CW Bandwidth (was: [Elecraft] K2 w/KSB CW reverse BFO)

[Vic Rosenthal]

"George, W5YR" wrote:
> 
> Good arguments can be made, and have been in other forums, that rise and
> fall times affect occupied bandwidth of CW signals to a large extent,
> possibly as much or more than mere keying speed.

I think the rise/fall times are the whole story here, George.  The bandwidth is
determined by sidebands inherent in the leading and trailing edges of the
pulse.  The relationship to keying speed is simply that the faster you key, the
sharper you need to make the waveshape so that the elements don't run into one
another!  To give an exaggerated example, if you are sending five dits at per
second and your rise/fall times are greater than 100 ms., then the dits actually
begin to run together.  

Vic K2VCO