2266c problem, looking for some guidance

[jskeen]

Ok guys, I got roped into trying to deal with this issue after another tech gave up on it, and I don't know the whole history. And Second, I'm not an Amp tech. I am an EE with 20+ years in electronics and medical instrument service, mostly dealing with Diagnostic Ultrasound systems. Don't exactly see many tubes in those anymore. But I'm doing the best I can with this one.

I was told that "it blows fuses after a while" probably referring to the t500ma HT fuse, but not sure of that. It came to me minus the cap to the bayonet fuse holder, so rather than replacing the soldered on board unit, I used a cooper bulkhead mount and soldered in some leads to the board with spade connectors. Easier to put an ammeter in series with, and the added bonus of not breaking off the board if somebody drops an fx rack on it or something like that. That was the easy part.

Hooked the speakers up, put in the correct fuse, turned down the master volume and was letting it warm up to check the bias, when I noticed some smoke trying to sneak out of the base of the v3 socket on the preamp board! Shut everything back down, checked fuses, pulled the power tubes and tried one more time. Definitely getting hot on pins 4-5 and 9 of v3, with maybe a little extra current getting pulled through pin 9 of v4 as well. Let it bleed down, pulled the 12ax7's out and measured pin 4 to 5. v 1, 2, and 4 read 33 to 37 ohms, V3 read 11. New tube on order. Got that far OK.

So, after I get the new v3 installed, and make sure the bias on the kt88's is between 43 and 48mv, and nothing else is trying to smoke, what else should I check before I turn the volume up a little and see how she runs? Current through FS2 seems like a good thing to watch. Anything else? If all looks good, plug in and run it for a while, then recheck bias? What else?

Thanks for the responses
James in Houston

[g1]

The preamp tube pins you are describing are the heater pins, so you can expect them to be hot. They all sound to be within the range of normal resistance.
I think you can continue without replacing V3.
Monitor the bias, you may find it drops out after awhile, or you may find one of the power tubes goes faulty after awhile.
Did the tech who did the previous work replace the power tubes?

[g1]

Oops. I missed the part about the smoke coming from V3 base.
Check for poor solder connections at those socket pins 4,5,& 9. Resolder as required. It sounds like there may be a poor solder connection that is arcing.

[g1]

schematics

[jskeen]

Thanks for the reply. Couple of things. I know that pins 4, 9, and 5 are the negative, center tap and positive terminals of the filament heater, but i'm not talking a little warm to the touch here, I'm talking Smoke coming off of the pcb as the varnish burns off the trace, within a few seconds of powering the main switch on. Seems that the difference between a normal current of 180 to 210 ma for a filament that reads 30 to 40 ohms at 6.3v vs over half an amp for a resistance of 11 ohms might be a bit outside normal?

I think I would prefer to end up with a spare preamp tube I really didn't need than to spend half an hour replacing a burned off trace.

Assuming that replacing the tube and or resoldering the socket pins successfully keeps all the valuable smoke in the preamp section, I can move on to the power amp, and check the bias. Assuming it is in range to start, I'll let the amp idle at it normal 8ohm load and recheck it every few minutes. If It stays stable for a half hour or so i'll walk the master volume up a little at a time and see what it does.

I'm kind of guessing here, but possible outcomes, if the bias voltage starts to sag and the current through the HT fuse climbs till it pops, that would tell me that the problem is something in the bias circut, as opposed to a bad valve?

If so, how might a bad valve present itself?

Thanks again
James in Houston

The preamp tube pins you are describing are the heater pins, so you can expect them to be hot. They all sound to be within the range of normal resistance.
I think you can continue without replacing V3.
Monitor the bias, you may find it drops out after awhile, or you may find one of the power tubes goes faulty after awhile.
Did the tech who did the previous work replace the power tubes?

[g1]

Yes, you are correct that it is better to replace the tube than repair a burnt trace. I was not aware the traces were burning up, apologies for the bum steer.
I've never seen a heater fault like that and thought the resistance differences may be due to cold vs. hot, I have no idea how a heater can decrease it's resistance with aging.
But this amp must all be looked at as unknown due to the previous repair attempts. So perhaps V3 is a new tube bad out of box. Anyway, best to be wary of any socket related shorts so you may want to check with no tube in V3 position that there is zero current through those traces.
As far as the bias for the power tubes, the only faults that could make the bias voltage sag would be a bad bias filter cap, and the current is so low in the bias circuit that you would get excess ripple on the bias even with the tubes out. If there is some bias drop-out issue I would expect it to be more likely caused by an intermittent connection, or iffy solder joint.
A bad power tube will probably start to run away current wise, and it's hard to say whether it is the fault of tube or bias circuit without having other tubes to try or another amp to try the tubes in.

