 15 20 POUT_MEAS (dBm) (nom) 25 30 Environmental Stability 23 22 0 45 90 135 180 225 270 315 VSWR Angle (deg) Increasing customer focus on antenna performance continues to drive complexity and increased margins into requirements 3 360 Bipolar Amplifier Bias Methods Vc • Saturated bias methods control Vb PA output power by limiting current or Ib voltage Gm RF • Linear bias methods maintain a fixed gain allowing for accurate control of output power through RF input drive Fundamentally – PA is a common emitter gain stage 4 Saturated Amplifier Bias Techniques Base-current control Vbias Amplifier Ib Base Bias Ibias OR Ib Amplifier Base Bias • Amplifier output power is adjusted by means of limiting DC base current • Implemented with voltage control through moderate impedance or high impedance current source 5 Output Power (dBm) Dynamic Range and Accuracy 40 35 30 25 20 15 10 5 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50 High impedance Base bias Low Drive, Nom High Drive, Nom Low Drive, Cold High Drive, Cold Low Drive, Hot High Drive, Hot 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Vbias (V) 2 Significant Variation with RF drive and Temperature 6 Output Power (dBm) Dynamic Range and Accuracy Current Source Base bias 40 35 30 25 20 15 10 5 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50 Low Drive, Nom High Drive, Nom Low Drive, Cold High Drive, Cold Low Drive, Hot High Drive, Hot 0 0.005 0.01 0.015 0.02 0.025 0.03 Base Current (A) Significant Variation with Temperature 7 Saturated Amplifier Bias Techniques Collector-Voltage control Venvelope Amplifier Vc Collector Bias • Amplifier output power is adjusted by means of voltage saturation by limiting DC collector voltage 8 Output Power (dBm) Dynamic Range and Accuracy Collector Voltage Control 40 35 30 25 20 15 10 5 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50 Low Drive, Cold High Drive, Cold Low Drive, Nom High Drive, Nom Low Drive, Hot High Drive, Hot 0 0.5 1 1.5 2 2.5 3 3.5 Vcoll (V) Nearly Constant Control with Temperature and Drive 9 Linear Amplifier Bias Technique Fixed-Gain Mirror Biasing Iref Vb Amplifier Base Bias • Amplifier output power is adjusted by means of RF input power • Amplifier bias point is held constant 10 Dynamic Range and Accuracy Fixed Gain Mirror Bias 40 35 Gain (dB) Gain (dB) 30 25 20 15 10 5 0 0 0.2 0.4 0.6 0.8 1 Icntrl (mA) 1.2 1.4 1.6 1.8 2 Bias Reference (mA) Moderate Gain Variation with Temperature Limited dynamic Range 11 Power Added Efficiency Efficiency Comparison 15 Base Current Bias Collector Voltage Bias 20 25 30 35 Output Power (dBm) Typically 4 – 5% trade-off 12 Bias Noise Consideration • Base current bias is the most challenging for noise – PA is not voltage saturated- High small signal gain – Significant conversion gain of bias noise to the carrier – Gain of the following stages drives challenging noise level requirements for bias of driver stages • Voltage saturation control offers the best noise performance – PA is constantly in deep voltage saturation – Noise applied to the final stage is the dominant source • Noise introduced on driver is AM and is suppressed by final stage compression • No gain follows the final stage Typically 1 – 3 dB trade-off 13 Indirect Vs. Direct Control Methods • Typical indirect power detection takes advantage of relationships between DC characteristics and RF output power – Simple circuitry, simple process technology – Limited visibility to antenna loading conditions – Potentially smaller size • Direct power detection monitors aspects of the RF waveform – High frequency circuitry, RF process technology – Increased complexity of dynamic range/isolation tradeoffs – Often requires coupler Indirect – Simple, Low cost, limited performance Direct – Increased complexity, expense, precision performance 14 Current Sense Indirect Power Control IDC Vbat Vbat Rc Vcemin Vceoffset Re Ground IDC * η => Vrfpk 15 Indirect Power Control Accuracy Current Sense Control Measured Power (dBm) 34 29 24 19 14 9 4 -1 -6 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 Desired Output Power (dBm) Precision over Wide Dynamic Range 16 Voltage Control Indirect power Control Vreg Vreg Rc Vcemin Vceoffset Re Ground Vreg - Vcemin => Vrfpk 17 Indirect Power Control Accuracy Collector Voltage Control Measured Power (dBm) 34 29 24 19 14 9 4 -1 -6 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 Desired Output power (dBm) Temperature compensation can improve low power range 18 RF Peak Detector Circuit Requires High Performance Process to support RF Input 19 3 Peak Detector Dynamic Range and Accuracy 2.5 2 1.5 Error (dB) 1 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 -5 0 5 10 15 20 25 30 35 Pout (dBm) Typically 30 dB dynamic range 20 Log Detector Block Diagram PTAT BIAS Band gap reference DET RFIN attenuator DET current summation DET 12 dB DET 12 dB VDET 1.67 DET 12 dB DET 12 dB /buffer offset compensation Current drain is higher due to multiple detection stages 21 Log Detector Dynamic Range and Accuracy 3 2.5 Delta POUT_MEAS (dBm) 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 -5 0 5 10 15 20 25 30 POUT_MEAS (dBm) (nom) 40 dB dynamic range easily obtained 22 Power Variation into VSWR 30 Indirect Methods 29 Delivered Pout (dBm) 28 8 mm 27 26 25 2.25 dB 24 23 22 0 45 90 135 180 225 270 315 360 8 mm Pout Delivered (dBm) VSWR Angle (deg) 26 25.9 25.8 25.7 25.6 25.5 25.4 25.3 25.2 25.1 25 24.9 24.8 24.7 24.6 24.5 24.4 24.3 24.2 24.1 24 Direct Methods 0.5 dB 0.5 dB 0 50 100 150 Coupler + complex termination 0.5 dB 200 250 300 350 Load Tuner S11 (ang) 23 Closed-Loop power control Direct Detection Requires Feedback 24 Transient Considerations • Lock time of the control loop is critical – Bandwidth must be constrained for stability and noise – For fixed bandwidth ΔTlock is inversely proportionate to ΔVin – Current Sense and Peak detect yield ~2 mV at pedestal – Log, RMS, Collector yield > 100 mV at pedestal • Saturation of the control loop must be prevented – Snap-down characteristic yields significant transient spectrum which violates system requirements Both Rising and Falling Edges Must be Considered 25 Transient Waveform Examples Lock Time Performance Saturation Performance 40 35 30 25 20 Normal response 15 (d B m ) 10 5 Saturated loop 0 -5 Pedestal -10 -15 -20 -25 -30 -35 -40 10u s 8us 10u 26 Conclusion • Several trade-offs must be considered when selecting the fundamental bias technique • Indirect power control can be simple and low cost but lack accuracy • Direct power control can provide precision control even in non-ideal load conditions • Feedback within the power control solution must be carefully designed with focus on transient performance characteristics 27 Bio David S. Ripley received his B.S. degree in electrical engineering from Iowa State University, Ames, in 1992 and the M.S. EE degree from National Technical University (NTU), Minneapolis, MN, in 2002. From 1992 to 1999, he worked in the Cellular Subscriber Division, Motorola, Libertyville, IL, where he was involved in the design and development of TDMA and AMPS handsets including RFIC design of receiver and synthesizer functions. Since 1999, he has been with Skyworks Solutions, Inc. (previously Conexant Systems, Inc.), Cedar Rapids, IA, where he has been involved with the design of multiband HBT power amplifiers for the GSM and CDMA cellular handsets. He holds fifteen patents. 28