Where there's smoke...

[Keppy]

I've been lurking at this forum for several months, and I finally have a question!

Last time I used my Mesa Boogie DC-5, it started smoking.

History:
- Amp is all original except for tubes and power cable
- Twice, the speaker cable has broken at the speaker, causing me to play it unloaded (not at high output) until I found the problem. The first occasion was over a decade ago, the second was a few months before this problem occurred. The amp functioned normally for several months between the last instance and the current issue.

Symptoms:
- Smoke
- The amp was still working when I smelled the smoke
- I was playing at a very low volume

I feared for the output transformer until I opened it and found the cooked resistor in the picture below. The red wire connects the blown resistor to the reverb transformer. According to the schematic I found that is a 22k resistor connected to 450v.

My question is, should I just replace that resistor, or should I suspect that another component (such as the reverb driver tube or the reverb transformer) caused the failure?

[eleventeen]

Myself, I kind of doubt the rev tranny is bad...if it was, that resistor would probably be vaporized, not just charred.

I think what I would do is to measure that resistor in-circuit. Make sure the amp has not been turned OFF for .... a while. 5 minutes. That resistor need not be unsoldered to measure its value.

Resistors cook because too much current is flowing through them. Generally, those type (carbon comp) resistors tend to drift upwards, which would reduce current. But they can also drift down, and if it has drifted downwards, then too much current might be flowing through it. (Once it is charred, there's no telling what actual circuit value it might have) That could produce a kind of runaway condition and scorch the poor dude.

It's also possible that the resistor will just fall apart or break in half if and when you try to unsolder it. Take a probe and push on the side of that resistor. It might very well just crack in half. (You're going to have to unsolder & replace it anyway)

Another approach would be to just replace that resistor and run the amp for a while, see if it smokes. Resistors are pretty cheap. Out tranny = not involved as far as I can see.  

The answer to your question "should I just replace or should I suspect..." is "yes" (meaning both), but you say the amp is a over a decade old. I would have to suspect the .002 cap placed around that tube, as well. If that cap is leaky, plenty of undesired current (over and above the nominal tube current...which, since the drop across the 22K resistor is 34 volts is only 34/22000 = 1.5 mils which is why you can use only half a tube instead of Fender which uses both halves of a 12AT7 as rev driver) So while you are measuring the charred resistor, measure that cap (with ohmmeter) as well and consider replacing it on General Principles.

I think I would replace the resistor and the .002 cap, let the thing run, be unsurprised if the new R smokes as well, but let that happen before going nuts taking it apart. Think of it as a sacrificial resistor.

[PRR]

The 0.002u cap should be a 900V part.

That may not have been obvious to the designer. A "400V" cap might last a very long time before quitting.

I'd go ahead and replace the R, and the C, but with a series-stack of two 0.0047u 600V caps.

Look at the bottom of this page, Hoffman Catalog, Capacitors
0.0047 mallory 472 - 630 volt $1.00
0.0047 sprague 472 - 600 volt $1.00
0.0047 xicon 472 - 630 volt $0.75

Since this is not a tone-critical cap, I'd go with a pair of the 75-cent jobs.

Resistors are on a nearby page.

Power-up with the tube out. Resistor must not smoke. Tube plate should run near 450V. Amp should play just-fine (except no reverb of course). Then stuff the tube back. If it now smokes, bad tube. (413V is technically over the tube's 330V rating, yet they usually last a while, but it may have done its time.)

[Keppy]

Thanks for the help!

Myself, I kind of doubt the rev tranny is bad...if it was, that resistor would probably be vaporized, not just charred.

Yeah, I was wondering more if the tube would've caused it.

I think what I would do is to measure that resistor in-circuit.

21.2k. I was extremely surprised to find it still in spec!

Another approach would be to just replace that resistor and run the amp for a while, see if it smokes. Resistors are pretty cheap.

Yep, that's the plan... Unless someone says "NOOO!!! You'll burn out the [whatever]!!!"

The 0.002u cap should be a 900V part.

If I'm reading it right, it's a 3,000V part. (It says 3KV) However, measured in-circuit it reads -10.5nF. Is that revealing in any way, or just one of those weird inaccuracies that happens on the lowest capacitance scale on a DMM?

Power-up with the tube out. Resistor must not smoke. Tube plate should run near 450V. Amp should play just-fine (except no reverb of course). Then stuff the tube back. If it now smokes, bad tube. (413V is technically over the tube's 330V rating, yet they usually last a while, but it may have done its time.)

Thanks, that helps a ton. Sounds like I need a handful of resistors and caps to complete the repair, unless the tube's bad. I'm a noob as far as amp repair goes, and this thing is all-PCB with all sorts of pot & jack wires tethering it down front and back. It's a pain in the rear to work on, so having a clear plan with a few basic steps helps motivate me to get the thing done. I was considering recapping it as well, since it's from the early '90s, but I just measured all the large electros and they still read in spec on the meter. Given that, I think I'll tackle one thing at a time.

[Jack1962]

I have worked on a ton of Mesa's I doubt if that cap is back , if it is rated a 3K volts I know it's not bad . Replace the resistor and check it out more than likely it's your only problem , worst case , the tube may be bad as well.

[silverfox]

The 0.002u cap should be a 900V part.

 

I'm uncertain here as to why it needs to be rated higher, as in, I don't know? Is it due to the inductive spike from the transformer? Twice the 'B' supply?

Silverfox

[HotBluePlates]

... Is it due to the inductive spike from the transformer? Twice the 'B' supply? ...

Exactly.

You should assume the voltage where the transformer meets the tube plate will swing from 450v down to 0v and up to 900v. Yes, the tube won't be able to pull its plate all the way to 0v, and the inductance of the transformer may not kick all the way to 900v when the tube is shutting off, but the designer should assume it could for safety.

[silverfox]

Thanks HBP,

Several thoughts- (off topic) started to pop up- Basically RF Transmitters etc.. Ended up at a HAM site.


Silverfox.

[Keppy]

Based on the info here, I decided the tube was more likely to be bad than the cap. Replaced the tube and resistor and all is well so far. Thanks again!

Now I just need to scrape together the cash to BUILD an amp...