Amp - Arc/Short to ground plane....

[karbomusic]

My beloved Egnator Rebel 30 died two years ago this month. Finally cracked it open this week. R51 was getting very hot which is part of the power to the plates of the preamp tubes. The other symptom was the preamp was not working but the power amp was. C43 looks like a smoothing cap that connects to R51 and appears to be the power for the plates of the preamp tubes. I removed it and 'thought' it had leaked electrolyte and went closed but the meter showed it was OK (at least at meter voltages).

I received replacement caps yesterday and as I was cleaning up the PCB, I noticed it wasn't electrolyte at all, it was an arc/short from the positive leg of C43 to the ground plane. This seems to explain all my symptoms including R51 heating up. My question is, that is a 1mm gap, is there any way possible way this could have arced on its own minus any other help such as foreign material etc.? Or is there some other root cause I'm completely overlooking here?



Schematic: http://bmamps.com/Schematics/Egnater/egnater_rebel30_sch.pdf

[thermionix]

Not trying to insult anyone's gear, but THIS is why tube amps shouldn't be built with PCBs.  I think you have your amp figured out.  It would be nice if you could scrape that ground plane back away from the B+ points with an exacto or something.

[vigilante397]

My question is, that is a 1mm gap, is there any way possible way this could have arced on its own minus any other help such as foreign material etc.? Or is there some other root cause I'm completely overlooking here?

Not likely. Air has a dielectric constant of 3MV/m, which means you would theoretically need 3000V to arc across 1mm.

If there was a previous arc or overheating that caused some charring that could be a contributing factor. I worked on an amp a while ago that had a melted trace from a power surge, so I jumpered it with a wire but it kept melting. I found out (from the brilliant people of this forum) that the charred silicon of the board made a great conductor, allowing it to continue to short to ground. If there's anything that looks like it might be burned or charred anywhere, clean it up as best you can.

[thermionix]

Even common household dust is primarily carbon-based, and can be somewhat conductive.  Regardless of what helped it along, you know you did get an arc.  Nathan is correct, you have to scrape away all burnt board material, as that is also a carbon-based conductor.

I find it silly, though, that they ran the ground plane that close to the HV.  All risk, no benefit.

[karbomusic]

Not trying to insult anyone's gear, but THIS is why tube amps shouldn't be built with PCBs.  I think you have your amp figured out.  It would be nice if you could scrape that ground plane back away from the B+ points with an exacto or something.

Not taken as in insult at all. I was thinking similar when I found the problem. I actually plan on adding another 2mm to that gap by removing some ground plane around it. It isn't needed there and there is plenty of real estate to do it. Also, I'll be cleaning all that residue/carbon out. Resistance shows infinite at the moment but I'm not taking any chances and will make sure it is 100% removed.

Not likely. Air has a dielectric constant of 3MV/m, which means you would theoretically need 3000V to arc across 1mm.

Thanks, I didn't think it would be remotely possible via air alone but didn't know the formula and wanted to check. I'll hope to complete the repair by replacing these then biasing this evening. I'll report back with a happy face or a picture of a plate of crow.

Thanks again to both of you.

[PRR]

That didn't arc from preamp voltage.

I don't think it arced at all.

I think it is just crappy PCB stuff. Expect future problems.

[karbomusic]

Expect future problems.

I do but at least I know more than I did before if it happens again.

There was some substance there around the whole area of the cap but not sure what it was; not the glue but a brown substance that I thought might be electrolyte but the cap seemed fine and not leaking. Anyway, I just finished up the repair, double checked the bias while I had it open and all good. Works great!

Thanks for the input guys.

[greaser_au]

I'd have to suggest that this possibly started with a scratch on the PCB between the pad and the ground plane.

If it was me I'd: 
- Drill/grind the PCB to remove ALL of the carbonised material, grinding back a bit the ground plane to have a look underneath . Remove the sharp-edged remnants of the pad (I assume this is the component side of the board).
- Clean the area with solvent e.g. pure alcohol (isopropyl/denatured) or  lighter fluid, and allow to completely dry.
- Fill the drilled area, and seal the scratches, and where the pad used to be  with a thin film of epoxy (24-hour if possible) and cure it under a desk lamp (with a filament bulb - assuming you still have such a beast) you're aiming for 60 deg C  (140F). for a few hours.

david

[karbomusic]

I cleaned it all up and removed an additional 2mm of ground plane around that pad and replaced all the caps in that section with Nichicon. I've gotten in 3-4 hours of heavy testing/playing on it since the repair and so far it's great. I still think this may be a one-off because although I can find various reliability reports on the net, I haven't found anything that made me think this particular root cause is common in this amp but could be wrong.

I can make lots of guesses but still don't know 'exactly' why it shorted. When it originally died, I had been testing an OD, then walked back into to the other room to do some tweaks and left the amp on (not in standby) for about 90 minutes and while sitting there this occurred; when I came back into the room I could smell the damage and it was no longer working.

Either way, I have the schematic from Egnater and I know/understand the amp better now that I've worked on it so hopefully I'm armed with the goods to keep it working.  I did send the picture above with explanation to the manufacturer as an FYI. It's really an outstanding sounding amp so I'm stoked to have it working again.

[vigilante397]

I cleaned it all up and removed an additional 2mm of ground plane around that pad and replaced all the caps in that section with Nichicon. I've gotten in 3-4 hours of heavy testing/playing on it since the repair and so far it's great. I still think this may be a one-off because although I can find various reliability reports on the net, I haven't found anything that made me think this particular root cause is common in this amp but could be wrong.

