Frequency: 8270.100 Hz,

start time 8:30:02, 

5 characters, 8K19A, 16 bit CRC, 10 second symbols,

total duration 5120 seconds,

end time 09:55:22.

Thanks Paul for trying!

 

Same here... Stefan sent me a 63 uHz spectrum data file from Heidelberg, which I unsuccessfully tried to decode. Unspreading the signal with the known message "df6nm" recovered the unmodulated carrier with 10.2 dB SNR in 1/5120 Hz, equivalent to Eb/N0=-4.3 dB for 28.5 net bits.

 

There seemed to be a significant degradation from Stefan's strong local test on 8270.005, which is close to but not exactly in the null of the 10 s symbol filter, and unfortunately also accompanied by weak noise sidebands. Notching out a 30 mHz wide section around the carrier improved Eb/N0 to -0.8 dB, which would have come close but still didn't produce a decode. 

 

I have uploaded IQ wav files and plots to
http://df6nm.bplaced.net/VLF/fec_tests/dk7fc_160207_0502.zip (130 kB).
The best unspread carrier was obtained using 3311 s start offset and +0.10022 Hz frequency offset.

 

Best 73,
Markus
 

-----Original Message-----
From: Paul Nicholson <[email protected]>
To: rsgb_lf_group <[email protected]>
Sent: Mon, Feb 8, 2016 8:36 pm
Subject: Re: VLF: EbNaut 8270.1 Sunday

> Just started another sequence with same parameters,
> running 04:00:02 to 5:25:22 UT.

I must admit defeat on this one Markus.

The signal is there but just a dB too weak to
decode. The following carrier was 0.07 fT which
is a decent signal but the noise was higher
than the other day.

Here I've been playing with ebsynth which is now
at version 0.5 and has some improvements. It uses
much less CPU and does the phase modulation using
a sigmoid shaped slew.

http://abelian.org/ebnaut/ebsynth.shtml

I built a multiplier chain, 53077.777778 * 3 * 3 and
have a nice 10V 477 kHz signal which is rock steady
and the modulation clicks don't seem to spread.

There was a gap in the rain on Sunday so I put up a
200 metre long wire.

--
Paul Nicholson
--