weird PI voltage mismatch

[andrew]

I could sure use some help in figuring this one out! The PI node reads 326.
The 100k side of the anode reads 322 at pin 6 on the PI tube.
The 82k side of the anode reads 173 at pin 1 on the PI tube.
I get 42 at each 220k/.022 junction.
I reversed the 82k and the 100k and I still get very close to the same voltages on pins 1 and 6. Pin 1 is reading low either way.
I tried three different 12ax7 and swapped power tubes and got the same results. I checked the resistor values and they do read 82k and 100k. Normally I get less than a 10v difference between pins 1 and 6. Where should I look next?
Thanks for any help.

[martin manning]

The 100K side's plate voltage (pin 6) is too high; it's barely conducting at all. There is a cap going from the grid to the tail resistor on that side, no? Is it leaky? That would drag the grid voltage on that side down and put it in cut-off.

[andrew]

I'll swap it out tonight. Thanks!

[andrew]

I swapped out the cap connected to the tail resistor but still got the same readings. I also disconnected the presence circuit and ran the tail resistor and cap to ground and got the same readings. This is a real puzzler.

[Structo]

Measure the value of your grid resistors on the power tubes.
Reflow all solder joints to and from the PI tube.

Measure caps from the last stage to the PI for leakage.
Same for the couplers to the power stage.

What amp is it anyway?

[andrew]

Structo, love that avatar. Looks like something I probably did as a child. I get 46v at the 220k/ cap junction going to the power tube and after the 5.6k grid stopper at the power tube. I resoldered the joints also. Here are more readings on the 12ax7: pin2-16v pin7-19v pin3,8-24
This was a Bluesbreaker style I am converting to plexi style.

[andrew]

I found it. It was the input cap to the PI. Thank you all for your time and advice.