ultra low esr replacements 1200uf

[momaka]

Re: ultra low esr replacements 1200uf

I haven't seen that happen with those KZJs yet but knowing KZJ it isn't impossible. Although I take it the ones that bulged in storage did so after being desoldered. goodpsusearch had an A0213 (Akita, Japan) dated 1000uF 10V Rubycon MBZ bulge in storage after being desoldered from a Dell Optiplex GX260 motherboard last year, after sitting in storage for four years.

Yes, my KZJs were desoldered as well.
And I had two 6.3V 1800 uF 8 mm MCZ bulge in storage too. But these were pulls from HP Compaq DC5000 SFF systems that put them through great thermal stress while in service. Many of these MCZs were already bulged when I started recapping those DC5000 SFF systems.

I don't think the desoldering is what causes the caps to bulged, though. Rather, I just think that these ultra-low ESR caps sitting in storage with no voltage applied is what causes the issues. That's why I went bananas this last winter/spring and decided to reform all of my capacitors in stock. For about 2/3 of them, I even decided to leave them charged as I put them in storage again, just to see if that helps vs. the caps that I discharged after restoring.

What were the date codes on those KZJs that bulged in storage?

Well, I could only find one of the failed ones... datecode: 5(2) 8L.

No, you're right, it does say something. Although HN certainly doesn't have a spotless track record but early HN (and HZ) datasheets rated those series for -40C rather than -25C (this was updated in August 2004 or thereabouts). So that suggests a change in the electrolytic chemistry. HMs were always rated for -40C.

Interesting.

I'm not trying to sound like a know-it-all. I just have this urge to correct misinformation

Well that's good.
It's okay to speak up (err. type up ) if you know you are right about something.

but I guess I'm not really any better what with my penchant for hijacking threads.

But we are on-topic here - we are discussing ultra-low ESR capacitors. I guess I stole that single-sentence being-on-topic from you here.

[Wester547]

Re: ultra low esr replacements 1200uf

Yes, my KZJs were desoldered as well.
And I had two 6.3V 1800 uF 8 mm MCZ bulge in storage too. But these were pulls from HP Compaq DC5000 SFF systems that put them through great thermal stress while in service. Many of these MCZs were already bulged when I started recapping those DC5000 SFF systems.

That's bad.

How long did it take those to bulge in storage after pulling them? Did they have 2004 datecodes?

I don't think the desoldering is what causes the caps to bulged, though. Rather, I just think that these ultra-low ESR caps sitting in storage with no voltage applied is what causes the issues. That's why I went bananas this last winter/spring and decided to reform all of my capacitors in stock. For about 2/3 of them, I even decided to leave them charged as I put them in storage again, just to see if that helps vs. the caps that I discharged after restoring.

As I understand it, leakage current increases after soldering and desoldering. So I thought it might have accelerated that failure. At least I've seen even older NOS MCZs and MBZs not bulge to this day. KZGs and KZJs are known to bulge in storage much more often than MBZ and MCZ and FJ and FL. Or so I thought. But now I am debating about whether I should recap my Xbox 360 Falcon which only has 1750 hours on it and 1500 power cycles according to the S.M.A.R.T. data from the Hitachi 60GB Travelstar 5K250 hard drive (but lots of time in standby). It has 1500uF 16V Rubycon MCZs with 2008 datecodes in the VRM in and 820uF 6.3V Rubycon MFZs and 820uF 2.5V Fujitsu polymers (2008 datecodes) in the VRM out. 100uF 16V Nichicon VZs on +5VSB, +3.3VSB, and +1.8VSB, and 220uF 10V Chemi-con KY and Nichicon HE otherwise (again 2008 datecodes). Not sure if it's worth it though, because the BGA RoHS GPU solder will probably fail first, and if not that the DVD laser optical pick-up.

I suppose those ultra-low ESR capacitors are formulated in such a way that they are somewhat unstable. Although I always thought Rubycons and Panasonics of that persuasion were at least -more- stable than Chemi-cons and early Nichicons of that ilk (and definitely more stable than Toshin Kogyo capacitors of that grade). I don't know about Sanyo/Suncon, maybe somewhere in between.

What I did notice is that all bad capacitors that bulge in storage have something in common - elevated H2O content. I would still think capacitors by good brands, even those ultra low ESR, would have special additives such as aromatic compounds and hydrogen absorbers to avert hydroxide molecules (the result of hydration and foil corrosion), the generation of gas, and excessive pressure build-up (although as I understand it, even those additives eventually get consumed or eventually lose their effectiveness).

Well, I could only find one of the failed ones... datecode: 5(2) 8L.

August 12th, 2005. I don't know what factory (2) denotes, although I'm guessing Miyagi, Japan. Behemot could clarify that.

