[Elecraft] K3: 630m amplifiers

[Jim Brown]

Hi Tom,

It's convenient to think that, but it's wrong.  The problem is that 
neither of those signals are continuous -- both are changed to convey 
information. Mother nature recognizes a keyed CW waveform as a square 
wave modulating a carrier. Any waveform more complicated than a 
continuous carrier has multiple components, and in the case of a square 
wave or impulse, an infinite number of them. If that waveform is passed 
through a non-linear amplifier, distortion will be produced. There will 
be both harmonic and intermodulation distortion.

Take a look at slides 13 and 15 in http://k9yc.com/FTDX5000_Report.pdf. 
Slide 13 is a continuous carrier. The sidebands at -48dBC are probably 
hum. Now look at Slide 15, which is the same radio transmitting a series 
of dits.

I haven't measured an un-keyed K3 yet, but I suspect it's cleaner, 
especially with the new synth board. I may get to that this week.

Several years ago, some of the serious engineers running RTTY noticed 
that the K3 running FSK wasn't as clean as when running AFSK. Wayne and 
the crew did some serious work on the firmware and it's now better.

73, Jim K9YC

On Sun,7/26/2015 6:57 PM, Tom Azlin W7SUA wrote:
> I was thinking one of the slow multi tone FSK modes that are a single 
> tone at a time. WSQ is in the same class.
>
> http://www.qsl.net/zl1bpu/SOFT/WSQ.htm
>
> CW would be another mode that would not need a linear amp I think.