... all the new resistors kept burning out so i just ran the amp on two power tubes instead of four and it works. ....
my next plan of action is to start testing caps correct?
i was actually able to watch the tube, and its flashing blue for a second when i have all 4 power tubes plugged in. almost like the tube is getting way too much voltage. explaining the screen resistor getting too much voltage, explaining its failure
so somewhere between the power supply and the power tube board, is a bad resistor or capacitor im guessing. sounds like a guessing game with the voltmeter is in session
No, don't waste time testing caps. Don't waste time guessing. Understand how the circuit works and what the resistor destruction tells you.
Current burns resistors, when there is much more than designed. Another poster said the resistors burned were 2
control grid resistors and a screen resistor. Unfortunately, the Crate schematic shows everything except the output tube sockets and the resistors attached to them, so I can't verify the above.
So the screen resistor for that (pair of?) tube(s) passed excessive current and burned. That could happen through normal use at full power output (during the instant when the plate current for the tube is at its peak) but is very unlikely because the designer would have taken care of that in the design, as it's "normal use". If the control grid was driven very, very positive then the resulting high peak currents could include a high enough screen current to burn a resistor. However, this too is very unlikely, as the circuit isn't capable of driving the control grid positive-enough in normal use to cause damage this way.
The 2 burned control grid resistors are
a clue. The control grid passes essentially zero current at all times under normal operation; there is some current in maybe the micro-ampere ange and would have to be a thousand times bigger to even register on your meter's most sensitive setting. So a current big enough to burn resistors is nearly impossible...
Unless... the screen voltage on pin 4 of the output tubes (about 450v) is able to come in contact with the control grid at pin 5 (and its bias of about -50v). If these two points came in direct contact, then the high current would flow through the screen and control grid resistors and probably burn them open before any serious damage occurred to the tubes. The resistors would have about 500v across the pair of them (450v + 50v) which could flow enough current to burn them open.
The output tube socket sits nice & high off the board. There's plenty of room for a bit of metal, loose solder blob or wire/component lead fragment to find its way over to the socket and touch pin 4 & pin 5. But that does seem like a one-time failure, given you don't see a bit of metal bridging those two pins in your picture, and the fault happened multiple times.
Turn the amp off, unplug it and measure the power supply voltage to be sure it has dropped to 0v across the board. Switch your meter to resistance mode (or continuity) and measure directly from the "bad tube socket" pin 5 to pin 4. Even if you get no continuity, re-check on resistance mode, as anything substantially below infinity Ω probably indicates a fault.
I suspect that the first occurrence happened when a bit of metal that you snipped (or was otherwise loose in the chassis) just happened to bridge those pins on the tube socket. The excessive current
may have damaged the board itself or burned a carbon track enable a recurring short-circuit, which could account for the follow-on burned resistors. Or maybe something else happened...
Either way, start your tests directly at the fault, looking for the most logical way for all clues to happen together.