Scary Trend of Rubycon MBZ failures - No bloating!!

[kc8adu]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

those tiny desktops have been killing mbz a long time.
saw the first 3+ years ago.
never thought to check the ones that did not bulge.
i simply replaced all without hesitation.figured if the ones in the hottest spots gave up the rest were beat too.
will autopsy the next ones.

[Topcat]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

^
To answer the previous questions in the thread though, was there anything that lead you to believe these MBZ's were fakes>? I found absolutely zero evidence of that.

[RJARRRPCGP]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

Yea, but Lenovo is a *Chinese* company... not Taiwanese.

Yes, I agree that I would rather see stuff from ROC than PRC!

Makes me miss my high school years, when there was more stuff made in ROC than PRC.

[kc8adu]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

only thing i see is a datecode of 0716.
were these old boxes in production that late?
was anyone even producing 478 boards in 07?

[mockingbird]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

That's not a 478 board... It's an LGA775 with either a 915 or 945 chipset.

[c_hegge]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

Test the seal on those bad caps. Apply too much voltage to burn the cap up. If the seal is good it will bloat and explode like they do for c_hegge the capacitor seal clubber. Test many burning them out slower and slower. They should all bloat if the seal is good.

If c_hegge used voltage instead of fire he would be performing valuable capacitor seal research and having fun at the same time.

Good idea. Do I get bonus points for testing caps without vent stamps?

[PeteS in CA]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

Good idea. Do I get bonus points for testing caps without vent stamps?

The lead engineering tech at Boschert showed me where the manufacturing test area had once been. The acoustic ceiling tiles had a bunch of round holes in them ...

[kc8adu]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

now i see that.still get both types in over here.
one of my biker friends has one dead i need to fix.will post results of the cap test.

That's not a 478 board... It's an LGA775 with either a 915 or 945 chipset.

[Krendoshazin]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

I've pulled a load of 2200uf caps from Dell 745s; most of them were Rubycon capacitors dated '06 and '07. I decided to test them for the sake of curiosity and most of them were around 500uf. Many of them had vented or had some electrolyte crust, but some of them still appeared to be normal.

[UserXP]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

I would also like to see the inners of those caps. I hope it will not be one of those horrifying cases when a cheap under-rated cap is packed in a larger can and then wrapped with a Rubycon sleeve - thus appearing genuine.

[kc8adu]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

looking at that again the logo is way off!

If you saw my Panasonic FC counterfeit (Last one on the left), you would think twice before considering them to be bona fide... You underestimate the counterfeiters... I would love to see microscope photos of the foil etching on them... and even if that matched, who knows about the electrolyte (Though I have to admit, this scenario would be very unlikely).

[kc8adu]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

nothing worth mentioning on this box.
full of kzg.
recapped with mbz.i guess i will know in a few years.this one will be close to me into the forseeable future.
its on a dyno.not going away.sent the guy a link to this thread.

now i see that.still get both types in over here.
one of my biker friends has one dead i need to fix.will post results of the cap test.

[DakotaKid]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

^
You're reading too far into it. it's no secret that ibm and lenovo are pals.....but on that note, name me one motherboard maker that isn't chinese besides tyan and supermicro.... The point being, the caps are real... There is absolutely no mistake about that. The pics shown are extremely close up with a DSLR camera, it does make them look a little different that close up, but I assure you they're real. I honestly wish I could say they weren't, it would explain this whole strange thread.

IBM and Lenovo aren't really "pals" or "in cahoots". In fact, IBM sold their personal computer line around 2005 to a Chinese electronics company which renamed itself and its computers Lenovo. They acquired a lot of IBM assets and intellectual property in the acquisition and IBM received about 20% of the comany but immediately began selling it off and may not own much at all any more. So, while IBM may have had interest in Lenovo, the previous IBM brands like Thinkpad have transitioned to Chinese control of and manufacture by Lenovo. The fact that Lenovo is a Chinese company probably doesn't increase or decrease the likelihood of counterfeit caps being used, more a matter of supply chain practices and controls. However, to think that any company has enough power, clout, and direct factory connection to 100% ensure against counterfeits is pretty naive and short sighted. Counterfeits can easily be inserted into the supply chain and it is known that the same factories producing legitimate goods, whether purses or electronics, are sometimes run "after hours" to produce the much more profitable (to the maker) counterfeit goods, often made with inferior or diluted grades of raw materials. The possibility ALWAYS exists no matter how absolutely certain someone thinks they are about authenticity. If someone accepts a good logo or proper appearing bung as confirmation that the caps are not counterfeit then they are at best naive, at worst arrogant. I'm not applying this to anyone or any comment in particular, it's just smart to be cautious and a bit skeptical when dealing with a marketplace where there are known counterfeits. Don't assume anything until proven.

