Re: LF: Phase shifts on 137.777 kHz

[Markus Vester <markusvester@aol.com>]

Hi Joe,

 

phase shift is working - two peaks at +- 10 mHz, as expected. Updating spectrogram:

https://dl.dropboxusercontent.com/u/26404526/EbNaut_spectrogram.png

 

Yes the start of the message and the bit timing need to be referenced to UT, and we have to provide this information to the decoder. You can make a choice on the symbol duration, based on available SNR, length of message etc.

 

Currently each of the two carriers is ~ 24 dB SNR in 1mHz here. A straight carrier would be 4 dB stronger, giving -2 dB SNR in 1 Hz. This means that you could theoretically achieve a 0 dB Eb/N0 decode here if you sent 0.63 info bits per second (6 characters per minute), using ~ 0.2 s symbols. But I would probably prefer to start slower, with something like 1 or 2 second symbols to have some margin for QSB and more distant receivers.

 

Best 73,

Markus

 

 

 

 

 

 

 

-----Ursprüngliche Mitteilung----- 
Von: jcraig <jcraig@mun.ca>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: Sa, 21 Nov 2015 8:47 pm
Betreff: LF: Phase shifts on 137.777 kHz

Dear Group, 

 

Hopefully I've finally got it right...  Inversions every 

50 seconds a dozen times then 100s of silence. 

 

Hopefully QRV on EbNaut in the near future.  I have a long 

list of 1s and 0's loaded and ready to send. Doing:  

./ebnaut -dt -N13 -p 4K19A < ebcode.txt  

gives back the message even if I try hard to mess up ebcode.txt 

 

If I scan ebcode.txt & use phase=bit*pi, how long should my 

bits be keyed? Is synchronisation against UT critical? 

 

Please excuse my earlier failed efforts & silly questions. 

 

73 

Joe VO1NA 

 

 

EbNaut_spectrogram_151121_2127a.png
Description: PNG image