Even Rubycon's can fail...

[Goonda]

A board arrived yesterday with a solitary bad cap. It was a Dell, S478 board. To my surprise, it was a Rubycon MCZ that had failed.

http://img187.imageshack.us/img187/7913/rubyconfailure001lg1.jpg


http://img299.imageshack.us/img299/8888/rubyconfailure002gr3.jpg

So I fired up the soldering Iron and replaced it with a 1500uF, 16V UCC- LXZ. I didn't have a 6.3V in hand so I just had to make do with what was available. I'm not sure why the Rubycon failed? Possibilities include random chance, it being under-rated i.e needed a 16V instead of a 6.3V or just a bad batch.

I exclusively use UCC/NCC in all my recap jobs (PSU's, MB's other devices..etc) and they have never failed me till date.


http://img45.imageshack.us/img45/8588/rubyconfailure003rr4.jpg


The astute reader would observe the snake like trace on the PCB. I suspect it has something to do with minimizing EMI along that trace. Never seen anything like that before. Have a snake like pattern is not generally good due to the additional stray capacitance it introduces, but I suppose that does not matter here...

[Shell]

Well that's a rarity, I've never seen a Ruby die without being killed.

Oh well, Ruby looks no worse in my eyes, I always found it strange that I've only come across a few dead ones out of several thousand.

[Goonda]

Same here. I have a few antique MB's laying about with Rubycons on them. Not a single one has any issues. Perhaps it is just a random incident. :shrug:

[akinsey]

I admit it, it was my board Goonda fixed. It must be a bad batch: this is the second Dell GX270 board I received with the very same blown capacitor.

If I misposted, sorry! :)

[Goonda]

AK has been instrumental in fueling my re-capping adventures. :clap:

The board has very high quality caps otherwise. An all Rubycon and Sanyo affair.

[bing]

Yep, just random incidents, few bad apples always exist everywhere in the bucnh of them, its just plain luck, but definately the luck is better with high quality brand. ;)

Wait until you measure their ESR at few of not so old mobos like 1-2 years, even they're good quality & brand, you will surprise especially at those used for cpu and ram vrm ! :D






.

[corpsejockey]

I see quite a few of these with the Dells. It's strange in that the caps near the dimms and the pci slots fail rather than the caps near the cpu which is usually more common. Could this be a design flaw?

[pby5cat]

The snaking traces are for signal timing, often done between the north bridge and the memory slots where timing is critical.

[bing]

That snake shape trace is common practice in reducing EMI residue (high freq ringing) from high speed switching signal or from power switching source.

Good reference reading -> HERE (http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1154,C1002,C1243, P1981,D11877)

[Goonda]

Actually pby is probably correct. It seems to be a timing thing. It could serve a dual purpose though. Provide precise timing and minimize cross-talk between critical traces. Modern MB's have precise digital control of timing so this may not be necessary. :shrug:

[El<(')>Maxi]

Nice, you should show the whole recapping procedure Goonda! Where do you source you caps? I've read som guys think it's not a good idea to stray too far out of spec when recapping or you may introduce problems in the power circut, personally I think thats a little over-blown. Lot's of guys go way up in uF with no problems at all.

[Goonda]

I'm using the same capacitance i.e. 1500uF, it looks bigger because it is rated for 16V as opposed to the 6.3V Rubycon that blew up.

Changing caps can be a dicey business depending on where they are located. You do not want to stray too much in general. Some locations are more sensitive to changes (like the CPU Power area or the buck regulator feedback loop) in capacitance. Bigger is not always better. More important parameters like ESR, ripple current rating and ESL need to be closely examined. If you are changing stuff dramatically without thinking, be prepared for problems. I've had devices where using ultra low ESR caps has actually introduced ringing/oscillation problems in the circuit. General rule of thumb would be to stick as closely as possible to the pre-existing design.

Chemicon, Panasonic FM, I buy at digikey.com. Rubycons are available at newark.com and SANYO-OSCON series can be purchased at Ten Four LTD.