Hi Paul,
I stand corrected by the RCA Radiotron Designers Handbook (27 engineers). Here is what they state:
"If all harmonics are within the frequency range of the amplifier it may be shown that---
1) If only second harmonic is present (a condition which never occurs in practice), the ratio of total intermodulation distortion to second harmonic distortion is 3.2
2) If only third harmonic is present (another condition which never occurs in practice), the ratio of total intermodulation distortion to third harmonic distortion is approximately 3.84 at low values of distortion."
3) If the distortion is small, the intermodulation sidebands are approximately given by
Modulation percentage of first intermodulation sideband} approx 2 times second harmonic distortion percent-
Modulation percentage of second intermodulation sideband} approx 3 times third harmonic distortion percent.
Thus intermodulation distortion is automatically weighted by the order of the distortion.
As a very rough approximation, the ratio I.M./H.D. may be taken as:
3.2 for single-ended triodes and 3.8 for PP (triodes or pentodes) where I.M. = total intermodulation distortion (r.m.s. sum)
and H.D. = total harmonic distortion," provided that the operation is restricted to the normal low-distortion region." The ratio tends to increase as the distortion increases."
As one can see, it is advantageous to keep the distortion as low as is reasonably possible.
--
The output Z of a tube varies with plate current. It is hard to find pictures or graphs showing Rp changing with changing plate current but imagine a snow ski jumper on the ramp. Near the bottom of the ramp is the normal plate Rp at idle current and tends to level off.
At the top of the ramp is tube cutoff, Rp is very very high.
In SET operation, the positive portion of the cycle is marked by a slightly lower Rp as the tube conducts more heavily. During the negative portion of the cycle, Rp increases dramatically as less and less current flows through the tube (at cutoff, infinite). It is similar to a perfect amp with a resistor in series with the speaker. Near constant Rp in the positive portion of the cycle, and constantly changing value during the negative portion of the cycle. So the output Z of the amp varies depending on the signal applied.
In PP, this radical change is much less (though not perfect) as one tube is conducting more as the other tube is conducting less.
Added: The sound quality of both SET and PP can vary alot, depending on quality of design and parts quality, price point. I just try to present views for better understanding.
Hope this helps to explain your two questions, Paul.
