Twin Reverb: Reverb driver tube shattered!

[Lambertini]

Hello!

During a rehearsal this week-end, after one hour of playing at a rather loud volume, the amp suddenly loudly crackled. I stopped it immediately, checked the tubes and noticed my lovely NOS GE 12AU7 has shattered all around. It worked just fine until now and hasn't been shocked at any time, handled with care at all time.
You can see below the unfortunate result.

Back home, I double checked the voltage of the tube plates. My PT is on the low side (330V at HT wires, instead of usual 340V) and V3 plates get 400V instead of the 440V noted on the schematics.

Did you already experience such a thing?
Could that be caused by heat on a fragile / defective tube? Or maybe 400V is definitely too much and I should use cheap tubes at this location?
Any other thoughts or suggestions?

Thanks!

[sluckey]

Replace with a tough 12AT7.

[Esquirefreak]

I'd check for shorts in the reverb transformer.

/Max

[tdvt]

It is my understanding that the reverb driver tubes in the late SF amps (which I think yours is?) run hotter as they have a different value cathode resistor than the earlier BF/SF & also current re-issues. 

[Lambertini]

Replace with a tough 12AT7.

I like the 12AU7 there, less gain, less agressive reverb, and I think a bit less noise.
I browsed several datasheets and didn't find tough 12AT7, specs were around 300V max on plates and 2.5W dissipation, while 12AU7 were around 300V and 2.75W.
BTW I still don't understand why the circuit is sending 400V to this tube.

It is my understanding that the reverb driver tubes in the late SF amps (which I think yours is?) run hotter as they have a different value cathode resistor than the earlier BF/SF & also current re-issues.

It was, until last week when I finished my AB763 build project ;) I'll post photos in a build report soon, nothing as clean as sluckey's builds but I did my best. Here's my layout, power wiring is missing but all the rest is what has been done.
I have 400V going to V3 plates, and the cathode resistor is 2K2 with a 25uF 25V cap in parallel.
I'll double check the resistor in case of anything failed since I soldered it.

[Dave]

You might try a 12BH7 as a reverb driver. I tend to like it. I think it makes a little smoother reverb than a T7 and is probably a little more robust than a U7.


Dave

[mresistor]

Maybe run a militarized 6189W instead   https://www.ebay.com/itm/395206746733

[mresistor]

Here's three of them for not much more money


https://www.ebay.com/itm/375224515851

[tubeswell]

Replace with a tough 12AT7.

I like the 12AU7 there, less gain, less agressive reverb, and I think a bit less noise.

If you're running a 12AU7 as the reverb driver tube, you'll likely need to alter the bias resistor, as it's a hotter tube than a 12AT7.

[Lambertini]

I did the following measurements, using a spare Tung Sol 12AU7, max plate voltage 300V, dissipation 5.5W (both triodes).

V3 plate voltage: 404V
V3 cathode voltage: 20V
V3 plate-to-cathode: 383V
Cathode resistor: 2216 ohms

Here a screenshot from Rob Robinette calculator. Everything looks perfect. What could be wrong?
Can that just be bad luck or a defective old NOS tube?

[sluckey]

Stuff happens. Maybe too hot, vibrations from too loud, weak glass, combination of all. Who can say. It's not likely to reoccur. That tube does run very hot. May be a good idea to remove the shield or replace the shield with one of these heat dissipating shields...

[Lambertini]

Stuff happens. Maybe too hot, vibrations from too loud, weak glass, combination of all. Who can say. It's not likely to reoccur. That tube does run very hot. May be a good idea to remove the shield or replace the shield with one of these heat dissipating shields...

Will do. Thank you! This reassures me, I'd like to avoid the replacement tube bursting in turn.