For those that swear by KZG... :(

[gg1978]

Re: For those that swear by KZG...

Well, i guess the only caps with the tristar vent i've seen fail are Samyoung then..

But the caps in Topcat's photos are defiantely UCC, as they don't have the dot in the center.. Whether they're real UCC or fake we can't be sure yet.. It's hard to tell from the picture the surface "shinyness" of the top of the caps.. The laptop in the other thread appeared to have the real UCC shield logo, but the vent cuts looked deeper and wider..

[Per Hansson]

Re: For those that swear by KZG...

Hmm, I've got a DFI Expert nForce4 mobo, it's using UCC KZG caps for most stuff...

I'm watercooling this thing so temp around VRM section etc gets pretty high... Will be interesting to see how she looks in another 6 months of use...

Note that it does use a few OS-Cons too...

[Chris1992]

Re: For those that swear by KZG...

w/r/t the crown logo, I have seen many Trigem boards with dotted-logo caps. They are identical to known genuine Chemicons in all other respects.

[trodas]

Re: For those that swear by KZG...

The second failure Topcat showed us looked exactly like mine failure (black dot in the middle):

Only far, far worser. In my case, the machine run fanless watercooled for about year and so at 200x13 - 2600Mhz with 1.800Vcore and folding all the time - except for gaming
Now my PSU was a Antec NEO 480W at that time, full of Fuhjyyu (surprisingly there is Rubycon in the input!) - so no wonder that these caps right on the 12V rail from PSU powering the Vcore give up sooner or later

I bet even a good cap could give up sooner or later in such enviroment (hot GFX card right bellow, very hot mosfets right on top and Fuhjyyu caps delivering enough ripple right on the PSU...) - but server board with lot's of them and on Zippy PSU - I can't believe how seriously these caps give up

It should be investigated with great care if they aren't fakes, Topcat. Could you please look?

[999999999]

Re: For those that swear by KZG...

> there is a dot in the center of the shield logo, while a KZG from a dead MSI Nforce 4 board has no dot.

I have here a Power General, non-PC-use switching supply with two sizes of UCC caps, 330uF/35V and 1000uF/35V. Below I elaborate, but this PSU has ran fine, no cap problems.

> KZG that appears to be real has relatively shallow relief for the tristar pattern on the top of the cap.. The "fake" LXY has significantly deeper cuts, and appear to be a little wider too..

Yup, I noticed this on a suspect KMG also. The vent cuts are wider and deeper, like they've been stamped. The real ones look much finer, with barely any signs of indentation.

The 1000uF on the PG supply are finer, barely indented and shiney topped. The 330uF are wider and deeper, duller on top. Point being, I would expect both were sourced direct from UCC, both real. Neither has any dot in the logo. The 1000uF are 12.5mm dia and 330uF are 10mm dia.

> the real ones have the aluminum case relatively "shiny" if held up to the light, while the fakes are more "dull" looking.

Yes, this is accurate - the real ones have a mirror-like aluminium finish (sometimes stained by atmospheric oxidation, but still very smooth). The suspect fakes have a duller matt finish.

These shiney tops are somewhat interesting to me, I'd wondered if they clear-coat them, but with dull probes I still get electrical conduction. Bare aluminum certainly should have oxidized long before now so it suggests to me it might be an unusual alloy.

From a spec sheet (page 3)
http://download.siliconexpert.com/pd.../LXY_Dec02.pdf
It appears the rubber plug pattern is different based on whether it's <12.5mm diameter or >= 12.5mm dia. That is how all the LXY I have here are, too, even the really old ones I'm about to mention...

Anyone know how to decipher the date codes on the LXY? I came across a bag full (of really old ones?) the other day, but I don't know just how old. The bag had an in-house sticker on it printed in 2001, but these have a blue sleeve instead of brown and I'd thought LXY went to brown sleeve before 2000, but I could be wrong. The only markings are:
(M)105C
1 (2)
7W

[linuxguru]

Re: For those that swear by KZG...

>
> It appears the rubber plug pattern is different based on whether it's <12.5mm
> diameter or >= 12.5mm dia.

Thanks for the datasheet - you're right about the pattern, both the recessed '-' and 'T' are known-valid patterns on the bottom plug, depending on the diameter. I look for a relatively wide recessed channel between the leads, as well as the embossed little triangle, delta or 'v' mark midway between the leads - it's always there on the good caps (perhaps the fakes will eventually get this right also).

[Rainbow]

Re: For those that swear by KZG...

The one in the middle is a bit bulged

Here you can compare it with a good one on the left

[toastygoodness]

Re: For those that swear by KZG...

Wait... since when were KZGs green? I always thought UCC's colors were Brown or Black or polymer. Man... I really want to swear by KZGs, but this thread is preventing me from doing so. Guess I have to use other UCC caps.

[linuxguru]

Re: For those that swear by KZG...

That's a KZE (not KZG), and I don't think there's any reason for concern. The bulge is barely detectable, and it must be a heat/ripple-current issue, probably compounded by a bad PSU. It can happen with any cap that is subjected to excess stress. In my experience, KZE is great in VRMs/PSUs in spite of being aqueous.

[PeteS in CA]

Re: For those that swear by KZG...

I've seen green KZEs, also brown. And I've seen LXYs and LXZs in blue as well as brown.

[kc8adu]

Re: For those that swear by KZG...

the lxz i have in stock are blue.

[Big Pope]

Re: For those that swear by KZG...

the lxz i have in stock are blue.

Yes, like this

[gdement]

Re: For those that swear by KZG...

Any update on the KZJ's? Is this issue isolated to the KZG series, or has anybody noticed this problem on the KZJ as well?

[bgavin]

Re: For those that swear by KZG...

I just finished recapping my new MSI 6670E (K7N2 Delta2-LSR). I replaced all the KZG with MCZ.

This board is running the Sempron 3300+ which is an OEM Fujitsu, I think. Clock is 2200 MHz, 400 FSB. I have two gigs of Corsair XMS in it now, and it passes all Memtest86+ tests, including Test #5.

[yanz]

Re: For those that swear by KZG...

kzg is done. i dont think it should ever be used for mobo..

this is a bulged kzg from a p3 mobo (shuttle mv19)with good ventilation :

[PCBONEZ]

Re: For those that swear by KZG...

Any update on the KZJ's? Is this issue isolated to the KZG series, or has anybody noticed this problem on the KZJ as well?

KZG and KZJ use the same electrolyte so expect the same problems.
-
KZJ is still uncommon and that's probably the reason for few reports.

[ant3202]

Re: For those that swear by KZG...

KZJ suffer a bad fate too

[japlytic]

Re: For those that swear by KZG...

Even the best electrolytics are not immune from failure... they may have been underrated (not in the data sheet) for the application...

[PCBONEZ]

Re: For those that swear by KZG...

There are very few rated higher that KZJ.

[kc8adu]

Re: For those that swear by KZG...

KZJ suffer a bad fate too

looks like the ost which are a common bad cap went and those kzj got left holding down all the ripple.likely that any cap in that situation would pop.