Measurement-Driven Automated Design of High-Efficiency Power Amplifiers

In modern power amplifier (PA) design, model-based approaches combining EDA tools are commonly employed. To avoid numerous iterations due to inaccuracies of transistor models and achieve first-design-pass goal, this paper proposes a measurement-driven automated PA design approach utilizing load-pull measurement data and machine learning methods for the design of harmonic controlled PA. To validate the effectiveness of the proposed approach, a PA operating at 2.4 GHz was designed and measured with a maximum power-added efficiency (PAE) of 73.4% and a maximum output power of 40 dBm.