LTPI and drive to power tubes... problem, maybe not?

[EFK]

I have an amp here that is driving a quad of 7591S (JJ) tubes, and I have something interesting going on that has me scratching my head. The PI is a typical ltpi, I started with a AC30 PI and have been tweaking it by ear. Currently, the PI cathode R is 820 ohms and the tail is split (for NFB addition on a switch) between a 27K and 4.7K. Grid leaks are 1M and both plate R are 100K. Amp is cathode biased, very hot (near 100% at idle) with a shared cathode R for each pair of tubes, but I am noticing something weird. When I monitor the cathode current when the amp is being cranked and driven (I'm using 120 hz and 1Khz into a resistive load), the power tubes being driven by the inverting/input side of the PI are showing almost no increase in cathode current from idle, while the power tubes being driven by the non inverting side of the PI are showing a climb of about 3 volts across the cathode R. When I check the voltage at the plates of the PI, they are both nearly identical at idle, but when driven, the inverting side is dropping around 10 to 15 volts more than the non inverting side. I have swapped around PI tubes, and I have swapped around the power tubes (they really are a very closely matched quad). I hope I'm being clear here. I don't know if this really is a problem or more just an observation, but any commentary would be welcome.

[Merlinb]

This is quite common. It's due to the two sides of the LTP clipping differently, and since you're seeing more drive from the non-inverting side, it suggests your LTP is cool biased. Hot biasing the LTP will drive the other valve harder instead, and somewhere in the middle you should get more-or-less equal drive to the power valves. Most amps do it, so it's nothing to panic about.

[EFK]

Thanks Merlin. I lowered the cathode resistor to 470 ohms and it seems to have brought the two halves closer together. I might play with the tail resistor a bit, but since it sounds good I'm inclined to just leave it be! My goal is to bias it as hot as possible without allowing the current to run away and over-dissipate either side, and right now I can just about bias it to 100% at idle and not have to worry about things climbing much, so I don't have to worry about any crappy sounding crossover distortion.