Here I got 'NA' from the 02:00 message at Eb/N0 somewhere
below -1.9 dB, BER 47.7%, uniform phase.

Very weak because the rx is off while the battery charges
and it is receiving only on the feeder or whatever leaks
through the dead pre-amp.

Nil from the other messages.

Edgar wrote:

> Have set-up to record using SL with 1.43051 mHz bin width,
> 30 second waterfall interval.

The signal is invisible on a spectrogram so you wont see anything.
Also invisible on a spectrum unless it is a strong signal.

Might be worth sending the recording 10:00:00 to 11:42:24 UT into
the decoder for a try.  Use a very large list size and check the log
for 'NA' in amongst all the false decodes. If you find NA, then
you get the carrier S/N in the log.

You should try a 3 hour carrier and record that.  Then plot
spectra (not spectrogram) with bin widths between 100uHz
and 300uHz, sampling from a range of time intervals.  Hopefully
there's a window in that 3 hour slot which will show a carrier
spike and from its S/N we can work out if a BPSK message will
decode.

--
Paul Nicholson
--