In a message dated 3/26/01 12:06:05 PM Eastern Standard Time,
[email protected] writes:
<< But unfortunately, any clipping or AGC will distort the amplitude
envelope, which in turn will re-introduce unwanted sidebands. >>
This may be a matter of semantics, but clipping won't actually "re-introduce"
sidebands. While AGC would indeed distort the amplitude envelope and worsen
sidebands, hard RF clipping eliminates envelope variations entirely. The
consequence of clipping is that we no longer have the envelope to help us
further suppress the phase-modulation sidebands. The sidebands themselves
remain just as they would for any continuous-shift PSK signal--less energetic
than for brute-force instantaneous PSK, but broader than if we retained the
amplitude variations that accompany appropriate filtering.
The ideal case would be to generate a waveform based on root-raised-cosine
filtering in the exciter, use envelope-elimination-and-recovery to amplify
the signal at maximum efficiency, and use a corresponding filter at the
receive end.
However, I believe the object of Rik's proposal was to decrease bandwidth
while retaining as much simplicity in the system as possible. If one can
abide the slower phase transitions of the filtered signal, and the modest
increase in bandwidth from not having dragged out the transitions even
further, the technique should work at moderate power levels. It's a
collection of trade-offs, as is anything in engineering.
73,
John
|
