Tom Bruhns wrote:
"Joel Kolstad" wrote in message
...
For standard broadcast, don't they always put L+R on I and L-R on Q,
so standard receivers get L+R?

They may well -- I'm not sure which 'standard' we're talking about here
(there seem to be several out there...). :-) But note that with L+R on I
and L-R on Q, you only get L+R if you manage to synchronize with the phase
(the original problem)! If you use a quadrature detector and extract the
magnitude, you of course get sqrt(L^2+R^2) -- good, but not exactly L+R
either. The Motorola C-QUAM AM stereo system forces the magnitude of the
I-Q phasor to be L+R, which makes it compatible with both envelope detectors
and quadrature detectors. (I probably sound like a Motorola sales guy these
day... I just think it's clever and the end goal here is to build a direct
conversion receiver to decode it...)

But I assume that in any transmission
standard, there is something transmitted that lets you recover the
right phase at the receiver.

For AM, it would appear that the carrier itself is what people use for that
purpose.

If it very clear to me now why TV transmission need to send the colorburst
sequence these days -- otherwise they'd have no way to synchronize the
chroma decoder, which is taking in a DSB-SC modulated in both I and Q.

One easy way is with a "Tayloe mixer" -- you should be able to find
info on that on the web, but it's basically just a commutating switch
that switches the signal (through its source resistance) among four
different capacitors.

I was just thinking of doing something like this -- from a square wave, get
a 4 bit shift register (that always has one output active) and feed the
outputs to a 4066 such that I sample the 'I' channel during, say, clocks #1
and #4 and the 'Q' channel during clocks #1 and #2. Add some filtering
and -- presto chango! -- we've got I and Q.

Thanks for all the help -- this just might end up working after all.

---Joel Kolstad