From: utzoo!decvax!harpo!eagle!karn
Newsgroups: net.ham-radio
Title: Re: Thoughts on CW
Article-I.D.: eagle.652
Posted: Wed Nov 24 21:52:23 1982
Received: Thu Nov 25 13:50:10 1982
References: sri-unix.4329 brl-bmd.453

Envelope shaping in a CW transmitter places a limit on the speed at which
it can be keyed with the dots remaining separable at the receiver.  Envelope
shaping is also what determines the bandwidth.

In the extreme case, a CW transmitter that could generate
distinguishable dots at 20,000 dots/second would be generating terrific
"key clicks" even if keyed at a much slower rate.

In order for a CW transmitter to take up the minimum possible bandwidth,
its keying envelope would have to be changed as a function of the operator's
sending speed.  If the operator were to send slower, the rise/fall times
of the transmitter envelope should be made longer, and the occupied bandwidth
would decrease.  Ideally, the envelope rise time should be adjusted so that the
transmitter just reaches full power at the point where the key is
released in a "dot", resulting in a sinusoidal envelope.  Otherwise, bandwidth
is being wasted by communicating the rise time of the envelope faster than
it need be.

The same effect is true in narrow CW filters for receivers.  If one uses
too narrow a filter (assuming the receiver remains stable), fast CW
would come out as a blur.  Any filter will spread out the rise and fall
times of the signals passing through it.

Phil Karn, KA9Q