I realize I should probably outline my thought process a little bit, as you're probably getting annoyed with my constant requests for voltages.
If we just consider the 12AU7 on its own with no 6080, we know you have 1.5V of cathode bias on both channels (which is good), 0V on the grids (which is also good), and then we can look at the voltage drop across the 22K plate load resistors for each half of the 12AU7.
1
On the T1/T2 side, you have 172V-88V=84V, this 84V is dropped across 22.1K, so Ohm's law says there's 3.8mA of current rolling through there. This is all OK.
On the T4/T5 side, you have 172V-118V=54V, across 22.1K, this is 2.4mA of current, which isn't particularly stellar.
Since this happens with the 6080 out of the tube socket, we can rule out tons and tons of random miswires, tube weirdness, and other issues that could take a few days of troubleshooting, and we can basically focus just on the 9 pin socket, and what might be happening there.
The last test, which should be nice and conclusive, would be to pop the red wires of T1 and T5, and wire them to the opposite terminals. These wires might not be long enough to reach, but you can temporarily tack in a jumper to make the connection, then see if the flaky voltage moves over to T1. If it does, the 12AU7 is just behaving a little weakly, and you can continue using it or send it in for an exchange. If the flaky voltage remains on T5, then there's likely a cold joint on that front 5-lug strip.
-PB
