On the Efficiency and AM/AM Flatness of Inverse Class-F Power Amplifiers

Inverse Class-F (Class F−1) power amplifiers (PAs) are widely known to deliver excellent performance at saturation. Due to this, Class F−1 PAs are not recommended for applications where PA operation is required to be backed-off from saturation. This paper explores optimizing Class F−1 PA performance and maintaining the performance over a high range of gain compression levels. A simple model of the Class F−1 amplifier is proposed to approximate its harmonic-dependent behavior as seen at the internal reference plane. A load-pull based experimental approach is then presented, according to which input second-harmonic terminations are tuned, while maintaining quasi-ideal inverse Class F output matching conditions. The device under test used in the experiment is a 0.7 mm Gallium Nitride (GaN) die from NXP at a frequency of 2.6 GHz. Measured results provide significant insight into the impact of input harmonic manipulation on the amplitude-to-amplitude modulation (AM/AM) profile. A design space for second harmonic input impedance was found to provide relatively flatter AM/AM response of the Class F−1 PA yielding significantly improved RF performance at the 3-dB compression point (P3dB) compared to the classical Class F−1 PA with no second harmonic manipulation at the input.