KZG, KZJ, -and- TMZ, TMV - Bad UCC series

[PCBONEZ]

[A little background for any cap newbies that wander in:]

Chemicon KZG & KZJ
These are well known to bloat and very often fail completely with no bloating at all.
A chemistry with the electrolyte formula used in these specific series is the suspected cause.
The Advertising [Announcement] that introduced KZJ mentioned it used [not quoting verbatim here] the same advanced electrolyte as their KZG series.

[This is the new info.]

Chemicon makes a TMZ series which is a custom/special order cap and not included in their catalog so there are no data sheets are available.
A few days ago someone asked me what the ESR on a TMZ was so I measured some TMZ.
Taking the can size into consideration they fall squarely between KZG and KZJ for ESR.
-
Presumably a cap with ESR between KZG and KZJ uses the same 'advanced electrolyte' as KZG and KZJ [or one very similar] and thus TMZ would be expected to have the same problems.
[TMZ -do- have some complaints here and there but the series isn't particularly common so complaints are far between in time. Rarity may be hiding the fact that they aren't so good of caps.]
-
There is another 'stealth' [undocumented] series named TMV but as yet no one has directly measured the ESR so the grade is unknown.
The TMV I've seen [online, not in person] are 680uF 4v used in VRM Vcore.
Their small size and application [where used] suggests their ESR would be near KZJ.
So it's -probable- TMV also shares the KZG/KZJ type electrolyte.

-
Incidently: I started researching this again when I found a leaking TMZ on an old but unused board dual P3 server board.
- No, that's not the one of those I measured.
- Caps I measured were unused NOS TMZ that had been reformed for an accurate measurement.

-
If someone has some TMV and an ESR meter I'd love to know what their ESR is.
.

[yyonline]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

You found TMZ to have lower ESR than KZG?

Awhile ago, I found 2 TMZ on some never used, but very dead Intel boards, and measured them. According to my readings, ESR was lower than KZE but higher than KZG, putting them about on-par with Panasonic FM. I got the identical readings on 2 TMZs. I compared them to readings with the same ESR meter of and the same can size (8x11.5mm) of MBZ and KZE/ZL to arrive at my conclusion.

Speaking of rare UCC series, I've also seen UCC KZV, but I've never found specs on it. Some of them were in rather hot Prescott P4 boards in 2004-05 Dells. They all passed the visual test, and the system was stable, so I didn't pull any to test or change them out.

[Scenic]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

i've got an abit IP35P with those TMV caps on the vcore (mixed with NCC polys IIRC), but i don't have an ESR meter
and as they're on huge ground traces, i probably won't be able to desolder them without damaging them :/

otherwise i'd send them out to someone with an ESR meter to check the specs..

[PCBONEZ]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

You found TMZ to have lower ESR than KZG?

Awhile ago, I found 2 TMZ on some never used, but very dead Intel boards, and measured them. According to my readings, ESR was lower than KZE but higher than KZG, putting them about on-par with Panasonic FM. I got the identical readings on 2 TMZs. I compared them to readings with the same ESR meter of and the same can size (8x11.5mm) of MBZ and KZE/ZL to arrive at my conclusion.

Speaking of rare UCC series, I've also seen UCC KZV, but I've never found specs on it. Some of them were in rather hot Prescott P4 boards in 2004-05 Dells. They all passed the visual test, and the system was stable, so I didn't pull any to test or change them out.

Mine was an unused NOS cap that had been reformed to correct for shelf-life issues. They actually sat for a little over a month after being reformed before it occurred to me to measure the ESR. That is about as close to having brand new caps to measure as one can get pulling caps from boards.

TMZ 8x11.5 (as read) ESR = 0.03
[between 0.025 and 0.035 ohms considering meter limitations]

KZG 8x11.5 ESR = 0.036
KZJ 8x11.5 ESR = 0.021

.

[Per Hansson]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

Do we have a general idea how long it takes for the KZG's to "overcook"?
I mean they are rated for 2000h at 105°C
Is it half that, a quarter or it differs allot from batch to batch?

Reason I ask is they are used in (very) expensive industrial systems and it would be nice to have a (very rough) idea of life expectancy...

[PCBONEZ]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

I don't have a reasonable guess and here's why.
-
When I get used boards I dunno how many actual hours are on them.
When I get new boards with KZG I either replace the KZG [if I'm going to use the board] or I flip the board [resell it as it came.]
I rarely see the whole time-line from new to failed KZG.

Closest I have seen...
The very first boards I saw KZG go bad on were about 4 years old when they just wouldn't boot after working fine.
[Know from date codes on caps. And I remember this because it was one of my first experiences with bad caps.]
BUT, I know that for roughly 3 of those 4 years the boards sat unused.
So, in summary:
Those boards had around a year actual use when the KZG failed.
But the caps themselves were around 4 years old.

Should note that:
Those particular PCs were occasionally run out doors on tables in the sun looping videos at swap meets and such.
Case temp was obviously a factor....
.

[Th3_uN1Qu3]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

My HD3870 has KZGs. 16v 680uF, but i didn't write down the date code. Chemi-Con logo, KZG, voltage and value in gold text, (M) 105C in white text on the other side. Still running fine after... guess it's been 3 years already.

[mockingbird]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

To confirm what you said PCBONEZ:

In my lap I have an Asus P5RD1-V motherboard. Around the VRM are 11 4V 680uF UCC "TMV" capacitors with no vents. All the tops except for two which are a bit farther from the CPU socket are bulging. The rest of the caps on the board are a mix of Rubycon MBZ and UCC KZG and a few other series of those to companies.

[retiredcaps]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

Chemicon KZG & KZJ
These are well known to bloat and very often fail completely with no bloating at all.

I know people here say KZG fails with no bloating and I'm curious if anyone measured a failed non bloated KZG cap?

For example, is the capacitance way off? Is the ESR really high? Is it shorted?

[PCBONEZ]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

If I ever did I don't remember.
.
When I found KZG problems I didn't own an ESR meter yet.
I found out KZG were the problem by replacing just the KZG with MBZ and the boards came back to life.
None of the KZG were bloated on the first 4 or 5 boards that worked with so since then I've been replacing KZG on sight.
.
I -vaguely- remember someone reporting high ESR but there are too many KZG posts to find it with a search.
.
I'll try and remember to check next time I run across a board full of KZG.
.

[retiredcaps]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

I'll try and remember to check next time I run across a board full of KZG.

That would be great.

[ben7]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

Perhaps there was a bad batch of these, just like the Nichicon HN?

-Ben

[severach]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

That bad batch would need to span many years. I see no evidence of the bad KZG and KZJ letting up. Good cap, weak formula.

[PCBONEZ]

Re: KZG, KZJ, -and- TMZ, TMV - Bad UCC series

That bad batch would need to span many years.

Yup. At least 9 or 10.