I can make lots of guesses but still don't know 'exactly' why it shorted. When it originally died, I had been testing an OD, then walked back into to the other room to do some tweaks and left the amp on (not in standby) for about 90 minutes and while sitting there this occurred; when I came back into the room I could smell the damage and it was no longer working.

Either way, I have the schematic from Egnater and I know/understand the amp better now that I've worked on it so hopefully I'm armed with the goods to keep it working.  I did send the picture above with explanation to the manufacturer as an FYI. It's really an outstanding sounding amp so I'm stoked to have it working again.

Awesome, glad you got it sorted Also glad it wasn't a bigger problem like a transformer!

Semi off-topic: I've never tried one of these, how does it sound?

[karbomusic]

Awesome, glad you got it sorted Also glad it wasn't a bigger problem like a transformer!

Semi off-topic: I've never tried one of these, how does it sound?

Hmm... maybe a boutique fender deluxe or tweed type sound but really depends on the settings/channel used. It can also sound a little vox or marshallish as well lots of tonal choices to be had. Basically it can do wonderful cleans but breaks up really well and screams with nice harmonics if asked to do so. A couple features that make it stand out are..

1. Two power tube output sections allowing phase aligned blending between EL84s and/or 6V6 power tubes.
2. Separate 1-30 watt adjustments per channel.
3. Takes pedals really well.

I bought it on an impulse trying to replace my Class A Mesa Maverick that I sold years ago, this does a really great job at replacing it and then some; it just sounds sweet and classy. Other than potential reliability issues, big thumbs up. Every time I gig with it people comment on it.

[J0K3RX]

I cleaned it all up and removed an additional 2mm of ground plane around that pad and replaced all the caps in that section with Nichicon. I've gotten in 3-4 hours of heavy testing/playing on it since the repair and so far it's great. I still think this may be a one-off because although I can find various reliability reports on the net, I haven't found anything that made me think this particular root cause is common in this amp but could be wrong.

I can make lots of guesses but still don't know 'exactly' why it shorted. When it originally died, I had been testing an OD, then walked back into to the other room to do some tweaks and left the amp on (not in standby) for about 90 minutes and while sitting there this occurred; when I came back into the room I could smell the damage and it was no longer working.

Either way, I have the schematic from Egnater and I know/understand the amp better now that I've worked on it so hopefully I'm armed with the goods to keep it working.  I did send the picture above with explanation to the manufacturer as an FYI. It's really an outstanding sounding amp so I'm stoked to have it working again.

Don't know if you have a facebook account but, if you do.. you might consider sending a PM to Bruce He is an awesome resource and he is a subject matter expert on all levels! He is also a super nice guy and often assists people with their amp woes regardless if the amp is Egnater or some other brand... He's a very active member of the "Tube Amp Builders" fb group and he is even a member of the "DIY Stompboxes (Unofficial)" group.   

https://www.facebook.com/bruce.egnater

https://www.facebook.com/groups/tubeampbuilders/

https://www.facebook.com/groups/6110761372/

[karbomusic]

That's a good idea Jim. Thanks!

[tubegeek]

Yeah based on seeing how he interacts with that Facebook group, he'd want to know about your problem and see the pics. Just to stave off trouble in the future. He likes his equipment to be trouble-free.

[karbomusic]

Thanks, so I sent Bruce a message with the pic. Here is a different view. From an academic standpoint, see all the brown stuff around the entirety of the cap area? What is that? Is it possible it could be electrolyte even though the cap reads OK on a meter? Seems like it could be a clue as I'm not entirely convinced the PCB suddenly decided to conduct at that level all by itself.

[thermionix]

Took me a minute to figure out what I was looking at, but this was a radial cap and the white stuff is just adhesive to hold it to the board, right?  I would doubt the brown ring is electrolyte, probably just funk gathered over time from dust, smoke, etc.  Maybe the adhesive hepled it stick around the perimeter of the cap. Also, when I've seen caps leaking goo, it's usually white or light gray, not brown.

[karbomusic]

Thanks, white = glue

I removed multiple caps and only that one had the brown substance - the rest were perfectly clean other than the white glue.

[thermionix]

The brown stuff could also be rust/corrosion related to the metal case of the capacitor, which could also be related to an electrolyte leak.  I just don't think it's electrolyte itself, because of the color.  You replaced the cap already, so you're good there.

[karbomusic]

Root cause still had academic interest nagging at me so I had a few free moments last night for some additional post-mortem investigation. Unless you guys tell me differently, I'm leaning towards the issue beginning with the cap. None of the other caps were in this condition. I suppose this could be a chicken/egg problem but thought I'd share a few macro shots just for fun. I do not think the brown stuff around the lead is flux FYI as there was no other evidence of flux on the entirety of the PCB.




Scraped on it a little bit...



Zoomed in..





Exposing a littler further by cutting away some of it...

[PRR]

> leaning towards the issue beginning with the cap.

I lean that way too.

It can't be electrical arcing, there's not enuff voltage at a peramp B+ node. (If there were, you'd have damage all along the B+ line.)

Caps are strong chemicals at low-bid prices. Sometimes a bit of "snot" gets in the vat. Chlorides do wicked things to e-caps. Other contaminants do too. I have never seen quite that type of corruption on an e-cap. However the strangest things happen. A hard-drive (or chips?) factory was built way out of town where land and labor was cheap. Initial production worked great. Then they had a month of defective drives (chips?). By the time they got the troubleshooters on site, the drives were running fine again. Then some time later, another long run of rejects. Someone finally stepped outside and realized the factory was in farm-land, the bad months were times when the farmers were spraying, and if you didn't know that you would never suspect.