Dell GX270 with bad batch of Rubycons or bad board design?

[Super Nade]

Well, a customer of mine contacted me about yet another GX270 with bad Rubycon MCZ near the VDRAM FET. She is afraid that almost all the stock of these boards are plagued by this. What gives? Bad batch of caps or just poor board design? Once she sends the board, I'm going to replace it with an OSCON SEPC 560uF after measuring the max potential between the legs does not exceed 4V.

This is the recap I did on one of the boards:
https://www.badcaps.net/forum/showthread.php?t=4412

I doubt it is a bad batch because ONLY that particular cap in that particular LOCATION blew up.

[gdement]

Re: Dell GX270 with bad batch of Rubycons or bad board design?

Are there any unpopulated cap locations in parallel with the one that fails? I'm guessing probably not, or Dell would have started using it. But if there is, it might help relieve some of the stress.

It would be interesting if somebody (probably with an oscilloscope) could figure out what the problem is with that cap. Maybe excessive ripple or bad voltage spikes or something. I don't have one of those boards but I find it interesting anyway.

[Super Nade]

Re: Dell GX270 with bad batch of Rubycons or bad board design?

I'll be getting a few boards to work on, so I'll try to provide some scope shots. Curiously, the boards seem to POST, enter Windows blah blah without any problems?

[gdement]

Re: Dell GX270 with bad batch of Rubycons or bad board design?

I'll be getting a few boards to work on, so I'll try to provide some scope shots. Curiously, the boards seem to POST, enter Windows blah blah without any problems?

Well, I'm no expert on analyzing such things, I'm just an interested observer. I've never even used an oscilloscope.

[syclonejim]

Re: Dell GX270 with bad batch of Rubycons or bad board design?

We have had 75 to 100 GX270 MBs replaced because of the failing capacitors. We are now starting to find replacement motherboards with the Rubycon 6.3v 1800uf MCZs blown. We also have found a few Nichion 1500uf 6.3v HN(M)s blown on these same motherboards. Since the Dell warranty has expired as of 1/31, I have been recapping the failed motherboards. I am starting to wonder if the cooling of the Dell cases are leading to failure of the 105c rated caps.

[willawake]

Re: Dell GX270 with bad batch of Rubycons or bad board design?

and the circuit design, psu etc.
nice to see you are dealing with it. got any pics of what you are doing?

[kc8adu]

Re: Dell GX270 with bad batch of Rubycons or bad board design?

i put 820@2.5 fujitsu re or ncc psa in all the vrm out and vdimm positions.
i have been using 2200@6.3 mbz at vdimm in +5
at vrm in +12 i use sanyo sep or ncc ps 330@16 and fill the 2 empty spots on the killerbee sff board.
i have never seen the 1800@16 mcz bad in several hundred of these i have rebuilt.
1 of the vdimm caps fails more often since it is near the(hot)vagp reg.

[kc8adu]

Re: Dell GX270 with bad batch of Rubycons or bad board design?

how to do it pics