Fused Capacitor lead

[Gaz]

When you said you were using silicone on the caps I didn't realize it was across the terminals?. Maybe it is conducting? Another possibility is lower rated caps sealed in a larger rated shell?

Thanks for the reply, Joe. Yes, it is across the terminal, but the type I use is 100% silicone, called GE Silicone II, which contains no acetic acid, which is supposed to be what causes corrosion of copper in airtight enclosures before curing. If if it had acetic acid, the was more than 48 hours to cure. I can't find any evidence of arc'ing to the negative terminal or can through the insulation.

I tore up all the caps in the power supply, and tried to remove the silicone for evidence of any corrosion, but can't see anything. I would chalk it up to a bad cap, or mishandling when I bent the lead, but having it happen twice is just too bizarre.

I'm totally stumped...

[mhuss]

It won't help for previous builds, but what I have done on some builds when using radial caps is to mount them upright. Unfortunately, I don't have any under-board pics to illustrate, but even though I'm using mainly turrets, I put eyelets in the board to mount the caps. Insert cap, solder, then bend the cap leads over and connect to nearby dropping resistor holes. This avoids wasting the nice long leads, and soldering first ensures that the bending causes no physical strain on the cap itself.

[Enzo]

Shorting? Corrosion? Am I missing something? It looks like a cap lead broke off. Wouldn't be the first or last time that happens. Vibration is usually the cause. the burnt/arc evidence you see is probably caused AFTER the lead broke free, and the end touches and sparks, just like shorting a charged cap with a piece of wire.

[Mark Black]

Yes I once had a batch of Sprague Atom small electrolytic caps where the leads broke free on the positive end while just handling, so I know it can happen. Maybe Nichicon had a troublesome batch.

[Leo_Gnardo]

Yes I once had a batch of Sprague Atom small electrolytic caps where the leads broke free on the positive end while just handling, so I know it can happen.

Same here with F+T, dual caps where hardly any bending broke leads loose. Some broke loose while simply being stored in a drawer. Gave me a bad impression of an otherwise good cap company.

[Gaz]

Shorting? Corrosion? Am I missing something? It looks like a cap lead broke off. Wouldn't be the first or last time that happens. Vibration is usually the cause. the burnt/arc evidence you see is probably caused AFTER the lead broke free, and the end touches and sparks, just like shorting a charged cap with a piece of wire.

No, I totally agree, that's the logical explanation, but this is the 3rd amp this has happened on, and the first one it happened on two caps at the same time. That one had a different power supply. Every time it happens in the power tube power supply, even though I mount the preamp filter caps the same way. The ones in the power supply are usually siliconed to one another as well, whereas the preamp ones are just free standing.

[Gaz]

It won't help for previous builds, but what I have done on some builds when using radial caps is to mount them upright. Unfortunately, I don't have any under-board pics to illustrate, but even though I'm using mainly turrets, I put eyelets in the board to mount the caps. Insert cap, solder, then bend the cap leads over and connect to nearby dropping resistor holes. This avoids wasting the nice long leads, and soldering first ensures that the bending causes no physical strain on the cap itself.
[ATTACH=CONFIG]37555[/ATTACH]

Mark, that's great and what I'm definitely going to do. Reason I've avoided this in the past is because I've always built my amps so that everything is accessible from the top. I guess what I'm realizing is that there are compromises if one wants to do that. Any tips on wiring so that getting to the underside is not a total nightmare? I usually builde Marshall style, but sometimes like Soldano with tubes up front.

Thanks.

[mhuss]

Easy servicing is of course the compromise! :-/ This particular amp was a typical "tubes on one side, controls on the other" arrangement.

[TimmyP1955]

I have found several adhesives that had some degree of conductivity even at Ohm-meter voltages (at least when fresh out of the container). I don't recall if a silicon was one of them.