[Fwd: LF: cw speeds]

Andre Kesteloot akestelo@bellatlantic.net
Wed, 04 Feb 1998 09:21:28 -0500

drassew2@interalpha.co.uk wrote:

> Re G3YMCs comments
> I don't know if Dave he missed the point of slow CW, but it is not for
> copying by ear, wouldn't have thought 3WPM (0.3s dot period?) could be
> decoded by this means anyway!  The 2 second dot periods as used by DA0LF for
> his high speed local working certainly isn't - although I suppose you could
> always use a stopwatch!  SLOWCW is for use by waterfall / spectrogram
> display methods using DSP techniques, which can be very simple interfaces on
> a PC plus basic FFT software.  If you want to try receiving this mode using
> stone-age technology, then a very narrow audio or crystal filter plus chart
> recorder or similar (stopwatch and meter ?) will suffice, but then the
> frequency needs to be tuned in very accurately.
> I certainly don't accept the comment about much stronger signals therefore
> not needing narrow modes - a number of us prefer to use technology to
> advance our contact potential.  With the whole band open to Europe I shall
> be expecting to use very narrow modes to work into DL, OH, EI etc.......
> Make use of the narrow bandwidth to get the DX if you cannot manage huge
> antenna arrays.
> Tests we've already made on 73kHz, with G4GVC, have shown the advantage of
> PSK over CW - we have demonstrated something like 6 - 10dB advantage at 1 -
> 2 characters per second (equivalent to about 12 WPM CW), and that was with a
> simple PSK system using hard keying of the transmit signal.   Using a more
> optimised linear mode such as PSK31 another few dB advantage should accrue
> on a bandwidth for bandwidth scale.   If you really think that most of the
> operators on 137 will be able to read 20 - 25WPM morse then good luck, but
> this band is (hopefully) unlikely to attract the type HF DX operators that
> CQWW thrives on.  If it does, then the experimenters won't stay long!  A lot
> of the 73kHz operators came from the microwave community where Morse is
> often used, but not for serious ragchewing, just as a necessity to overcome
> weak or frequency spreading signal paths.  DSP techniques are of less value
> there (at the moment anyway) due to the frequency stability problem.
> I'll leave the actual mechanics of the bandplan to the committees.  It
> doesn't really matter where what goes, but history has shown, on the
> Microwave bands in particular, that bandplans come about via a pragmatic
> (and sometimes dogmatic) approach.  eg if one person operates a mode at one
> particular frequency (perhaps because he happened to have a crystal there?)
> that becomes the centre of activity for eternity - certainly not ideal but
> that's how it happens.  Hence my suggestion of 137.1 as the centre for
> narrowband modes as that is where DA0LF currently sits.
> I have no sympathy whatsoever for users of VFOs who cannot measure their
> frequency accuracy to crystal controled levels.  The band is only 2100 Hz
> wide and just measuring frequency by listening to an uncertain audio tone in
> a speaker is fraught with dangers.  If at least 10 Hz absolute certainty in
> frequency setting is not possible then perhaps you should be keeping to
> crystal controlled drivers.   And what has happened to high stability VFO
> technology these days.  If some of the VFOs I have heard of 73 were scaled
> to HF they would be drifting and chirping so much no one would hear
> anything.  If 500 Hz stability on a 5 MHz tuneable VFO is taken as being
> adequate at HF (many constructors have done much better than this - look at
> Tech Topics articles in the early 1980s) this translates to 7 Hz at 73k and
> pro-rata.  Incidently, nearly all 10GHz operators these days regularly, and
> easily, get 5kHz frequency stability from hill top operated portable
> equipment.  Translated to 73kHz this corresponds to 36 milli Hz ...  enough
> said.
> On a different note, my linear Tx driver is just about finished so I will be
> able to use the PSK31 mode on 73 and 137.  Experience of this mode on 3.5MHz
> suggests hand typing speed communication (as fast as normal RTTY) with good
> copy on signals barely audible - and its only 31Hz wide at the -60dB points!
> The DSP receive filter is so sharp that it just about kills QRM that would
> otherwise be considered co-channel.  Using this system it could even be
> possible to have a contact in the middle of the tones of the RTTY signal on
> 73kHz - if the rest of the receiver had a sufficiently wide dynamic range to
> cope with this.
> Finally - sorry for not reporting on your beacon Dave G0MRF!  Heard all of
> Friday and Saturday's transmissions when it was on.  Is this really intended
> to be a 24 hour system as you originally said, in which case I could try
> some long term propagation monitoring tests.  Far more interesting than
> ragchewing and knocking off callsigns in one morning.  Could you close up
> the period of no carrier between callsign sequences to make continuous
> monitoring easier.
> A reminder - if anyone needs a beacon keyer, the PIC based design by G0IAY
> provides a non volatile memory based design, programmable via an RS232
> interface from any terminal, in a unit the size of a postage stamp.  A
> version is also available for slow CW.  Contact G0IAY QTHR
> Andy  G4JNT