From: DL4YHF@aol.com
Message-ID: <11.d5dbfa2.27710908@aol.com>
Date: Tue, 19 Dec 2000 13:55:04 EST
Subject: LF: Re: PSK experiments, QSO between DK8KW and DF0WD
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 Geri, Rik, and others,

Concerning the PSK31-signal generated from DF0WD:
Oh yes, the signal must have been a little "broad" because the transverter 
which I used at DF0WD was a bit overdriven and so not really as 'linear' as 
it should have been. The absolute maximum output is about 100 Watts (with 
square waves at the drains), the output in PSK31 was much less, but still 
above the compression point of the transverter's PA. The bias current was 
quite low because usually I operate in CW.
I was quite surprised to hear Geri's transmission to DL7YA, so I tried to 
call him (with success), though my rig was not properly set up. Furthermore, 
there was not even a PTT control from the computer to the radio, so I had to 
ask a "co-operator" to press the PTT of the IC706's microphone. Some letters 
of DF0WD's transmission have been missing because of the quite large 
delay-time of the 'human PTT switch' !

But things will improve. I am currently modifying the PSK TX algorithm (taken 
from WinPSK) so it can generate "phase ramps" instead of "amplitude shaping". 
This should reduce the bandwidth, even if the PA operates in non-linear mode 
as Klaus (DJ5HG) explained in an earlier posting to the group. Also thanks to 
Claudio (IK2PII) for his contribution. I have to think about using a 
four-quadrant multiplier, maybe this is another way to produce the slow phase 
shift. An I/Q multiplier is exactly what the 'original' amplitude-shaped BSK 
modulator uses. At the present time, I generate a kind of PSK with "soft" 
phase transitions using a simple sine function with a slowly moving "phase 
offset" to the argument, like 
    y(t) = A * sin(  F * t + phase),     where 
       'A' is the output amplitude (constant !).
       'F' is proportional to the frequency of the transmitted signal,
       't' is the time (or the audio sample counter, which is equivalent)
       'phase' generates the phase-shift modulation.
Note that this is a rough simplification ! The trick is to find a proper 
function for the 'phase' to move between 0 and 180 degrees. But I don't want 
to bore you with details, because experiments are still going on.

So thanks again to Geri for my first 2-way PSK QSO on LF and to Rik for the 
report and the comparison of both signals.

Happy Christmas to all, and all the best for 2001,
     73 from Wolf (DL4YHF, DF0WD)