TABLE OF CONTENTS AM-PM effects seem to have something of a poor-relation status in the RF power amplifier world. Amplitude distortion is well understood, and has an established litany of rules of thumb and design guidelines. But phase distortion effects seem more inscrutable, despite their obvious potentially disruptive effects in phase modulated systems. One factor is that until recently, AM-PM was a tricky measurement to perform on an RF power amplifier. Older vector network analyzers had to be calibrated at very low fixed power levels, and a change in the level could produce a phase error greater than the amplitude dependent phase change of the device under test. Fortunately, modern instruments are much less susceptible to these effects, and the phase of the transfer characteristic can be confidently swept over the same power range as the P in -P out characteristic. But there is still an almost complete absence of calibration points, such as intercept or 1 dB compression points, which provide a rough roadmap for amplitude distortion effects.
There is also a good deal of mystery about exactly what causes AM-PM distortion in the first place. Clipping on supply rails can, as we have shown, do quite a reasonable job of explaining gain compression, but it is not clear where phase distortion comes from, even when looking at well-clipped voltage and current waveforms. A good starting point, therefore, is to look at some measured AM-PM data. Figure 9.13 shows data taken for the actual amplifier...
