[CW] Speed vs Bandwidth

Ken Brown [email protected]
Sat, 13 Mar 2004 09:08:53 -1000


Hi Buck,

Here are a few ways to look at it.

A CW carrier that is not switched on and off and has no noise and no 
amplitude or frequency instability has zero bandwidth.

There are various reason a transmitter may have some noise in it's 
carrier which will make the bandwidth greater than zero, even when the 
carrier is not switched on and off. That is not really what were are 
interested in with regards to keying speed though.

As soon as you switch the carrier on and off you can say that it is 
being "modulated" by another waveform. When you do this the bandwidth 
increases to at least the frequency of the modulating waveform. If the 
keying waveform was square waves with zero rise time and zero fall time 
the bandwidth would be infinite. You could transfer information at an 
infinite speed too.

If the rise time and fall times are finite, then the bandwidth is also 
finite, and the maximum CW speed that can be used is also finite. If you 
sent slower, the bandwidth occupied by the signal would still be 
determined by the rise and fall times (the slope of the leading edge and 
trailing edge of the keying envelope) and not by the speed at which you 
are sending code.

A CW transmitter can be narrower than another only if it's maximum 
keying speed is lower.

All of the above is oversimplified to make the basic principal clearer. 
There are a lot of other details that enter into the situation in real 
life. The keying envelope can have overshoot and ringing. The oscillator 
can be noisy or have some instability as is gets locked on to the 
transmit frequency. There can be phase noise, etc, etc.

Hope this helps,

Ken N6KB