[Toasty]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

Good idea. Do I get bonus points for testing caps without vent stamps?

Only if you can get them to double-ricochet and land on the teacher's desk...


@Topcat - Autopsy...?

[Topcat]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

@Topcat - Autopsy...?

They're still sitting in the ash tray on the desk....I just haven't had a chance....but I'll get to it.

[bigbeark]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

I was in a BestBuy store recently and had a look inside a new Lenovo desktop.
Superficially they look like the old ThinkCentre desktops, but they've really cheaped out on the quality of the case and the motherboard was covered in crap caps.

The thing just reeked of cheapness. The old IBM boxes were solid, the new stuff is junk!

[ben7]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

The MCZ are also known to go bad. I actually had received an RROD Xbox one time, some of the MCZ in it were slightly bulging, but even replacing them didn't fix the problem, I just kinda gave up...

[mockingbird]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

MCZ in Xbox? You mean MFZ? If MFZ bloated, it means the guy tried the 'oven trick'.

[ben7]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

MCZ in Xbox? You mean MFZ? If MFZ bloated, it means the guy tried the 'oven trick'.

There was MFZ and MCZ in it. The MCZ were on the +12v.

At least the MBZ failures don't end in a huuuge explosion xD

[momaka]

Re: Scary Trend of Rubycon MBZ failures - No bloating!!

There was MFZ and MCZ in it. The MCZ were on the +12v.

Yes, the CPU and GPU VRM on Xbox 360s is usually Rubycon MFZ (either 820 uF in 8 mm or 2700uF in 10 mm, depending on what model 360 you have - i.e. HDMI or non-HDMI). Sometimes, they do use Nichicon as well - again, either 820 uF HN series in 8 mm or 2200 uF HZ series in 10 mm - but these are rare. Some HDMI versions have complete 2.5V 820 uF polymer coverage (United Chemicon PSC or equivalent).

As for the 16V 1500uF caps on the 12V rail - those are mixed. Normally Rubycon MCZ (both with gold and silver markings), Nichicon HN, Panasonic FL, and in rare cases United Chemicon KZJ and Sanyo WG.

However, mockingbird is right - if they are bulged, they probably have been subjected to too much heat likely from an attempted reflow. This isn't only caused by an oven reflow, though. They can get damaged with professional equipment too.

The old machine we have at work often would do that, especially if the techs we have forgot to cover them with Al tape.

The good news, however, is that if they are only slightly bulged (or at least not leaking), then they are actually fine. I tried pushing the top back down on several that were little to moderately bulged (but not leaking). Now, I know this raises a red flag for most people here, but that experiment turned some useful results:
On all of the bulged caps that I pushed the tops down, none of them bulged back again. Moreover, pushing the tops down was easy, meaning there was no pressure buildup inside the cap (compare to a cap that has failed... more on that in a bit). I did this for 2 Xbox 360 motherboards. While they have seen only intermittent use in the last 3-4 months, none of those caps have bulged again. Moreover, when I removed some of them to test on the ESR meter, they were all in spec for capacitance and ESR.

Now compare that to bulged caps that have failed. Last month I was building a linear DC-DC power supply but accidentally put one of those 16V 1500 uF caps backwards on my breadboard. Right within a few seconds it started bulging. Luckily I caught it immediately so I pulled it out before it vented. After letting it sit for a day, I decided to try and push the top back down on it but it was hard as a rock - an obvious sign that the cap had failed and pressure had built up inside the cap. So then I tried busting a small hole in the vent to really see if there was any pressure. Sure enough, a large amount of gas and electrolyte flew out when I did that (good thing I was wearing safety glasses cause some electrolyte did hit me in the face).

My guess as to why those reflow-overheated bulged caps were fine is most likely because when they are over-heated, their contents inside expand too much causing them to bulge slightly (and to vent if you really really overheat them - one of our techs did that multiple times while doing tests with the reflow machine on a random board). With the lack of current in the circuit however, the caps' electrolyte probably did not break down, thus the reason why there was no pressure buildup inside those caps. After cooling down, their contents likely went back to normal, thus allowing the top to be pushed back down.

Now I'm not saying those caps are 100% safe to use again - their seals on the bottom could have been compromised. But at least they haven't failed yet and their specs should still be good. As I mentioned the ones I tested were several months old and still fine. But I haven't tested any longer so I can't say how long they will last.

So anyways, that's the whole story so far. The sad thing is, that 16V 1500 uF cap I killed was a Rubycon MCZ .