But we are on-topic here - we are discussing ultra-low ESR capacitors. I guess I stole that single-sentence being-on-topic from you here.

I know you called me out for that first. I just thought it would make for a good slogan.

[halaster79]

Re: ultra low esr replacements 1200uf

are 2004 rubycon mbz caps probably bad?

[Wester547]

Re: ultra low esr replacements 1200uf

Rubycon MBZ isn't a bad series regardless of datecode. Although capacitors of that age may need reforming. I think I may have been jumping the gun earlier regarding the stability of MBZ and MCZ - I'd guess if they bulge in storage that the electrolyte was already starting to break down because of thermal duress.

[halaster79]

Re: ultra low esr replacements 1200uf

are 2009 rubycon mbz and mcz both black with greyish with - bar?

[stj]

Re: ultra low esr replacements 1200uf

the only time i saw a black mcz it had a suspicious "target" / concentric-rings type bung.

[Wester547]

Re: ultra low esr replacements 1200uf

The black / white MCZs and MBZs use a different material for the sleeve (PET or Polyethylene teraphthalate), as part of the RoHS standard (rather than PVC / Polyvinyl chloride). That said, MBZs or MCZs with concentric circles embossed into the bung instead of the regular convex standoff are definitely suspect.

[stj]

Re: ultra low esr replacements 1200uf

it was on an Intel Little-Falls2 motherboard!

[momaka]

Re: ultra low esr replacements 1200uf

the only time i saw a black mcz it had a suspicious "target" / concentric-rings type bung.

Black Rubycon MCZ do exist - again something I know from the Xbox 360
Even got more than a few in my stash. But they all do have a proper Rubycon -style bung, though.

That's bad.

How long did it take those to bulge in storage after pulling them? Did they have 2004 datecodes?

Well, I've had them since Fall of 2012 or thereabouts. Noticed they were bulged this past winter... so 3.5 years give or take.

As I understand it, leakage current increases after soldering and desoldering. So I thought it might have accelerated that failure.

Why would that be - the heat?

But now I am debating about whether I should recap my Xbox 360 Falcon which only has 1750 hours on it and 1500 power cycles...
It has 1500uF 16V Rubycon MCZs with 2008 datecodes in the VRM in and 820uF 6.3V Rubycon MFZs

Nah, that's a pretty new board. I don't recall getting any caps bulge from systems that new - not unless they were baked in an oven or improperly with a heat gun.
I think you should leave it as is. Like you said, the GPU BGA will likely fail before that. If you want to keep it working for as long as possible, ditch the stock GPU heatsink and put a CPU heatsink on it. But you will need to move the optical drive elsewhere (possibly doing lots of ghetto-modding).

What I did notice is that all bad capacitors that bulge in storage have something in common - elevated H2O content.

Yup, looks like it for sure.
I guess that supports my suspicion that ultra-low ESR caps are generally unstable.

[Wester547]

Re: ultra low esr replacements 1200uf

Well, I've had them since Fall of 2012 or thereabouts. Noticed they were bulged this past winter... so 3.5 years give or take.

I think the electrolyte was already starting to break down given that they were subjected to prodigious levels of heat. Same goes for the MBZs in goodpsusearch's old Optiplex. I've never seen a NOS MBZ or MCZ die in storage. IMHO, a truly decent cap would not bulge sitting in storage for 10-15 years.

Why would that be - the heat?

Yes. But then I realized that leakage current already increases as they sit in storage without any bias applied to the plates...

I think I should have rephrased what I said to elucidate. Really crappy brands seem to die at the slightest hint of heat (like Sacon) - so I wouldn't be surprised if the heat from soldering/desoldering alone shortens their life a bit (even if it only lasts for a few seconds - some brands are that bad!). Although for the "good" brands, it probably wouldn't make much of a difference if any.

Nah, that's a pretty new board. I don't recall getting any caps bulge from systems that new - not unless they were baked in an oven or improperly with a heat gun.
I think you should leave it as is. Like you said, the GPU BGA will likely fail before that. If you want to keep it working for as long as possible, ditch the stock GPU heatsink and put a CPU heatsink on it. But you will need to move the optical drive elsewhere (possibly doing lots of ghetto-modding).

Yes, those old 360s were designed really poorly on account of their cooling and ventilation.

Yup, looks like it for sure.
I guess that supports my suspicion that ultra-low ESR caps are generally unstable.

Only those from Chemi-con, early ones from Nichicon, some from Sanyo. I don't see too much evidence that those from Panasonic or Rubycon are unstable (although MCZs and MFZs are a bit too sensitive to heat). I suppose your opinion of Panasonic FJs and FLs wouldn't change if you saw them bulge in storage one of these days?