[Fwd: LF: IK1ODO Tests]
Sun, 16 Aug 1998 12:09:30 -0400
> >From Toni, HB9ASB:
> Dear Andy, I strongly support your proposition: we should get a band plan agreement as soon as possible. From my point of view, the "computer-segment" should be at a band edge and not in the middle. So it would be "under attack" only from one side. May be separated by a segment for local low power experiments or just a "guard band".
> If possible, we should even reserve both band edges (the first and the last 100 Hz) for computer-communications. For two reasons: First, the QRM-situation is different from one region to another, second: also some "mega-strong" stations will enter this mode in the future. At least I probably will, when the improvement of my station and antenna (limited by the 1W ERP rule, my garden and my budget, HI) comes to an end and the limits for conventional QSO's are reached.
> Regarding the length of CQ's, I'm happy with longer calls - otherwise I may miss a weak signal when tuning slowly over the band with a narrow bandwidth. This takes me - with my equipment - about 2 minutes. So It may take me some minutes to detect a signal and some more minutes to identify it! I think 10 minutes for a call is just fine - we are not on short wave and as I remember my microwave time, we used also very long calls to attract the attention. But, if it's too long, we may miss the good propagation. I'm aware that we have a different situation down here: The band is not crowded yet, in central Europe.
> I feel that computer-communication is a good way to push the limits on long wave and it could probably be the dominating mode on longwave in the future. May be, one day we will have to inverse the situation and give the small computer-segment to the "old fashioned", but "mega-strong" operators, to protect them from the common traffic, HI.
> I'm looking forward to see the first real computer-QSO.
> 73 de Toni
> Andy Talbot wrote:
> > Just about managed to detect a signal from IK1ODO initially around 0400 -
> > 0500z, this morning (Sunday) but the signal was so weak that I had to use a
> > bandwidth that smeared the CW characters and so couldn't read the callsign
> > properly. However, the waterfall plot does show some structure that appears
> > to be consistent with 20 second dot length CW so I'm reasonably sure the
> > signal received was genuine (0.21 Hz high)
> > THe signal was detected initially using a 10 Hz frequency span and 256 point
> > FFT giving an effective bandwidth of around 0.08 Hz. The signal was too
> > weak to read the CW and switching to 5 Hz span smeared the characters too
> > much. Perhaps a 40 - 60 second dot length will work, but only if a
> > frequency is used well away from the strong fast CW operators around here.
> > After 0500 ish the signal was clobbered by keyclicks / sidebands from G3LDO
> > calling CQ for long periods just 40 Hz away which completely obliterated any
> > further serious experimentation. The actual strength of the signal was no
> > problem, after all 16 bit sampling of the audio gives over 80dB dynamic
> > range, but AGC on the receiver needs to be switched off for obvious reasons
> > and frequency spacing has to be sufficient to remove keying sidebands which
> > are frequently 50 - 100 Hz wide for normal CW.
> > We seriously need a bandplan, or at least some agreement, on 137 now. Put
> > aside a segment for really weak signal working using SLOWCW, PSK or
> > whatever and make it absolutely inviolate - It only needs to be 50 Hz or so
> > wide. Also, if all the mega-strong stations just want to work eachother on
> > manual CW then please keep the CQ calls short - after all you wouldn't dream
> > of calling for 10 minutes continuously on the HF bands.
> > Incidently, Peter, you're still drifting - I spotted 10 Hz shift on your
> > transmission this morning at the start of the keying cycle.
> > Andy