[volt-nuts] volt-nuts Digest, Vol 7, Issue 9

[Dick Moore]

Alan, if it's possible to just leave the hole in the board, that may be the best solution overall. BTW, Nichicon has always had a very good reputation for quality electrolytics, as have the Japanese suppliers in general. The 2001 is certainly worth the effort to repair. I had a similar problem with a PCB in a Fluke 732A DC Reference. An epoxy coated Ta cap burned up and took a lot of the board with it and then a resistor burned as a result and took more board. I scraped and dug away the carbon and left the board as is -- of course this was not a multi-layer board. Most instruments don't need the sealer protection once in a lab in relatively dry, warm air where they'll be used for a long time. I wouldn't be sure that using the automotive epoxy would be a good idea...

Dick Moore


On Mar 17, 2010, at 5:00 AM, volt-nuts-request at febo.com wrote:

> 
> Message: 6
> Date: Wed, 17 Mar 2010 11:20:37 -0000
> From: "Alan Scrimgeour" <scrimgap at blueyonder.co.uk>
> Subject: Re: [volt-nuts] Keithley 2001 Multimeter Fault - Update
> To: <jfor at quik.com>,	"Discussion of precise voltage measurement"
> 	<volt-nuts at febo.com>
> Message-ID: <AE27F40615CE4CA08494C3AD153F4414 at AlanPC>
> Content-Type: text/plain; format=flowed; charset="iso-8859-1";
> 	reply-type=original
> 
> Perhaps only the copper was 'etched', but I can imagine even epoxy being 
> attacked by hot electrolysis taking place right next to it. Electrolysis 
> could produce some very active compounds, depending on the electrolyte (and, 
> I'm not sure, but in such close proximity to the electrodes there may also 
> be some extremely reactive short lived species, or is that just Sci-Fi?). 
> But by some mechanism the upper layer of glass fiber in the board was 
> visible before I started digging it away.
> 
> The excavation is progressing with care. In the centre the damage has gone 
> right through the board to virtually the other side. Presumably air cooling 
> stopped or slowed the progress of the damage actually through the very last 
> layers of the board.  It appears that once the hot electrolysis had begun to 
> damage the pcb it carbonised and in that conductive state, drew current and 
> generated yet more heat leading to a chain reaction in the form of a growing 
> carbonised region.
> 
> I've just had to cut a wide buried copper track in order to be able to 
> remove the carbonised pcb beneath it, which is disconcerting, but it will 
> just need soldering, or replacing with a piece of wire. I'm more worried 
> about what to use as a 'filling' in this cavity. I said I'd use epoxy resin, 
> but the usual stuff is damaged by soldering temperatures. I have some 'Auto 
> Weld' which says it's resistant to a constant 300C and should do. Once I 
> fill that hole back I'll never get it out again, so I'd better fix it 
> properly!
> I'd like to add that I'm feeling pretty angry about those electrolytic 
> capacitors. They are sheer vandalism! Time for some companies heads to bow 
> down and appologise! Those unstable low dropout regulators are another 
> annoying self destruct mechanism too!!!
> 
> Alan