Re: VLF: EbNaut tonite, 'SMS' via VLF

[DK7FC <selberdenken@posteo.de>]

Hi Markus, Paul,

It was a long transmission with relatively short symbols. Apart from the 
accuracy that is required for the timing there is also a higher 
requirement for the accuracy of the frequency. In my attempts of getting 
a decode from the recent 970 Hz test i saw that there is a frequency 
offset of -1/(transmission time) or probably rather -1/(FFT length). I'm 
still not absolutely sure but maybe this also plays a role here?
A frequency offset was also observed by RN3AUS. Maybe this is not 
propagation but an effect that happens in the SL instance?
Something we can find out in coming transmissions.
I could run something like 30...40 characters, which is close to the 
limits for SQ5BPF and DL0AO...
73, Stefan


Am 29.05.2017 19:58, schrieb Paul Nicholson:
> Markus wrote:
>
> > I managed to produce a seamless wav with 0.514 Hz samplerate:
> > 
> http://df6nm.bplaced.net/VLF/fec_tests/dl0ao_170524-25_join31uHz_wav.zip
> Still no decode.  The correlation peaks at 20.44 dB in 12.6 uHz with
> offsets +2058 seconds and +0.1 Hz so it looks like the settings are
> correct.    I'm using the sample rate 0.5143924/sec from the
> WAV inf1 header.
>
> It's about 9 or 10dB short of decoding.
>
> The previous attempt gave 22.09 dB in 12.6 uHz so it's lost a
> couple of dB somewhere.
>
> --
> Paul Nicholson
> --