Re: LF: Re: EbNaut test transmission

[Markus Vester <markusvester@aol.com>]

Well done Andy!

 

This decode was from a message before the NTP adjust. Your transmit frequency looks stable and accurate.

 

I have temporarily moved the opds SpecLab instance to your frequency, and inverse Fourier transformed a saved opds spectrum data file.

 

Domenico and I have successfully tried speeds up to 0.1 second symbols, with no additional tweaking after we'd once got the time offset right.

 

Best 73,

Markus (DF6NM)

-----Ursprüngliche Mitteilung----- 
Von: Andy Talbot <andy.g4jnt@gmail.com>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>; rsgb_lf_group <rsgb_lf_group@yahoogroups.co.uk>
Verschickt: Di, 17 Nov 2015 5:26 pm
Betreff: LF: Re: EbNaut test transmission

Its still going.

At 1611z I ran rsNTP and it made a PC clock adjustment of 0.4 seconds, which with the 0.5s symbols used for this transmission will glitch slightly .

Pleased to note that a local recording I made using the  new LF Receiver hardware and my own software to generate a .WAV file compatible with Spectrum Lab's did generate a valid decode. Recorded at fs = 31.25Hz  That was before the PC time had been corrected and I had to do a fair bit of shuffling with the Start Offset value to get a decode.   Values from 0.45s to 0.85s worked - much of that presumably due to the PC time offset in the transmission.   But once a value in that range had been entered, zero decode errors - which is what one would expect of a very strong local copy !

It does appear that using EbNaut with rates much faster that 2S/s could end up very fraught with timing offsets.  Fortunately, weak signalling with longer symbols of several seconds will have less of a problem with timing offsets.

Andy  G4JNT

sym_11171616.png
Description: PNG image

11171616-log.txt
Description: Text document