From: DL4YHF@aol.com
Message-ID: <cb.cfe19d6.2789c517@aol.com>
Date: Sun, 7 Jan 2001 08:11:51 EST
Subject: LF: Tests with PSK31, PSK10 and more..
To: rsgb_lf_group@blacksheep.org
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

Hello group,

Just did a few tests (off-air) using a slow variant of PSK31 which could be 
called  PSK10, PSK05 etc.
Instead of the signaling rate of 31.25 symbols/second from G3PLX's original 
PSK31, I used only 10 or 5 symbol/second. This reduces the bandwith to 1/3rd 
or 1/6th of the original PSK31 signal, and thanks to the nice coding scheme 
used in PSK31, the character transmission speed is still acceptable for a 
'real' QSO. 
( For those who are not familiar with PSK31: 
   It uses a code set called 'Varicode', 
   which uses fewer bits for frequent letters like 'e' 
   than the others, similar to the Morse code.)

To test the effect of the reduced symbol rates, I added white noise to the 
signal and compared the decoded text with the original PSK31. I decreased the 
signal/noise ratio until the decoded text became undetectable.
The results were as expected: Using "PSK10" instead of "PSK31", the noise 
could be increased about 3 times in amplitude for the same detection quality.

However, if you want to test the 'slow PSK31 variants' on air. Because all 
PSK31 programs I had did not support other symbol rates than 31.25sym/sec, I 
modfied the PSK modulator and detector from WinPSK (by AE4JY) and stuffed 
them into Spectrum Lab.
I just uploaded a new release of SpecLab (V1.63) with a built-in 'digimode 
terminal'. It is available at www.qsl.net/dl4yhf/spectra1.html. Beta testers 
are welcome.  Bug reports please to DL4YHF@qsl.net. How about some 'QRPP' 
tests,  Geri ?

------------ snip ------------
Drifting frequency of DF0WD's CW signal:
The problem is a result of the poor stability of the main oscillator in my 
old Icom IC706. The fan blows warm air from the final amp towards the 'master 
crystal'. I watched the Loran lines yesterday, and found them 'wandering' 
around the sreen for about 30 seconds after one over (using only normal CW 
here). 
I thought about buying a 'TCXO' for the IC706, but the cost of about 120euro 
for that tiny piece of PCB kept me off. Any better ideas ?
----------- snip -----------

73's Wolf  (DL4YHF, qrv on LF only from DF0WD).