[1000mp] ARRL testing of Clicks

[Tom Rauch]

> > Key clicks are *not* generated by short rise and fall times 
> > alone - the overall shaping of the wave is very important to 
> > the spectral output.  I've done key click testing on rigs 
> > with rise times in the 2ms ballpark and found they were quite 
> > clean. 

I guess that depends on what we consider quite clean, and what shape 
they use. Correct me if I am missing anything, but here is how I see 
it.

Imagine a pure sinewave modulating a carrier. Assume the negative 
peak is zero power, and positive peak is maximum power. It takes 1/2 
cycle to reach full power, so the modulating frequency must be half 
of .002sec/1 or 250Hz. 

Since CW is amplitude modulation two sidebands are produced that 
contain all the harmonics of the keying waveshape. With a sine shape, 
there are no harmonics but it is also important all following stages 
have nearly-perfect linearity.

Under this perfect condition the sidebands will be 500Hz total width, 
or 250Hz up and down and very strong (without significant 
attenuation) no matter how fast or slow we send.   

While that certainly is better than most radios, it is not as narrow 
as it could be. It would also be a disaster if rise or fall harmonics 
are allowed to pass through the system, or be generated in later 
stages.

While a 5mS long rise and fall do not guarantee clean operation, a 
short rise and fall guarantees clicks no matter what the rise and 
fall shape.

73, Tom W8JI
W8JI@contesting.com