[jcassity]

try something new...............

condition- unplugged and the loctal tube removed.
check to see if there is **ANY** resistance from any pin at the top side to chassis. If your in the same boat i was in,, my loctal was suffering from a minor carbon trace. or carbon strike,, whatever its called. i had no socket replacement so i proceeded against the advice i was given and pulled the socket out. i took it apart to reveal the two wafers . I laid the wafer with the obvious gully of a carbon strike down on different sand papers until i sanded way past the carbon trace. i put it back together and that solved that problem.
although its not typical ,,, i actually desoldered all terminations to this socket so i was sure my meter would read just the pins to chassis.

Check to see if the theory on my set holds true to yours.... see if you have an AC and DC voltage drop across that resistor. I could not actually detect it on the meter but,, it should be there invisible to your meter or not. A oscope would surely see it. Because of where it is actually living in the circuit, the thought on my set for this same problem resistor was that it not only has dc but also some ac on it,,,
I was able to gain control of how i calibrated it by observing my voltage at my 6k6 to get it below 300v or close to what the schematic says.