Efficiency of Doherty RF Power-Amplifier Systems

TLDR

A Doherty RF power‑amplifier system combines multiple linear PAs through an impedance‑inverting coupler, operating with the first PA linear at low output and the second PA linear when the first is saturated at higher output levels. The authors analytically derive instantaneous and average efficiency curves for ideal class‑B RF PAs and determine maximum‑efficiency transition points for various amplitude‑modulated signals, enabling easy scaling to real‑world PAs. The two‑stage Doherty configuration raises average efficiencies from 28 % and 8.9 % (class‑B) to 60 % and 48 % for 10‑ and 20‑dB peak‑to‑average ratios, and adding a third stage further improves them to 70 % and 66 %.

Abstract

A Doherty system combines the outputs of two or more linear RF power amplifiers (PAs) through an impedance-inverting coupler such as a quarter-wave transmission line. At low output levels, the first PA operates linearly, reaching saturation (and maximum efficiency) at some transition voltage below the system peak-output voltage. At higher output levels, the first PA remains saturated and the second PA operates linearly. The instantaneous efficiency and power characteristics of a Doherty system are derived using ideal class-B RF PAs so that the results can easily be scaled for use with real-world PAs. The average efficiency and maximum-efficiency transition points are then determined for a variety of amplitude-modulated signals. The Doherty amplifier can be considerably more efficient than a conventional class-B linear PA. For example, the 28-and 8.9-percent average efficiencies of a class-B PA with Rayleigh-distri buted envelopes with 10-and 20-dB peak-to-average ratios are improved to 60 and 48 percent, respectively, by a two-stage Doherty system. The addition of a third stage further improves the efficiencies to 70 and 66 percent, respectively.