[52 Bill]

I know that pins 4, 9, and 5 are the negative, center tap and positive terminals of the filament heater, but i'm not talking a little warm to the touch here, I'm talking Smoke coming off of the pcb as the varnish burns off the trace, within a few seconds of powering the main switch on.

I'm a little confused here. V3 and V4 heaters are wired for 6.3vac, with pins 4&5 connected together as one side of the filament and pin 9 as the common second side. Both tubes are wired in parallel so either one can be removed without turning off the other, unlike V1 and V2 that run two twelve volt filaments in series. Please explain what negative, center tap and positive terminals are.

As G1 suggested, pull V3 and measure the ac filament voltage at pins 9 and 4/5. Watch the traces for any signs of heating. The ac filament voltage for V3 and V4 is connected by one connector (CON12), which would make for an easy spot to measure current draw of the filaments. You can also pull V3 and then install V4 in its place to see what happens.

[jskeen]

Snipped the relevant portions of the schematics. I was incorrect in detail but still managed to detect the problem. There is in fact a center tap on the secondary side of the transformer section that powers the filaments, but is tied to ground, and the voltages used at the tubes are -3.15 to +3.15, as opposed to +6.3 to ground.

In any case, my problem is that the factory Marshall ECC-83 tube in the v3 position has somehow shorted a portion of the filament such that the total dc resistance of the filament, as read from pin 4 to 9 is 5.7ohms, pin 5 to 9 is 5.9 ohms, and inclusive from 4 to 5 is 11.1 ohms. No other pins read conductive to the filament as well, so the mechanism of this failure is something of a puzzle. This is pretty far outside the normal value of 15 to 20 each side, and 30 to 40 total. I know that dc resistance is not necessarily equal to active impedance under load, but as a troubleshooting guide, it seems prudent to replace that tube before reapplying power. Again, I don't know what was done previously, nor do I have a coherent view of the problems over time, so I am forced to troubleshoot in stages as issues present themselves.

If anyone is interested in performing a postmortem exam of the tube in question, I would be happy to make it available. :-)

James in Houston

[g1]

The point about the transformer heater winding is that it is AC, not DC, so there is no + or -.
The center tap is connected to ground as a ground reference, but the tube is only concerned with the 6.3VAC across the whole winding.
Tubes can not run on 3V heaters. The 2 heater connections that come from CON12 must have 6.3VAC between them.

Bill's point about trying V4 over in the V3 spot while you wait for your tube is a good point. If there are any other issues related to the V3 heater connections, you can work on them in the meantime.

[jskeen]

All true granted. May try that when I put the amp back up on the bench. Playing with this at the moment.

[g1]

Playing with this at the moment.
[ATTACH=CONFIG]36114[/ATTACH]

You gonna find a way to use those lovely GEC's in there?

[jskeen]

well, not directly, no. :-)

But after I get the strat and the marshall both put back together and working, I plan to do some EXTENSIVE performance testing as a pair before they go back home to their respective owners.

[jskeen]

Just wanted to loop back on this one and update the thread. I replaced the original marshall ECC-83 in v3 with a groove tube GT-12ax7-M, powered with the gcc's out, and watched everything for a while, All the smoke stayed right where it was supposed to. powered down, waited, put the GEC's back in and lit it back off. Watched everything for a while and all looked good. walked the master volume up a bit at a time for about 5 minutes, everything looked fine. Turned back down to one, plugged in my newly shielded strat, and proceeded to make a little noise while watching for signs of anything getting hot. checked the current through the HT fuse, about 270ma. Everything seemed happy. Proceeded to seriously annoy the squirrels and birds outside the shop with my abysmal playing. for about an hour, everything stayed kosher. turned down to 1 and went to lunch. Came back and checked everything over again, still good. Bias balanced at 45 and 46ma. Played around with it a couple more hours, till I noticed my ears were ringing, and called it a day. Next morning I looked everything over carefully, checked s few voltages and went ahead and put the chassis back in the cabinet. I rolled the preamp valves around over the next couple days and decided I liked the mullard in the v4 and the sovtek that was there in v3 with the original marshalls in v1 and v2 best. Little warmer and thicker on the LP but still not icepicky on the strat even intentionally misadjusted. The owner played his first gig last night with his baby back and called me at 11:30 to tell me how happy he was. Ya hoo. I gave him the bad tube (very well labeled as BAD) but now I wish I had kept it for autopsy, just to scratch my curiosity bump. C'est la vie.