"Richard Harrison" wrote
The FM amplifier does not need linearity. Amplitude distortion is
irrevelant. Severe clipping to remove amplitude variations
is common practice.

Not so. You confuse receivers with transmitters. Limiting is supplied by
the IF strips of FM receivers to reduce/remove AM components on the incoming
wave, but FM broadcast transmitters are operated well below any
limiting/clipping level, and that is probably true of ham FM txs also.

Synchronous and asynchcronous AM are low in broadcast FM tx RF stages (the
FCC spec is -50dBc), but not because the FM amplifiers are "clipping."
Broadcast FM txs easily can be adjusted over an output power range of ~25%
to 105% or more simply by adjusting drive to the PA (keeping constant PA
volts). This technique often is used for output power control/VSWR
foldback, actually.

Clipping generates harmonics and FCC rules limit harmonic
transmission in all services. .. I would be surprised if some
final filter were not used to guarantee compliance with the rules.

You are confused again. I wrote that no "tank circuit or in-band filter(s)"
were necessary to achieve the high efficiency I described. Your post I was
responding to states that a "tank circuit and other filters" are necessary
for high efficiency -- that is not true.

Harmonics are present at the PA output of an FM transmitter, but "clipping"
is not the process whereby they are generated, as I state above. They are
reduced to legal values using a lowpass/harmonic filter. The FCC
attenuation spec for harmonics and spurs more than 600kHz from Fc is 80dB
below the unmodulated carrier.

The lowpass/harmonic filter does not improve efficiency--it has a small
amount of insertion loss in the FM band.

RF