Here is the result list for VO1NA last night:

 

   2015 Nov 27/28

start  Eb/N0 time offset
 UT      dB     seconds

 

2030     2.2     1.2
2100    -0.9     1.2  (after notching QRM on 137776.98)
2130     5.7     1.0
2200     3.1     1.1
2230     4.5     1.2
2300     3.7     1.2
2330     4.9     1.4
0000     3.5     1.3
0030     2.1     1.3
0100     7.1     1.3
0130     5.5     1.6
0200     nil
0230     5.4     1.7
0300     5.8     2.0
0330     3.5     2.0
0400     9.6     1.9
0430     3.3     1.9
0500     3.5     1.9
0530     nil
0600     3.0     2.0
0630     2.8     2.0
0700     2.1     2.2
0730     nil
0800     nil
0830     nil
0900     nil

 

Only the 4 UT sequence produced a faint blue haze on the TA spectrogram.

 

Note that the "nil" decodes usually produced a couple of false outputs at high rank numbers. A few of these consisted solely of A-Z and 0-9, so reducing the character set would not really solve it (and i.m.o. would not be desirable anyway). Using a longer CRC checksum (eg. 24 bits) would take care of it but of course somewhat reduce coding efficiency. On the other hand, I don't feel that occasional false decodes are a big problem.

 

All the best,

Markus (DF6NM)

 

 

Impressive.  That's within a decibel of the limit. 

 

The TX power is 30 watts.  Less than 10 should give 

valid decodes.