Here’s another example of what’s happening. So to turn the amp on “naturally” (no light bulb current limiter in place) I have to pull two output tubes before starting the amp. Once the amp is on I can insert the remaining two output tubes and then I flip it off standby after they’ve warmed up and I watch the bias (the amp isn’t biased yet but the bias trimpot is set to the middle of its range). Here are some pics taken showing bias readings from both meters over a span of about 15 seconds after switching standby off before shutting the amp down:
As you can see the bias current is skyrocketing on one pair of output tubes while the other pair stays about where it should be. I duplicated this twice with the probes on different tubes to confirm the runaway bias is happening only on the same two tubes, V4 +V5.
One of the first things you should validate is that you get the right negative bias at both sides and on all 4 tubes on the 5th pin. do these tests with the tubes out, and on a limiter as well. I'd think at this point you need to ensure you're getting solid negative bias to all 4 tubes, if not that could be what is causing one side to go into runaway. Oddly on this amp per the schematic, they have 1.5k to the first grids, then they jumper those off to the next tube with another 1.5k, so in theory two tubes get 1.5k grid stoppers and 2 tubes get 3k grid stoppers... very weird but whatever. I'd make sure you can guarantee the power tubes have equal negative bias on both pairs. If not try and trace backwards in the circuit to the junction of R26 and R27 to see if it's good there, if so, maybe the 110k bias resistor on the other side is having issues. If one side has the tube going into runaway like that it seems like the negative bias isn't right on that side only. If you can get consistent negative bias on both, then keep an eye on the negative voltages at pin 5 on all with limiter on and with tubes in, to see if you see them dropping for some reason.
BTW You'll want the negative bias not mid way like you have it, but at it's maximum negative voltage so you can get the tubes biased really cold and slowly work up until you see them running away and you can quickly roll it back or shut down
You may have already cooked some of those power tubes if they redplated.
Okay so then at that point, you're getting a good negative bias to the tubes, the schematic says to set it about -38 to -40 VDC at the point you're seeing -47 so you're in a safe bias point. It would be interesting to measure the 150k and 1.5k resistors on the side that's trying to go into runaway at cold temperatures and then right after you have to shut it down, I'm wondering if one of those resistors, when it heats up, is doing something bad. If not, what other components on that side could cause the tubes to go into runaway?
pompeiisneaks wrote: ↑Sun Mar 14, 2021 11:48 pm If not, what other components on that side could cause the tubes to go into runaway?
Leaky coupling cap.
i.e. positive DC leaking from B+ on the previous stage? Oh that would be bad since the bias point is supposed to be negative.... yeah that's definitely something else you need to look for.
I put the tubes back in with the bias pot still at max negative voltage. With the amp on a current limiter I’m getting about 7mA on both sides but after a few minutes the problematic side slowly rises to about 10mA and stays there, I suspect because there’s not enough current for it to keep climbing? I think with the current limiter in place I can’t get a normal bias current flow, the bias is just really cold so I don’t have an accurate idea of how the amp is actually running unless I take it off the limiter but then I have to do the dog-and-pony show I previously described of popping tubes in while on in order have the amp plugged straight into a wall without blowing the fuse. Sometimes the amp will turn on without the limiter and NOT blow the fuse but not always and I don’t have a lot of fuses on hand so I’m forced to use the current limiter a lot.
I will give those resistors a look before and after warm up to see if their values change or something else weird is going on.
I just got the amp turned on with no current limiter and no probes, bias as low as possible. After about less than 2 minutes of being fully on, V4 and V5 actually redplated before blowing the fuse. This was my last fuse so I’ll have to go buy more before I can continue with this thing.
At least I’m sure it’s not the transformers now so that’s good news.