The Hall of Shame - Badcaps Photo Montage

[RukyCon]

Correct me if i'm wrong but isn't NCC's KZG line of capacitors know to be bad?

[Behemot]

They can usually work for some time but harsh conditions make them go bad sooner.

[momaka]

Originally Posted by Behemot

That's quite poor VRM and the PSU also does not seem to be much good. They must have been pretty stressed.

Looks like ASUS with that yellow color PCB. They either do a really good job with the CPU VRM (where it's efficient and there aren't any hot components at all) or they do a crappy job - at least for their older stuff. Not sure how it is nowadays with new boards... or ever since Core i3/5/7 came out anyways.

Originally Posted by ruky con

Correct me if i'm wrong but isn't NCC's KZG line of capacitors know to be bad?

Yeah, they are pretty shoddy - especially the 6.3V rated KZGs. The 16V rated caps seem to do a lot better.
In any case, if I see them on a buck regulator output (like CPU VRM out), I always replace them.

Originally Posted by Behemot

They can usually work for some time but harsh conditions make them go bad sooner.

Nah, the 6.3V KZGs are so bad, they will bulge by themselves in storage. I've seen it enough times now that I don't trust them at all. They are no better in use either, especially if there is even a tiny amount of heat present.

[Behemot]

I have seen enough of them working for years to state the contrary. I mean MANY do that but not all.

In the end, are you sure you have not exceeded shelf time? I would guess you 1000% did as the shell life is about 42 days. So even these caps meet their specs

[momaka]

Originally Posted by Behemot

In the end, are you sure you have not exceeded shelf time? I would guess you 1000% did as the shell life is about 42 days. So even these caps meet their specs

Your shelf life calculations are wrong.

Shelf life is typically stated as X many hours at max. rated temperature (so usually either 85C or 105C). Unfortunately, datasheets don't say much about shelf life at avg. room temperature.

[Behemot]

In that case it obviously makes sense to consider them the same…as usually, they MAY handle more, but this is the worst case.

[momaka]

Originally Posted by Behemot

In that case it obviously makes sense to consider them the same…

Not really.
It probably works somewhat similar to expected lifetime, which approximately doubles for every 10C drop in temperature.

Also, most datasheets state for shelf life that a capacitor should be in spec up to X hours. That doesn't mean the capacitor should fail as soon as that shelf life is crossed. Also, datasheets often state that reforming a capacitor after it has been in storage for full rated shelf life, should bring it back to spec.

Clearly, that is not the case with KZGs bulging left and right.

Yes, I too have a lot of stuff with KZG caps that haven't gone bad and have lasted many years. But I just have enough stuff as well that hasn't lasted at all with KZG, so that's why I consider them not reliable.

[Behemot]

Not reliable!=not delivering spec. The thing is just they do not usually live way longer than specified as most other NCC series, as we are used to.

The ultra-low ESR series usually have very low endurance, other series from Rubycon, Samxon, Suncon (Sanyo) also fail after many years of use. Not arguing that KZG are generally even worse though.

[Wester547]

AFAIK, the ultra-low impedance series (NOS liquid electrolytics) from Rubycon, Samxon, Panasonic, Nichicon post-2005, and most from Suncon (except the WF series), did not have issues with generating hydrogen gas just sitting on the shelf. I would have to concur with momaka that KZG (and KZJ, as well as TMJ and TMV) are bad capacitors in earnest. So bad, in fact, that even though they can last many years before they fail in well ventilated environments, once they finally do start failing, they have an exceptionally high failure rate and nothing will change that. The issue is probably complicated... because they pushed the chemistry too far, the electrolyte is doomed to failure, and if you use too many stabilizing additives, you can reduce the conductivity of the electrolyte, along with the breakdown voltage, and unnecessarily increase the cost of manufacturing them. So they had to make them unstable to yield good performance, even though they don't last.

[Behemot]

You sure it is hydrogen, you analysed that using spectroscope?

Increasing numbers of bad ultra-low ESR Rubycon, Samxon and Suncon caps are reported lately. So

once they finally do start failing, they have an exceptionally high failure rate and nothing will change that

can be pretty much said about those as well. Just that it takes longer. I mean this is like all the others were immortal…no, they are not. They are still ultra-low ESR series with way too low endurance. You may most likely not run into the life expectancy of 6-10k hours caps as chances are something else will fail long before them. But 2k caps under heavy load even from reputable brand even from reputable series will die sooner rather than later. I am just not so surprised as you ppl are when said caps die after 6+ years of usually daily usage.

Cause yes, ffs, time is running and all those 939, AM2, 775 boards with these bloated caps are 8-13 years old now! Socket 939 came in 2004, AM2 in 2006, same for Core 2 architecture. Nehalem arrived in 2008, that's 9 years ago! We will surelly see some of 115x boards with wet lytics dying soon, if we have not already.

[RukyCon]

Originally Posted by momaka

Looks like ASUS with that yellow color PCB. They either do a really good job with the CPU VRM (where it's efficient and there aren't any hot components at all) or they do a crappy job - at least for their older stuff. Not sure how it is nowadays with new boards... or ever since Core i3/5/7 came out anyways.

