Successful message decode of an amateur signal at over 3 wavelengths at ULF

Paul Nicholson has successfully decoded a message on 2.97kHz from DK7FC using clever analysis.

Hi Paul,

Am 25.01.2017 19:30, schrieb Paul Nicholson:

> f = 2970.000000 Hz
> Start time: 25.Jan.2017   07:00:00 UTC
> Symbol period: 30 s
> Characters: 5
> CRC bits: 16
> Coding 16K25A

Decoded at Bielefeld via DL4YHF (303.8km).  Copied 'DK7FC' with
Eb/N0 +4.5 dB and constant reference phase.

S/N 19.1 dB in 29.8 uHz, -60.1 dB in 2.5kHz.

Just excellent! Thanks for working out the decode!
My call passed over 3 wavelengths on ULF, that's a dream! I am sure we will manage even more.

Very interesting to see the high Eb/N0 relative to yesterday. From the calculator it should have been weaker. And the QRN didn't look much lower than yesterday, at least from here.
Some there's a relatively high uncertainty down there, and the risk of getting no decode when coming to close to the limit. The stacking is an advantage again.

I'm still working on the recordings from the weekend.  Gained a further
3dB by using very strong hum filtering, a threshold only 1.5 times the
mean amplitude and narrow notches.  Enough to suppress not just the
harmonics but also the sidebands of the harmonics.   Still no trace
here in Todmorden - yet.  Still some permutations of polarisation
and azimuth to try.

Am 23.01.2017 18:52, schrieb Paul Nicholson:

Still no trace of the signal here.  But tweaking the blanker
isn't going to go from 'no trace' to 'significant detection'.

3 dB is very significant! Congrats to that. It will also help to get better results from Renatos data. But as far as i remember he is offering the stream with the blanker already applied.
We could ask him to disable it temporarily, if that helps. If he knows what we're trying, he will do it i expect :-)

We are having fun with this!  Constantly pushing at the limits.

Oh yes, and also a lot to learn about the optimisation work for the different bands.

73, Stefan