Hi Paul,

Thanks for the efforts and feedback.

Its strange that the 3 char message didn't decode with 60s symbols!

I could start in the morning hours, 3 characters at first.
So be prepared! :-)

73, Stefan


Am 27.10.2016 21:10, schrieb Paul Nicholson:

Recent transmissions from Stefan

 2016-10-05_21:00:02,+138880 6470.0 S140 C15 8K19A
 2016-10-07_21:00:02,+166400 6470.0 S80 C15 16K25A

 2016-10-10_14,+29h 6470.0025 carrier

 2016-10-12_14:30:00,+16640 6470.0 S40 C3 8K19A
 2016-10-13_06:00:00,+41600 6470.0 S20 C15 16K25A
 2016-10-14_13:15:00,+24960 6470.0 S60 C3 8K19A

Sadly, unable to decode any of the EbNaut messages
although 2016-10-07_21:00:02,+166400 looked close
with a couple of partial decodes:

        'DK7FC4R,3ZSB35X'  Eb/N0 = -0.4dB
        'DK7FC4273 U8Z  '  Eb/N0 = -1.1dB

The carrier has been useful and I managed to gain
nearly a dB using it to play with blanker settings,
E-field mix and loop pointing.  E-field mix is
fiddly, amplitude depends on the weather!

Using the carrier (which decodes to **** messages)
I obtained the following all-star decodes from
the 06:00 to 18:00 signal,

        16K25A S30 C8  Eb/N0 = +2.3 dB ber 37.1%
        16K25A S25 C11 Eb/N0 = +1.2 dB ber 37.8%
        16K25A S20 C15 Eb/N0 = -1.0 dB ber 40.0%

         8K19A S60 C9  Eb/N0 = +0.9 dB ber 33.8%
         8K19A S45 C14 Eb/N0 =  0.0 dB ber 34.1%
         8K19A S40 C15 no decode

I continue to think that the best prospect for a 15 char
success is the 12 hour daytime signal with 16K25A 20
second symbols.  It might take a couple of attempts.

We seem to struggle with transmissions that use both
day and night.  I've tried compensating for the extra
~60 degree nighttime phase lag but that didn't help.

I'm running some trials with 24 hours of the carrier
to see what can be done.   I've also been playing
with a suggestion that Markus made some time ago to
normalise the day/night amplitudes in various ways.

--
Paul Nicholson
--