The motherboard is from a HP machine.

The computer has not been touched ever since it died years ago.

Still runs windows XP tho.

[Wester547]

Originally Posted by Behemot

You sure it is hydrogen, you analysed that using spectroscope?

I don't have such equipment, no. It's technically true that if a capacitor bulges and becomes as hard as a rock on the shelf, that while there is certainly gas formation, hydrogen formation would require bias across the plates.

Increasing numbers of bad ultra-low ESR Rubycon, Samxon and Suncon caps are reported lately. Socan be pretty much said about those as well. Just that it takes longer. I mean this is like all the others were immortal…no, they are not. They are still ultra-low ESR series with way too low endurance. You may most likely not run into the life expectancy of 6-10k hours caps as chances are something else will fail long before them. But 2k caps under heavy load even from reputable brand even from reputable series will die sooner rather than later. I am just not so surprised as you ppl are when said caps die after 6+ years of usually daily usage.

But those failures are almost all heat and duress-related (with the exception of Sanyo WF, again, along with pre-2006 Nichicon and TK ATWY/ATWB which I forgot to mention beforehand), giving off the false impression of dreck quality as Chaos said in another thread. The vast majority of motherboard cap failures on this forum, at least in recent years, have been KZG failures (many KZJ failures too, and of course many bad TK ATWY/ATWB). MCZ mostly die in SFF and USFF machines, sometimes not. KZG and KZJ die in well-cooled and well-ventilated machines, left, right, up, down, oy vey, and every which way, which is why people willingly complain of them more than other ultra low-impedance equivalents from the "good" brands (again with notable exceptions, for the sake of simplicity).

As an aside, I'm not really convinced that the 16V-rated KZG and KZJ do much better than the 6.3V-related KZG and KZJ (or the 10V-rated ones, for that matter)... they may "last" longer because they're usually under less stress but I think they're doomed to failure too. Also, the endurance tests don't test the capacitors until they fail. They simply test them long enough for them to remain within the specified conditions even after however many hours at the rated temperature, maximum rated ripple current (particularly for low-ESR capacitors), maximum rated voltage, etc.

[RukyCon]

Failing OST's and Capxon's in a power supply...

[Behemot]

Would any of the contributors mind if I borrowed some pics for a banner? I have my own stock of HW with bloated caps, but don't have time to take good pics, it may be easier to just find some good ones here…at least I hope so

[RukyCon]

I would not mind at all.

If you like, i can leave links to all of the photos i've taken so far. [Of failing caps of course.]

[Uranium-235]

My 12 year old yahama reciever I got from Radio shack for a bitching $99. Had been working fantastic for so long, but last year it started to get a cold hum, that would go away after it warmed up. Eventually the hum would decrease when it got warm but not go fully away. I go tired of it today and finally took it apart. Pretty sure these are the culprit. I'm going to replace them with UCC KY's. Close match, slightly better specs, higher life

I love living 5 minutes from mouser headquarters. No shipping, can pick up locally.

[momaka]

Originally Posted by Behemot

Would any of the contributors mind if I borrowed some pics for a banner? I have my own stock of HW with bloated caps, but don't have time to take good pics, it may be easier to just find some good ones here…at least I hope so

I don't remember if I contributed any pics in this thread, but feel free to use any of the pics I have uploaded to BCN.

Originally Posted by Uranium-235

My 12 year old yahama reciever I got from Radio shack for a bitching $99. Had been working fantastic for so long, but last year it started to get a cold hum, that would go away after it warmed up. Eventually the hum would decrease when it got warm but not go fully away. I go tired of it today and finally took it apart. Pretty sure these are the culprit.
...
http://www.badcaps.net/forum/attachm...1&d=1491708964

Only 12 years old?! That's not a whole lot. All of my amps from the early 80's are still going strong with their original caps.

[Uranium-235]

they were made better in the 80s. This was around 2005

son of a bitch, the hardened goop held a diode under it and it cracked when I removed the small cap. I'm not too familiar with diode replacement, anyone know what to get? it was in a circuit with a 25v 1kuf cap

[Uranium-235]

UGH I cannot find a schematic for this amp board. I googled the board model and got NOTHING. googling the amp gets me owners manuals.

I know its a glass diode, not sure if its germanium, schottky, or zenier. All I know is its in this circut with this 25v cap, other side connects to a resistor (circuit below, can't see the resistor). Oh yeah. and its the main part of this boards VRM, past the LARGE transformer

I can't see the models on any of these diodes, they're too damn small, even with my phone zoom, and it has good zoom

[televizora]

These are a low power shottky diodes. Perhaps 1A/50V shottky diode will replace it just fine. Unless it's something special that requires this exact signal diode. D should be rectifier diode. If it was ZD, then It would be zener diode.
And you are right. They were producing shottky, zeners and other stuff in glass package.
p.s Also the sign on the back of the board says rectifier. If it was zener, then the sign would be different.