The Eimac tube manual says a single 4-400 typical operation at 3000 volts on
the plate is 630 watts high level plate modulated carrier output. So, 630
watts X 4 = 2520 watts pep capability.
So, a more conservative 375 watts X 4 X 1.30 = 1950 watts pos. peak.
Dividing that by 2 = 975 watts of peak audio output required from the
positive producing modulator tube to make 1950 pos. peak RF output.

He'd be better off to consider what the modulator is actually doing.
The modulator is DOUBLING the anode voltage for 100% positive peaks.
Say the anode voltage is 1000 volts and current is 1 ampere. The
modulator drives a load impedance of 1000/1 or 1000 ohms. It has to
produce 1000 volts PEAK, which on a sine wave is 707 volts RMS.

707 volts RMS into a 1000 ohm load is 500 watts of sine wave power.

The modulator SINE WAVE power for 100% positive peak is exactly half
the plate input power. This ASSUMES a triode well into class C for the
PA. A tetrode will actually often use a bit less power because it often
must be modulated on the screen (or control grid) and anode at the same
time in order to be linear. The anode of a tetrode or pentode does not
follow square law power rules like a low mu class C triode does.

You can generally consider for voice even LESS power is required, a
good rule of thumb is about 25% of dc input power. So as a general rule
a 250 watt average audio power modulator will modulate a 1000 watt
transmitter 100% on voice.

I suspect he has something wrong with the PA not following square law
power change with anode voltage changes. In order to be linear with
modulating voltage changes the PA has to be well into class C
(switching very hard) and it must, with a terode, have some portion of
audio power applied to the screen (you could do the drive power
instead, but that's tricky).

He could have an audio mismatch or the modulator system might be
incapable of supplying the necessary power which is half the dc plate
input power for 100% modulation with a sine wave. He could have a PA
that does not follow square law power as voltage is changed. There is
absolutely NOTHING wrong with the rule that a modulator capable of
supplying half the dc input will modulate a properly designed PA 100%.

73 Tom



The 4-125 typical operation indicates 330 watts of AB1 is available. This
is a far cry from 975 that I need. On the other hand a 4-400 in AB1 will
deliver 1540 watts.

The negative vs. positive peaks will be controlled by a broadcast quality
Innovonics AM asymmetrical compressor limiter which I am currently using in
the shack.

The simple unexplained rule of thumb I came to understand early in my early
ham days is that a 375 watt transmitter requires 187.5 watts of audio for
100% modulation. So, a 4-125, with 330 watts of capability, should be more
than what is needed. In practical terms this is very misleading. That is
why I been beating a dead horse for three years while two broadcast
engineers could not explain why I could barely attain near 100 modulation.
Little did I know that I was driving the pants of the poor 4-125's while
trying!

Hence, I ordered a custom built filament xfmr that will power two 4-400
modulators. Hope to have all that working in about two weeks.

73,