The cat sat under the stars...

 

Confirming Paul's observation. The strong unmodulated carrier was well visible in the grabber spectrograms. Got an error-free ******* decode of the 0:00 transmission:

 

found *******
list rank 0
reference phase    0    0    0    0
carrier RMS 4.357e+005 noise RMS 5.605e+003
carrier S/N 37.81 dB in 822.4 uHz
carrier Es/N0 9.97 dB
carrier Eb/N0 21.80 dB
info bit period 30.48 seconds
symbol error rate 0/608 = 0.000 %
Shannon capacity 6213.1 bits/hour
Shannon efficiency 1.9 %
symbols file 11240014-symbols.csv
elapsed 2 seconds

 

 

-----Ursprüngliche Mitteilung-----
Von: Paul Nicholson <[email protected]>
An: rsgb_lf_group <[email protected]>
Verschickt: Di, 24 Nov 2015 9:20 am
Betreff: Re: LF: VLF in VO

  

 

Joe VO1NA wrote: 

 

 > An 8K19A sequence with 7 characters &  2 second long 

 > symbols was started around 2300 (I hope). At 2330 I will 

 > try to sync it against UT and send it again. 

 

For both messages I got an 'all-star' decode 

 

  23:00 

  found rank 0: *******  ps [  8  -60  -60  -60  -60] 

  re-encode 167/608 ber 2.75e-01  Es/N0=-7.4 Eb/N0=4.2 

 

  23:20 

  found rank 0: *******  ps [  0    0    0    0    0] 

  re-encode 1/608 ber 1.64e-03  Es/N0=6.4 Eb/N0=18.0 

 

which means you were sending pure carrier.  (The CRC16 isn't 

initialised to 0xffff as is usually done, instead zero so that 

a payload of all zeros encodes to all zeros and transmits/receives 

as a pure carrier). 

 

The signal at 23:20 was very strong, only one of the 608 symbols 

was demodulated incorrectly. 

 

 > Next start time 0000 UTC 24 Nov. 

 

Also unmodulated carrier 

 

  found rank 0: *******  ps [  1  180  180  180  180] 

  re-encode 1/608 ber 1.64e-03  Es/N0=6.4 Eb/N0=18.0 effrate=1/14.48 

 

with again (suspiciously) just one symbol error. 

Checked: the 23:20 symbol error was at offset 283, the 00:00 

was at 512. 

 

Something wrong with the modulator but the signal is huge. By the 

time it arrives here you've enough spare dB to cross another 

ocean.   Once you've got things fettled we should try daylight 

path which would a better test of the mode. 

 

-- 

Paul Nicholson 

--