Mesa .50 caliber with red plates

[Rhodesplyr]

I've got a 1988 Mesa Boogie .50 Caliber amp in the shop with the owner complaining of red plates on the output tubes. He says someone else worked on it, but, though it will work fine for 10-15 minutes, they still go into runaway. I haven't gotten too far into it, but I suspect that the output tubes were damaged in the initial incident, probably a failed capacitor or bias problem, and now suffer grid emission as a result.

This is the second amp I've had arrive with these symptoms, and I wonder why more techs don't recommend replacing output tubes when they've gone cherry-red. I was under the impression that they almost always suffer permanent damage when this happens and should be replaced. My first experience of this was with a Fender Super Reverb; another tech replaced a coupling capacitor, but, sadly, the Sylvania 7581s were ruined and would no longer hold a stable bias.

The tubes I found in this amp are regular 2003 Sovtek EL84s, not EL84Ms. With 436V on the plates and 366V on the screens, according to the schematic, it seems to me like Mesa was tempting fate here. I'm assuming the original tubes were EL84Ms, although the only tube ever made in the 6BQ5 family that would really be rated for these voltages would be the now-rare 7189A. The owner innocently installed ordinary EL84s because the owner's manual does not stipulate that you need EL84M/7189 rated output tubes. Hmmmm....

I'm recommending to the owner that we go with 1980s production EL84Ms (I can't ever remember the Cyrillic designation) as replacements. I am, of course, going to do a careful diagnostic before installing new tubes. No reason to burn up another set.

[Robert M. Martinelli]

I'm recommending to the owner that we go with 1980s production EL84Ms (I can't ever remember the Cyrillic designation) as replacements. I am, of course, going to do a careful diagnostic before installing new tubes. No reason to burn up another set.

I'm getting old, so maybe my memory is failing me, but it seems to me the EL84/6BQ5 Russian replacement type is named 6P14P ( I seem to remember that, actually, with a max pd of 14 W and a max plate voltage of 440 V, they're beefier than EL84s and more similar to 7189s, so they should fit the bill quite well ).

HTH

Best regards

Bob

[Rhodesplyr]

You made me look it up :-)

6P14P-EV transliterated into Latin/English characters

6П14П-ЕВ in Cyrillic

Even for this tube, however, I think the official limit on the screen is 300V. I'm going to call a Russian friend and have him take a look at the datasheet for me.

[Enzo]

Don't obsess on the data sheets. If we did that we'd never plug another 6V6 into a Fender operating at 100v over its "maximum" plate voltage.

When the amp fails do a survey of power tube voltages. Are plate and screens about where they ought to be? IS the bias disappearing on either tube? DO new tubes make the difference?

[Rhodesplyr]

When the amp fails do a survey of power tube voltages. Are plate and screens about where they ought to be? IS the bias disappearing on either tube? DO new tubes make the difference?

Would a competent tube amp tech not do all those things? :-)

I don't swear by tube datasheets, but if there's a higher voltage rated tube for a circuit with high voltages for that type, I will admit than I am tempted to go that way for replacements. To be clear, I do NOT consider the tubes themselves the cause of the problem.

[Rhodesplyr]

Here's the tube data on the 6P14P-EV output tube, translated with the help of a native Muscovite. The datasheet calls it an "EL84 analog." It's the same tube--or at least the ancestor of--the EL84M as far as I know.

Max plate voltage at greater than 8 Watts dissipation: 300V
Max plate voltage at less than 8 Watts dissipation: 400V
Max plate voltage at cutoff: 500V
Max screen voltage: 300V
Max supply voltage: 500V
Max cathode current: 65mA
Max plate dissipation: 14W
Max screen dissipation: 2W
Max grid resistor: 1Meg (doesn't say fixed or cathode biased)

In part, this was just an exercise for me in seeing if I could read a Russian datasheet, but I thought it was interesting that they gave different max plate voltages based on dissipation. And something I'd never have picked up on was my friend's observation that the datasheet seemed to him to be written in an odd dialect. When I told him that the factory was in Saratov, he said that that explained why it sounded "provincial" to him.

[Enzo]

Would a competent tube amp tech not do all those things? :-)

Maybe he would. Thing is, we have the amp now, not him. One never knows who saw what and said what to whom about anything. I have a GK head here that someone told us had been to several techs and no one could fix it. I fired it up and some parks flew from the LM3886 IC on the heat sink. One new LM3886 and the amp works. So much for "can;t be fixed." To you and me, a "competent" tech is one thing, but out there in the real world, some kid who read half of a Gerald Weber book can make people think he knows amps inside and out.


Or am I looking past your point?

[Rhodesplyr]

I was referring to myself as the "competent tube amp tech." I've had a number of amps brought to me that had been to several techs, and, thus far, I've been able to fix them all.

I'm not assuming anything in this case, but I'm hypothesizing that the output tubes are damaged. The tendency to grid emission doesn't show up on most standard tube tests, at least in my experience. The 7581s in the Fender Super Reverb tested fine, but would not hold a stable bias. You could watch the grid emission creep up on a meter across a grid-stop resistor, then accelerate, even in a cathode-biased circuit. New tubes solved the problem.

[dai h.]

I think the original tubes were Sylvania 6BQ5. I have a .22+ and IIRC the plate voltage was 420V(screens much lower).

[Rhodesplyr]

The conclusion: the Sovtek EL84s were damaged and would no longer hold a stable bias. The grid emission problem followed one pair of the tubes in particular. I replaced them with a matched quartet of 1987 Russian 6P14P-EV, and, so far, all is well.

I checked everything out before putting new tubes in, and I found some really bad circuit board work done by someone before me. There is general darkening of the board around two of the output tube sockets, and another tech had replaced those two sockets. I'm not sure if it was from the overheated tubes or the repair, but some of the traces around one of the tube sockets were in very poor shape. I suspect it was from someone not being very good at desoldering components from a board with plated-through holes, a PITA, I admit. The soldering of the new sockets was dull, rough, and grainy. I used my vacuum desoldering setup to remove the bad work and resoldered them properly, and I added jumper wires to a couple of the questionable solder pads as insurance. This is one reason I don't like the practice of mounting output tubes on circuit boards.

This is my first Mesa Boogie amp repair, and, like others before me, I was cursing a chassis layout that requires you to desolder a bunch of wires to get access to the PCB. To quote Jack Nicholson's character from "The Witches of Eastwick," "A mistake? Or did he do it to us ON PURPOSE?" :-) It would certainly discourage casual attempts at modifications. I wonder--for factory repairs, does Mesa do repairs on the boards, or do they just drop in new circuit boards?

I think that this amp would benefit from some degree of adjustable bias to compensate for output tube variations and aging, but the owner's manual discusses replacing output tubes every six months. I guess the idea was to burn them up, throw them out, and buy new ones.

Speaking of tubes, the owner has Ruby-branded Chinese 12AX7s in it, and though they measure OK in a tester and in circuit, they make the amp sound dull and flat. I stuck some old stock Sylvania, GE, Mullard, etc... in it, and they livened it up considerably. What type of 12AX7s does Mesa typically sell as replacements for their amps these days? Two Mullard short-plate ECC83s as V1 and V2 sound really nice, but those are my own private stock.

David

[Enzo]

They probably sell whatever they can reliably get production quantities of. And it likely varies month to month.