Generic load-pull-based design methodology for performance optimisation of Doherty amplifiers

Abstract

In this study, a systematic design methodology is proposed to optimise the operation of Doherty power amplifiers (PAs). The proposed approach makes use of two sets of load-pull data to enhance the performance of Doherty PAs at low- and high-power-drive levels. The first load-pull, which is performed on the device operating at saturation, permits one to maximise the performance at a high-power region. The second load-pull, which is performed at the power level associated with the turn-on of the peaking amplifier, aims to boost the performance at back-off. To assess its effectiveness, the proposed methodology is applied to design three Doherty PAs sought for power efficiency, linearity and gain, respectively. Around the Doherty turn-on point, these circuits achieved up to 9% efficiency improvement, up to 10 dB inter-modulation reduction and up to 2 dB gain improvement, respectively. For experimental validation, a gallium-nitride (GaN)-based Doherty PA prototype sought for efficiency was implemented. The fabricated Doherty PA demonstrated a power-added efficiency (PAE) higher than 40% over an output power back-off (OPBO) range of 8 dB, with two peak PAE points of 52 and 62% located at 6.8 dB OPBO and at saturation, respectively.