Hello Walt,

I did some simulations with EL34.

Conditions:
Common Grid circuit, Valve connected as triode, 3.6 MHz.
Vsupply, 600V, Vgrid -58V, bias current 40mA (yes I know that it will
run very hot).

Maximum available input power: 9.2W out of 600 Ohms, sinusoidal.

Anode loaded with 20uH, 96pF, 3800 Ohms (parallel circuit), QL= 8.4,
so the anode voltage has sinusoidal shape.

Without changing the drive (9.2W), 3800 Ohms with 96pF gives highest
output power of 64W (Vanode = 698Vp). So power gain is 7.0.

Output impedance:
Change to 4100 Ohms load gives Zout = 1826 (voltage saturation),
Change to 3500 Ohms load gives Zout = 4350 (current saturation),
So for small change in Zload, Zout will be close the load resistor
that gives maximum power. This supports that under well-matched
conditions, given a certain drive level, Zout = ZLoad.

Input power reduced with 3 dB, output becomes 32.6W (Vanode = 497.5Vp,
gain = 7.08). There is a slight increase in gain showing that at 64W
the amplifier is not into deep voltage saturation.

Output impedance at 32.6W output:
Change from 3800 to 3500 Ohms gives Zout = 8100 Ohms.
This shows that reducing the drive level (without changing the
matching) results in Zout not equal to Zload for this triode common
grid circuit.

Also shown that small changes in load (reduction or increase seen at
the anode) results in changes in output impedance. That is why I used
three resistance values for determining Zout at well-matched output.
I used low frequency to avoid endless matching because of parasitic
capacitances and I also skipped the output matching (I have to work
sometimes, so time is limited).

It is not strange to see an RF swing above supply voltage as this is a
common grid circuit with cathode grounded for DC.

If you like, I could do same exercise for modern amplifier topologies,
but be prepared to see unexpected results (As Tom, K7ITM already
noted).

Best regards,


Wim
PA3DJS
www.tetech.nl
remove abc first in case of PM.