I understood your post as if you would like to gain some understanding (without having to consult Terman) on why the voltage & current figures you read on your RCA TT-3 manual for the 833 tube do not apparently match the 400-W peak-envelope-power figure commonly quoted for a 100%-modulated 100-W AM carrier.

In my understanding, there are two main issues that you did not consider in your reasoning:

1) the voltage / current pairs you quoted for the 833 tube correspond to a fixed grid voltage (either 0V or -50V). In practice the grid voltage is sinusoidal (at the carrier frequency) and, for a class-C amplifer, it only causes plate current to flow when getting above the interdiction threshold. So, plate current only circulates for a fraction of the RF cycle (180 degrees). Therefore plate current is not sinusoidal and its average value (along half a cycle), i.e. what you read on a DC meter, is closer to the peak value compared to a purely sinusoidal waveform. When plate voltage varies due to modulation, the grid interdiction threshold varies and so the circulation angle does. This turns into a change of the average-to-peak current relationship. That said, it immediately follows that simply taking voltage & current figures at a fixed grid voltage would not make much sense. All what said does not take into account the presence of the plate tank circuit which, as someone else has noted, yields a remarkable effect on actual plate voltage & current figures

2) the modulator is designed to feed a given load resistance (depending on the modulation transformer winding ratio). If the modulator load (i.e. the class-C final stage) does not show a linear current / voltage relationship, the plate voltage will, partially, self-adapt itself along the audio-frequency cycle, thus smoothing the tube non-linear behavior.

In addition to that, there are other effects (like the grid leak bias causing grid voltage to somewhat follow the modulation) that someone else has already pointed out.

The above arguments do not provide a justification of why a class-C tube shows, in practice, a reasonably linear voltage / current behavior, but they should at least give you sufficient evidence why your reasoning is way too simplified to credibly deny the 1:4 PEP ratio of AM-modulated signals.

73

Tony I0JX