28 Feb
2010
28 Feb
'10
8:17 p.m.
On Sun, Feb 28, 2010 at 11:46:21PM +0000, Gordon JC Pearce wrote:
The incoming RF is downconverted to the audio range,
but not
demodulated. Effectively you've mixed a chunk of RF (7MHz band, in the
case of my sample file) down to an audio range but not yet demodulated.
At this point you may have a signal at 7066kHz which has been downmixed
to 10kHz, but that's outside the receiver passband - you can't hear it.
By mixing a further 10kHz signal into the audio you bring that down to
0kHz, and if it's an LSB speech signal it now occupies a range of
frequencies up to about 3kHz with an "image" at 20kHz. Now we've got it
in the speech range and we can demodulate it by summing the imaginary
and real components so that the image cancels out.
Think of the SDR hardware as being a pair of direct-conversion receivers
with the local oscillators 90 degrees out of phase so the output is a
vector rather than a scalar.
It's similar to Steve Harris' Bode Shifter plugin,
except that this takes a single (real) signal and
converts it to complex internally using a Hilbert
transform.
Ciao,
--
FA
O tu, che porte, correndo si ?
E guerra e morte !