"K7ITM" wrote in message
...
"I'm not sure how they work, but suppose that something similar to
what's used for an audio phase shift network for use generating SSB
could be used."

I suspect it's actually something a bit simpler in that the LO is supposed to
be one frequency (perhaps with harmonics) and not something that moves around
quickly (hence the idea that you can just use a feedback loop and track it),
whereas with audio you're dealing with a whole bunch of frequencies at once
(in a sense) and have to phase-shift them all 90 degree simultaneously.

I was planning on designing (or at least implmenting -- all the real design
was by those filter guys decades back) an audio phase shift network, but since
you need an IQ (-like) modulator anyway, and since I already needed a
microcontroller for this design, I'm thinking these days it's just as easy to
perform a Hilbert transform inside the microcontroller and spit out the audio
and it's phase-shifted version on two DAC pins and call it good.

---Joel