[Elecraft] K3 CW rise time mod - Definition

[Ron D'Eau Claire]

    I would like to see someone please take this discussion one step 
further and define the terms "hard" keying and "soft" keying, and 
describe how either an operator or someone listening would be able to 
discriminate between the two.  My rig is #095 and I frequently contest 
in the 30-34 WPM range.  How would I know if I'm guilty, if guilty is 
the right word,  of "soft" keying?!  Is this an issue only for those who 
are keying external amplifiers?

    73,
Gary, VE1RGB

---------------------------------

It's completely subjective, Gary, as applied here.

"Hard" keying used to be keying with obvious clicks caused by the sidebands
produced by a very abrupt rise from zero to full power and then the drop to
zero at the end of each dit or dah. Extremely "soft" keying is one where you
aren't sure a dit or dah has been started until it's nearly done. 

Obviously "hard" keying can be easier to copy, since the transition at the
start and end of every dit and dah is absolutely clear, but the clicks take
up excessive spectrum space. They can be heard off to the sides of the main
carrier. In an extreme case, they might be heard several kHz from the
signal. 

The only CW signal with absolutely NO sidebands (clicks) is one that isn't
keyed, but careful shaping of the rise from zero to full power and the drop
back can minimize them to the point where they are almost inaudible, even
very close to the carrier, with rise and fall times appropriate for easy
copy at a wide range of CW speeds. It's a trade-off between speed and
sidebands. The faster the data rate (CW speed) the wider the signal. It's
basic physics, and the reason why an intelligible voice signal must use much
more spectrum bandwidth than a CW signal, for example. 

In an absurdly extreme case for CW, having the keying rise and fall times
too slow for the speed would result in a steady CW signal that only varied
slightly in strength between dits, dahs and pauses. The signal would still
be building up when the dit or dah ended, and then wouldn't reach zero
output before the next dit or dah started the rise again. It would sound
like a steady carrier with, perhaps, a tiny amount of modulation barely
audible at the keying rate. 

Obviously, such a signal would be useless for CW communications, just as a
signal with too hard keying that produces wide sidebands uses up excessive
spectrum. 

The "right" keying is somewhere in the middle and, obviously, the right
characteristic depends upon the keying (CW) speed used. Most rigs, including
the Elecraft rigs, adopt a compromise that produces a good, readable signal
over the normal keying speeds used by Amateurs.

The original values used in the K3 were fine at moderate speeds but, as the
speed was increased, it was too "soft" on the leading edge of each dit or
dah for optimum readability. That's what the change in the value of C222
affects. 

The end result is keying that is easier to read at higher speeds. 

I normally operate at a maximum speed of 20 or 25 wpm, and didn't think the
keying sounded "soft" at all, nor did anyone I worked mention it. Even so
it's a good idea to make the change. The notes with the mod suggest Wayne
plans to offer the ability to change the keying characteristic with a menu
setting. If so, the values he provides will assume all K3s have the same
hardware characteristics. If you don't have the recommended value for C222,
the results you get with various menu settings might be quite different than
expected. 

Ron AC7AC