KT88 filament failure - twice on same socket

[rps944]

OK, I did a few searches and if this has been covered already, my apologies - maybe my kewords were wrong. Anyway....

The amp in question is a Rogue 88 (not Magnum build). I've had two filament failures on the same KT88 position/socket, running intriode if that matters. Bought the amp used and the first failure was after one month. I retubed the entire thing and all was fine for another month and one night I hear popping. Inspection shows the same tube position is out. After the failure, fireworks begin in the other 88s (there are four of them) with seemingly random selection. The plate fuse (1/4 amp slow-blo) also blows, sometimes the one of the filament failure side and sometimes both. A couple of the 88s also heat up to the point of the silver top getting dark - ouch. I've also replaced the crispy output resistors to the fuses (100 ohm, 1/2w but replaced with 1w). The problems start once I replace the fuses (guess that makes sense?). Without the plate connected, things are seemingly OK.

Before I fry more of my money, am I missing something simple like a bad socket/finger connection? FYI, I've replaced the bypass caps (don't recall value) and the resistors measure OK. Also, I've replaced the current regulator in the PS (LM317 if I remember?) and the PS seems to be OK. The B+ runs higher than the schematic notes but I attribute this to the 122vac in my house. I tend to overanalyze (EE!!) so I could use a clean perspective. Any advice on modifications to improve the stability of the circuit type is also welcome, but I've already read some good suggestions on the basics.

Many thanks for the anticipated good advice!!
Bob

[SY]

If the voltage at the heater pins is 6.3 or thereabouts, I'd immediately suspect the socket. A few rapid on-off cycles at a bad time could do the trick.

[wboyd]

Have you contacted Rogue Audio....I would try that before frying anymore tubes. KT88's and 6550's can get expensive (as you may well know).

Also as Sy suggested...check the socket...it may be bad.

Wayne

[Wavebourn]

Pin2 (filament) and pin3 (anode) are too close to each other. If an ark happens it may leave conductive path between pins. Or some wires may be soldered too close to each other...

[HollowState]

What kind of sockets are in your amp for the KT88s? If they are black, they are inferior for that tube. The socket is colored with "carbon black" which is a conductor, albeit a very high resistance one. The better sockets are brown mica filled or even better white ceramic.

edit: the black ones are also hygroscoipc.

[kimjand]

Originally posted by rps944
OK, I did a few searches and if this has been covered already, my apologies - maybe my kewords were wrong. Anyway....

The amp in question is a Rogue 88 (not Magnum build).

I have had a look at a few Rogue products. They seem to have a general problem with the heater supply... Either too high, or too low. A possible fix may be a mod back to AC-heating, at least for the finals.
Also, a delayed HT supply seems to be the norm. Unfortunately the bias supply is derived from the HT windings, which results in the finals taking a beating on each and every startup. Since the bias supply has a longer time-constant than the HT.
This may not apply to your particular model, but appears in other Rogue products.

[rps944]

Thanks everyone. All good suggestions. I intend to work on the amp tomorrow and will post my (hopefully successful) outcome.

The sockets are white porcelain and I will check for carbon tracks, as well as tighten up the connectors, resolder the board connections, and clean the connections in general.

I spoke to Mike at Rogue, he's a good guy and supplied me the schematic to work from. I'm betting on the socket as the filament supply is good and I seem to have ruled out everything else. I hate taking the "shotgun" approach as I always considered myself a good troubleshooter/tech. But I've learned over the years to step back and seek fresh advice - I rarely see the forest for the trees!!

Another question: I don't have access to a tube tester and since I intend to expand my tube-based components, any suggestions on a good and reasonably priced tester I can tinker with? As always, many thanks for the suggestions. I hope to return the favor soon!!
Bob

[keithgreenhalgh]

how about an intermittent short in the winding of the power transformer.When it heats up it puts a high voltage across your heater winding. This is just a random idea

[rps944]

Everyone,
Again, thanks for the suggestions. Looking at this amp again with "fresh" eyes, I quickly found the culprit: no surprise, a cathode bypass resistor opening up on the tube that was burning out. Either I missed it the first time through or wasn't seeing the meter straight.

After a few days, things are seemingly fine so I've got my fingers crossed.
Bob