[Elecraft] K2 CW speed

[Ron D'Eau Claire]

Kevin, AC0H, wrote:

How do the rise and fall times of the SSB and CW waveforms compare? 
Isn't SSB, even with VOX delay turned all the way down, a lot slower 
than CW? If the SSB rise and fall times are slower than CW how do you 
get clicks?

I would agree that nobody in their right mind would cut loose with 
800wpm HSCW on the HF bands becasue of the bandwidth used. I can't 
remember the old formula for figuring the bandwidth for CW signals of 
any arbitary speed.

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

With proper waveform shaping, the CW bandwidth in Hz that is required is
approximately 4 times the speed in WPM. So a 20 wpm signal needs about 80 Hz
to pass the on/off transitions without stretching them out and impairing the
readability. SSB allows 2.5 kHz or more bandwidth which would support about
800 wpm. That is, of course, way too much for the CW bands. 

As others have pointed out, that takes being careful about the nature of the
driving signal, possible noise introduced by the oscillator-rig connection,
and a clean audio waveform, as well as proper keying waveform. 

A number of SSB rigs used that technique to produce CW in years past. Some
early SSB transceivers had no ready means to key the RF source, so they
provided a keyable sidetone oscillator that also fed the TX chain. 

As a brass pounder 99% of the time, I didn't follow the transition from
audio-injection to keyed RF sources in the newer transceivers, but I suspect
it had to do with the issues of keeping the signal clean with adequate
sideband and carrier suppression.

Current rules require such spurious emissions to be at least 40 dB below the
peak carrier for transmitter > 5 watts (with a 50 mW limit no matter the
power) or at least 30 dB down for a QRP transmitter (up to 5 watts). 

Elecraft spec's the carrier suppression at -40 dB (typical) for a K2 with
the SSB adapter. I suspect that's very conservative, and allows for some
sloppiness in setting the carrier null. Still, I'd check the actual level
from my K2 if I tried that scheme running QRO. The opposite sideband
rejection should be no concern considering the K2 OPT1 filter
characteristics. 

I would not expect the VOX to be at all useful in such CW. Actually, a lot
of the HSCW is done like we used to do ALL CW just a few years ago: no QSK!
For scatter work, one doesn't know what one received for a period of time
anyway. It has to be played back slowly while the op tries to hear the
fragments of CW in all the noise. It's about as far from QSK as one can get.


Ron AC7AC