[1000mp] ARRL testing of Clicks

[Tom Rauch]

> I could find no qualitative judgment about the CW operation of the
> Mark V in the Nov. 2000 QST review.  Certainly, if you look at fig. 4,
> the initial dit has a very sharp "on" transition leading one to
> suspect that the rig might be "clicky," but nowhere in the review is
> this deficiency addressed. 

Here's the short form of this all....

The slope of the rise and fall at every instant of time controls the 
frequency width of clicks. You do not want anything even approaching 
vertical. The amount of level change at a given slope determines the 
amplitude of the clicks at that bandwidth. 

We do want anything sloped steep even for a change of a percent or 
two of the overall amplitude, at any point in the entire output 
period. That means we have to look for steep slopes down low, just as 
much as up high. 

Obviously some rigs that have 200% overshoot won't click as much as 
other rigs that have a really sharp rise or drop near zero amplitude 
but "look excellent" near the tops.

Even if the envelope is perfect, and has a raised sine or other nice 
multi-pole filtered shape, it can still click. How? It can have FM 
problems, or a VCO that is swooping to a new frequency while the RF 
is on.

The only way to compare one rig to another is to listen to it sending 
a string of dots both in QSK and non-QSK while working split and 
working simplex on a selective system that indicates peak power level 
in reasonably narrow slices of frequency, or do a multiple sweep peak 
storage on a peak-storage spectrum analyzer and look at the picture.

This is similar to what the FCC now requires with commercial FM 
transmitters. They want to know actual bandwidth, not just what a 
deviation meter says.

73, Tom W8JI
W8JI@contesting.com