AT100 blowing random power resistors

[bay4hotrod]

Hey fellas, you've helped me fix this amp before - wondering if you'd like to go another round? I'm not an amp tech so I regret not being able to contribute more; just a player that likes to tinker!

This amp sounds so good, I hate to part ways, but I'm starting to have second thoughts. Amp Kat, among others suggested hard-wiring all the board connections in lieu of the crummy push-ons... that is starting to appeal!

Here's the chain of events: this amp has been really humming for the last year until last week when I kicked on the power after warm-up and purple sparks were jumping around in PT V7 or 8. Kick myself for not noting.
So I found R105 burnt.. Ordered a pack of spec resistors and replaced. warm-up/power up sparks in 2 power tubes. R107 and R104 burnt.

Now... you would think I would have taken a moment of clarity to start testing .here. nope. replaced those resistors, same power up, same purple sparks in V7, blown R107.

Moment of clarity has arrived. I remember Enzo saying something about pulling PT's with lights off and check for arcing, then measuring B+ voltages. Pin 4 is obvious B+, what is normal on what pins?
Amp is fitted with originals except power section are Svetlana (winged C). Would like to go to full set JJ's, but was wondering how to figure out if a bad tube(s) are causing the trouble, or something else is causing the pretty light show.
Also, something weird: none of the replacement resistors are blowing- just moving down the line of originals. well the last original just blew- still not feeling lucky!

I'm running out of resistors, so I thought better to ask somebody that can point me in the right direction.

Any help with thanks!

[bay4hotrod]

not to bump my own thread, I just like talking to myself............but in the automotive industry where I (belong), diodes are commonly used to protect the pcm from feedback and can be quickly tested with an ohm meter. I noticed that all the 1N4004 diodes have a reading of roughly 375 while D7A and D6A have infinite readings on all scales. This is an active project and am replacing R107 right now....

[Chuck H]

Those all look to be screen grid resistors. Obviously R107 is the one with the most current across it. It may be telling though that the other screen resistors went also. One bad tube would have seemed likely if not for that. As it is I'll speculate that the power tubes are drawing excessive screen current because of something other than a gross tube failure. That isn't to say that you may not have bad tubes though. Possibilities I would check first would be whether the amp is being played into a proper load (or plugged into a load properly).?. And is the bias supply functioning and adjusted correctly? I remain hopeful that your OT is alright. With all of the screen grids drawing excessive current it could be a suspect.

EDIT: the amp looks to be a combination bias. That is, cathode/grid voltage combined. So the grid will sit at a fixed voltage and the cathode will somewhat self adjust the circumstances. THIS ONLY APPLIES IF YOU USE THE EXACT SAME TUBES THE AMP WAS DESIGNED WITH!!! Power tubes of different models and from different manufacturers have widely variable bias requirements. There IS an adjust pot on the fixed grid voltage supply. Whether that is enough to compensate for different tubes and makers I can't say without having the amp on the bench. It does appear that there is some bias type switching. Also some heavy use of switching jacks for the load connection. There may be some problems with some of these switching functions. Also, you mentioned excessive hum for an extended time. This could be indicative of an imbalance on the OT due to a power tube failure since a lot of hum is intended to be cancelled there if the tubes are somewhat balanced. Continually operating a bad power tube may have been taxing other circuits, like the bias supply, and resulted in an operating condition that has since compromised other power tubes and culminating in the condition you're dealing with now.

[bay4hotrod]

8 ohm output into mesa recto 8 ohm input for load. .. . resistor replaced. wow I can't tell my soldering from the factory. that only says one thing: the factory needs some more practice lol! I have the option of moving suspect tube into different location and blowing another resistor, or using known failing set of Sovtek 5881's.......hmmm. should power up w/o power tubes and check voltage first?

[bay4hotrod]

pin 4 voltage: 485 across the board

[bay4hotrod]

substituted Sovtek 5881. no flaring but one is dead and minimal output... what is the advice for temporarily mixing a 6L6 in with 3 5881's?

[Chuck H]

substituted Sovtek 5881. no flaring but one is dead and minimal output... what is the advice for temporarily mixing a 6L6 in with 3 5881's?

Don't

Do pull power tubes to check the power tube grid voltages and the cathode resistance. Don't neglect to check these in all switch settings. I hope the "mesa recto 8 ohm input" means that you're plugging an actual speaker into the "out jack" and not JR7 rather than plugging the amp into another amplifier?!? (Why would Mesa apply the moniker "Rectifier" to a cabinet??? Not that they didn't. I'm just not familiar with it.) It might also be a good idea to build a light bulb current limiter at this time (I'd cover it but you can look it up easier). I'd just order a quad of the OEM power tubes now if I were you. I think you'll need them before this is over.

[Enzo]

When I see sparking INSIDE the tube, I generally consider that the kiss of death - bad tube.

[Chuck H]

When I see sparking INSIDE the tube, I generally consider that the kiss of death - bad tube.

I think that's likely too. But I'm also considering that one bad tube may have caused another failure in a circuit since all the other screen resistors have popped too. Or, if all the screens are seemingly compromised, why?

[Silvertone Jockey]

When I see sparking INSIDE the tube, I generally consider that the kiss of death - bad tube.

Definitely. Grid, and or screen grid wires will be melted. Always discard any tube that has been arcing internally

[Enzo]

Chuck, I agree other damage may have happened, but as soon as I see arcs inside tubes, I get rid of them before going any further.

[bay4hotrod]

Don't

I hope the "mesa recto 8 ohm input" means that you're plugging an actual speaker into the "out jack" and not JR7 rather than plugging the amp into another amplifier?!?

lol, well I suppose that's a legitimate question isn't it? stupid is as stupid does!.. but yes Mesa slant 4x12, 8 out to 8 in. I realize that mixing tube types for any length of time is about a half dozen cherios short of a full box, but for a
3- second test it works just fine. Filaments on, grid stable, no lightning, but the output was severely muted.

I might be rather embarrassed if this just turns out to be a foul 6L6. What bums me is this set of PT's only have about 10 or so hours. The sparking and the blown resistor happened in less than 5 seconds the very first time. What bothers me is there were no indicators, visually or soundwise to alert me of imminent failure and/or damage. New tubes in the works. I live 100+ miles from the nearest tube mart so it'll be a minute.

I'm still interested if anyone knows about the diodes 6A and 7A I mentioned before?

Chuck, I don't remember any humming, certainly not for extended periods - maybe from another post?

Thanks to Enzo and Silvertone for your comments too.

[The Dude]

Diodes 6A and 7A are flyback diodes for high voltage protection. Unless they are shorted or leaky, not a problem.

[Chuck H]

Chuck, I don't remember any humming, certainly not for extended periods - maybe from another post?

Yes. It is from another post. That's the second time in a week I've done that!?! I must be tired because I'm not drinking any more than usual

[The Dude]

......I'm not drinking any more than usual

Maybe you should.