Rubycon bad cap series..

[4tified]

I apologize if there is another thread dedicated to this topic, but my question is about the series number on the side of the caps. For example, I have a busted Rubycon MCZ cap with a series number of "T0343", but on another board the same age, there is a cap labeled "T0524" in the same area that is in perfect condition. ( I haven't measured resistance as I don't have an ohm-meter to do so.) Both caps are 16v, 1800uf. I don't want to replace all of the caps of this size on the board if they are good, so is there a way to tell by the series? (or manufacturing number, whicever it is.) Can I assume that a higher series means that they fixed the bad electrolytic flaw at some point during the manufacturing process? Any help is appreciated, I don't have the money to buy an ohmmeter right now to test every cap above 1000uf on the board.

P.S. - These are Dell SX270 boards I'm working with.

[willawake]

Re: Rubycon bad cap series..

i havent seen any total failure of MCZ on a board to indicate they had some manufacturing problem like the nichicon HN series. for instance the HN of a certain date code make failure of many or all caps on the board.

but there are some boards like the Dell SX270 where one or two MCZ commonly fail. leading me to believe that they are caps which are sensitive to an unknown parameter, perhaps heat and ripple seen on this board. certainly their water based electrolyte aspect is commonly criticised.

T0343 is read simply as 2003 week 43

so there is not any information to assist you i am afraid. personally i would replace the failed MCZ only and see how that goes. however do you have any nichicon HN on that board which should definitely all go..

[4tified]

Re: Rubycon bad cap series..

So perhaps the boards themeselves are flawed due to bad design. Even the caps I replaced two years ago with MBZ serires on an older SX270 are starting to rupture now.

[bgavin]

Re: Rubycon bad cap series..

Bad power supply is more likely.

[Topcat]

Re: Rubycon bad cap series..

Welcome to the world!! No cap is indestructible. Even Rubycon's will fail if ran hot enough, I see blown MCZ's in SX270's and SX280's all the time. They simply overheat. The GX270 breathes a lot better, but a blown MCZ is not unheard of in one, I see them once in a while. Make sure the system has adequate airflow. If it does, you either got fakes or did something wrong. Even crap caps will last 2 years.

[gdement]

Re: Rubycon bad cap series..

Based on all the reports of similar issues on those Dell boards, it does seem to be a design flaw. Lots of people have had those caps go bad, yet they're known to be good caps on everything else.

Some caps are more durable than others, and I would imagine that the more aqueous caps like MCZ's are probably more sensitive to overheating. Simplest fix is to just replace the caps and plan on them going bad again in a couple years.

For a more permanent fix, there's a couple ways you could go. You could try a more conservative and less aqueous cap like ZL series, but caps like that don't meet the performance specs of the originals so they might not run very well.

The other, more ideal option would be to use solid polymers. But figuring out the correct polymer substitute isn't entirely straightforward. I think there's been some people who have worked this out for the Dell boards in question but I don't know the layout myself.

[kc8adu]

Re: Rubycon bad cap series..

i only see blown mcz on dell boards.
i have yet to see a blown mbz on anything.
as for a fix for these cap cookers i use polymer caps in the vrm out and vdimm spots.
i did some testing with 820@2.5 in both places since i have a ton of these caps and spare boards.
this eliminated some buggyness in a 270 that was being used to run a cnc truss saw.
this has worked out so well that boards of this type get polys.
look at https://www.badcaps.net/forum/showthread.php?t=4969
i consider this mod to be ready for widespread use.

[spiff59]

Re: Rubycon bad cap series..

Hah! I'm back from last January and still whining about my dying SX270 SFF PC's. I'd begged a batch of mobo's out of Dell prior to their January bad-cap-replacement cutoff and am just now running out of spares. kc8, you'd mentioned you were going to try a polymer fix, I'm off to read the thread you referenced!!!

PS - Seeing just one or two erupted MCZ's in these Vesuvius motherboards is the exception for me, the attached photo I put in this January thread shows *15* toasted Ruby MCZ's: https://www.badcaps.net/forum/showth...3&page=3&pp=20
The board I pulled yesterday only had 9 crusty MCZ's.

[linuxguru]

Re: Rubycon bad cap series..

Maybe Rubycon outsourced certain batches (to Indonesia?). But then again, any cap can fail in a hot Prescott VRM environment.

[kc8adu]

Re: Rubycon bad cap series..

these boxes are simply little blast furnaces.
polys are the only thing that will last longterm.
hell i am now seeing panasonic fj and fl blowing up in the boards dell replaced in a local school.
i am getting over 100 to poly.
told the i.t guy there that every last one of them will need to be recapped.he didnt believe me till he opened some at random.
he called back and said "oh shit!"
he could not find any without at least 1 bulging cap!
explains all the bsod issues.