[CW] CW Bandwidth

[David J. Ring, Jr.]

Ken,

I remember in "information theory" that a constant amplitude "always" on
carrier had zero bandwidth, but as soon as the amplitude changed in some
way, or the carrier was even keyed off, it has bandwidth.

I agree with you - the waveshape of the keying is the biggest part of the CW
bandwidth.

I've never measured the bandwidth - but there seems to be somewhat of a
controversy about this.  Traditionally, the bandwidth of a signal is
proportional to the speed of the keying.  I know it is proportional to the
wave shape.  I also know that to be able to send faster you have to make the
waveshape "sharper" which increases the bandwidth.

What I don't know - but it is interesting if someone does - is that NAA used
to send CW.  They were on (around 17 kHz) and they couldn't send faster than
13 wpm.  If they went to 16 wpm the transmitter would arc over because the
antenna couldn't handle the bandwidth.

What Navy finally did was have two tanks of final amplifiers, one for
"space" and one for "mark" - each with a separate tuner for the massive 17
kHz antenna.  This way they were able to go on RTTY and beat the speed
restriction.

What I don't know is "when Navy tried to send faster, did they change the
keying constants so that the signals weren't mushy?  Is that what caused the
arcing?  Or was it simply because of the speed increase?

I would love to know that answer!

73

David Ring, N1EA

----- Original Message ----- 
From: "Ken Brown" <ken.d.brown at verizon.net>
To: "CW" <CW at mailman.qth.net>
Sent: Tuesday, September 07, 2004 1:09 PM
Subject: Re: [CW] CW Bandwidth


CW bandwidth is inversely proportional to keying rise and fall times. CW
speed capability (of the transmitter) is limited by the keying rise and
fall times. Slow speed CW sent with a transmitter that has rise and fall
times fast enough for high speed CW will use the same bandwidth as it
would use if it were sending at high speed. A narrower band width
receive system than necessary for high speed CW reception could be be
used to copy the slower speed CW sent using a transmitter with fast
enough rise and fall times to operate at high speed.

CW bandwidth is proportional to the CW speed capability of the
equipment, not proportional to the CW speed it is being operated at.

Maybe some newer DSP radios actually adust the rise and fall times to
what is needed for the keying speed being used, though I have not heard
that any really do this. It would be a good way to minimize the
bandwidth used by slower speed operations. Mostly though the rise and
fall times are fairly constant, determined by fixed component values in
the keying circuitry.

DE N6KB

>Fact was the controller was getting the CW just fine at its input, but to
>send the on-off signals at that speed, meant that those signals were taking
>most of the 9600 baud
>bandwidth, which meant the transmitter wasn't really responding to the
>keying signals
>as witnessed by the monitoring receiver! Sounded pretty weird also.
>
>
>
>>Info about CW bandwidth has been in the ARRL Handbook for many years;
>>and it does say bandwidth is proportional to speed.  Just refer to the
>>handbook if anyone disagrees with you on this one DR.
>>
>>
>>
>



_______________________________________________
CW mailing list
CW at mailman.qth.net
http://mailman.qth.net/mailman/listinfo/cw