On a sunny day (Mon, 22 May 2006 15:32:12 GMT) it happened Jan Panteltje
wrote in :

On a sunny day (22 May 2006 07:47:53 -0700) it happened "Telstar Electronics"
wrote in
s.com:

Jan,

Sure you can run class D at 27MHz... the problem is distortion. Class D
is very efficient but is really only meant for digital signals.

www.telstar-electronics.com

Hi, you know class D audio amps exist I suppose?
There are even class D power amp chips for audio.
These use PWM (pulse width modulation).
You can make any output waveform with PWM (as long as the switching frequency
is much higher then the output frequency).

This links shows some basics for a modulator (driving class D for example):
http://www.netway.com/~stevec/ham/pd...lid_state.html

I can also think of a 'digital' solution, say you have 8 RF amps,
1W, 2W, 4W, 8W, 16W, 32W, 64W, and 128W,

Actually I think I have this wrong, should be Volts (RF output) ratio or
powers of 2.
So 1 W in 50 Ohm: U^2 / 50 = 1, so U = sqrt(50) = 7V eff, the next higher
higher one would be 14V eff or 3.92 W,
better start at the high side...

So for your 300W in the highest bit: U^2 / 50 = 300, U^2 = 15000,
so U = sqrt(15000)= 122V eff.
The next lower one (bit 7) would be 122 / 2 = 61V eff, so 61^2 / 50 = 74.42W,
bit 6 would be 61 / 2 = 30.05 V, makes 18.06W
bit 5 would be about 15 *15 / 50 = 4.5W
bit 4 would be 7.5 x 7.5 / 50 = 1.125 W
etc etc.
I have also been thinking about doing SSB this way.
For SSB we need half the bits, and one bit to change the phase of the driving
carrier 180 degrees.
Hope I got that right.

You'd be selling lots of amps!!!