> >From W9AC:
> "Correct to the extent that CW bandwidth may be minimized. I don't
> necessarily agree with Tom's view that the "correct" CW waveform is
> one in which the transmitted bandwidth is minimized by extending the
> rise and decay of the envelope and turning the wave edges into a
> sine-slope characteristic. The problem with Tom's approach is that in
> minimum signal conditions and at relatively high CW speeds, his
> preferred wave form becomes difficult read. In essence, the "correct"
> CW wave form is purely subjective.
That is not true by any observation I have gathered in well over 50
tests when my signal was weak on 160 meters.
I now use a waveshape that is similar to a sine rise and fall, and no
one so far can tell any difference at all when I "square it" except
the people up or down a few kHz.
I work just as much DX when my signal is marginal, and when I do
an A-B test no one so far can tell the difference unless I AM
strong.
What proponents of wide bandwidth-wasting clicking CW signals
forget is that the ear and the receiver system up to your ear, unlike
a power meter, does not measure or store "average power" or
accumulated power. Until the accumulated rise and fall time
approaches the width of a "dit", it makes no difference at all in
weak signal copy.
We have to have some respect for other people on the bands. As
far as I am concerned, people can click and clack all they like as
long as they respect my space, and stay a few kHz away from all
other operators on the band...especially weak signals.
> "My preference is for a fast leading edhe in the order of 1.5 mS, and
> a trailing decay in the order of 5-6 mS, much the same as is often
> shown in the older ARRL handbooks when grid-block keying is used.
That will click, and occupy excessive and unnecessary bandwidth.
The closest possible sideband would be almost 700Hz away, no
matter what speed you are sending. Even if you used a filtered sine-
shaped envelope, your signal would be 1.5 kHz wide. With a
normal single pole R/C filter, which gives the poorly filtered
waveshape the handbook calls "ideal", you would have garbage out
a few kHz each direction every time the envelope rose. That
garbage would be there no matter if you sent one word per minute,
or a hundred words per minute.
While you may "like hearing" an excessively-short clicking rise
time, it is poor engineering and wasteful of spectrum to transmit
that way on crowded bands. It is poor form to promote something
that will cause all CW signals to be well over 1.5kHz wide.
> This produces a percussive CW note with a slight "bell-sound." No, it
> is not as spectrally pure as what Tom suggests, but the note is very
> distinctive and easy to copy. A bit of extra bandwidth can sometimes
> be a good thing."
I'm sure it is a "good thing" for some of us when we are dishing it
out. For one thing it ensures others can't stand operating within a
kHz or two of you. When we are on the receiving end of the
needless clicks, it isn't so wonderful!
My goal is to have a click-less signal that occupies no more
bandwidth than necessary.
There is no logical reason in the world for a CW signal to occupy
more than a few hundred Hz of bandwidth.
73, Tom W8JI
W8JI@contesting.com
|