Hi Wolf and all,

The  SNR of the LORAN lines in the receiver's passband makes it difficult
to track them, and as I understand it I need two of them separated by a

couple of Hertz to use them to calibrate both sample rate and LO drift.

There are a couple of hundred such lines in SSB bandwidth; the total power
is a robust signal which is easy to track.  You need not process separate
"lines".  Each LORAN pulse group produces a damped audio waveform, whose
shape depends on the instantaneous LO phase when the pulse hits.  If you
frequency shift the audio (in software) so that the effective LO is a
multiple of 2*GRI, then the pulses all line up.  A software PLL maintains
that alignment, which also stabilizes the signal being measured, except
for sound card drift.  The software also needs to measure the apparent
GRI, and adjust the demodulation of the signal under test accordingly.

Sorry that the TV thing won't work.  Probably the result of meeting
the EMI emissions regulations :)

For the GPS scheme, you do need to know the effective LO frequency within,
say, +/- 0.4 Hz.  There are a number of ways of (manually) measuring it
(once).  The "calibration" doesn't have to be right - you can set the
receiver dial to one value, and tell the software the true value.
You are also correct in that fairly good stability is needed, perhaps
within a couple of tenths of a Hertz per minute.  However, anyone serious
about QRSS already has better stability than that.

73,

Stewart KK7KA