Red Plate 6V6 in Princeton Reverb.

[varaso]

Hi, I've got a Princeton reverb clone that I built and 2 times I had issues with super hot power tubes.

The first time that this happened, I thought it was the tube's fault because previously that day the tube had fallen down hard by accident. I replaced the tube and everything was fine for about a month. Yesterday while the amp was on for about an hour or more I saw the new tube glowing red hot. Today I opened the amp and it operated fine. The only issue is the current of the tube that overheated increased slowly up to the desired amount.

I bias the amp using the OT resistance and the voltage across the secondary ( B+ and pin 3 of the output tubes).
So I'll be refering to this voltage as the tube's voltage t1 or t2 and the sockets will be 1 and 2.
The OT resistance is 215Ω for s1 and 205 for s2. The t1 voltage used to be 4,3V giving us 20mA of current. The t2 was lower at about 3.8V. So today's issue was that the t2 voltage went from 3.3V to 3.8V very slowly (about 5-8minutes)

The issue appeared only in t2 that's biased colder in combarison to t1. I've checked many things but with no luck.
For today's test I changed the tubes' positions and now t1 was about 3.3V increasing slowly to 3.8 and t2 4.1 (again 20mA).

The issue must be after the bias circuit, so I checked PI coupling cap for leackage, changed grid and screen resistors on the socket, changed the 220k resistor from the bias to the coupling cap. I also inspected the socket and it seemed fine (it's a belton socket btw). I measured the components in the bias circuit and everything seemed normal. The bias pot also is a 10k and has a nice usable range. It's schematic is uncle doug's in youtube and will be attached bellow. The bias is somewhere near -39V.

Finally I got to the tremolo circuit and found out something odd. When I increased the intensity pot, the t1 and t2 voltages shifted and boosted that's normal but, the t1 voltage was boosted from 3.8 to something like 4.5 V and going back and forth between 3 - 4.5V with the pot halfway and when cranked it went rarely to 6V, while the t2 had a crazy range of 1V to 8V (up to 40mA) when the pot was cranked and 1 to 7V halfway. Then I recalled that yesterday I did had the tremolo on for a long time so I guess that's the problem.. That huge voltage range happens only in the socket 2 that the issue happened twice.

For a fix I tried a diode in the outer lugs of the intensity pot and the t1 and t2 voltages weren't boosted only decreased but the tremolo was weak and perhaps there was a little drop in volume and increase in the treble. It's an easy fix but I didn't liked it at all. The tremolo without the diode is nice and lush and rich I'd like to keep it that way.
 
I can't figure out why is this happening and want to make the amp stable.. It operates fine now but the red plate thing didn't happen without a reason.. The first time this happened, the amp was cranked, now it was with the volume at 0 or something like 2 when it starts working, and it waited for me to play it.

Sorry for the long post and thanks in advance!

[shooter]

I bias the amp using

I've found a 1ohm R from cathode to ground, and plate vdc to be the best way to know what current is flowing at idle.  once you start playing it becomes a dynamic exercise in futility.  Start with low current (cold bias) then sneak it up til you and the tubes are at a happy place.

[varaso]

Yes that's on the schedule for tomorrow..

so this is how the amp really works and I should just compromise? Shouldn't I try to find the cause of the red plate in the tube? Because it's not a constant condition, I can't really find what's causing it.. I just guess.. unless it happens again.. and as I said I didn't even play the amp when that happened.

I really like the sound at 70% Plate Dissipation, it's a nice warm sound.. If I could I'd like it at 80% but I know it's not recommended so I don't do it.. in less than that the amp is kind of weaker and more harsh in the treble with less mids and bass..

[shooter]

try to find the cause of the red plate

cause, (I would put north of 90%), current
too much volts would probably arc
that leaves a small margin for gremlins 

[varaso]

yes of course but if it's biased correctly that shouldn't happen.. I mean something's wrong if the tube goes over 90% when it shouldn't.

The 6v6 can handle 28mA max @425V (100%) and I bias them at 20mA (70%)

With my readings, when the tremolo is cranked and at it's slowest speed, the current to socket 1 goes from 17.6 to 26.5mA and that is within the range (but's its still a little too much) but at socket 2 goes from 7,8mA to 35,1mA which is WAY more than the tube can handle.. shouldn't these voltages be the same? Why do the differ so much?

[sluckey]

The issue appeared only in t2 that's biased colder in combarison to t1. I've checked many things but with no luck.
For today's test I changed the tubes' positions and now t1 was about 3.3V increasing slowly to 3.8 and t2 4.1 (again 20mA).

Since the voltage conditions follow the tube, I want to believe the issue is inside the tube, not in the external circuit.

I suggest you add another bias cap to the wiper of the bias pot. If you only want one bias cap, then move the existing cap to the wiper of the bias pot. Look at page 4 of this schematic...

     https://el34world.com/Hoffman/files/Hoffman_PrincetonReverb.pdf

[varaso]

Since the voltage conditions follow the tube, I want to believe the issue is inside the tube, not in the external circuit

Sorry I might not explained this right.. the huge voltage swing is always at the same socket, it doesn't follow the tube.. I guess the tubes are not ideally matched that's why there are different voltages. But whichever tube I put in that socket the voltage swing is crazy..

I suggest you add another bias cap to the wiper of the bias pot. If you only want one bias cap, then move the existing cap to the wiper of the bias pot. Look at page 4 of this schematic...

     https://el34world.com/Hoffman/files/Hoffman_PrincetonReverb.pdf

that's nice, I may try this tomorrow.. or on Monday ! but what exactly will I gain by doing this? stability?

[sluckey]

You will be able to dial the tremolo intensity to zero.