Wow, that looks bad! What values and voltages are those caps? I have a few polymer/hybrids laying around. I've been meaning to test them, but there has been no opportunity to do so till now. You are welcome to try 'em. Made by NIC components.
Q6700 @ 3.6 GHz
Zippy GSM-6600P
Curcial Ballistix PC6400 (4 x 1Gb) Micron D9GMH
Abit IP35Pro
ATi HD4870
How close are they, in system, to hot parts or heatsinks? Maybe aqueous parts are not suited in those locations in the context of that system?
PeteS in CA
Power Supplies should be boring: No loud noises, no bright flashes, and no bad smells.
****************************
To kill personal responsibility, initiative or success, punish it by taxing it. To encourage irresponsibility, improvidence, dependence and failure, reward it by subsidizing it.
****************************
The ones on the right edge are part of a linear regulator, so ripple current is probably not an issue - so I'm guessing that heat is the culprit, as PeteS, Gonzo and I have been conjecturing for a while regarding aqueous electrolytics. FJ is even smaller than FM, so it probably can't take the heat either. Looks like FC/FK are still the premium Pannies, notwithstanding their larger sizes and (sometimes) limited availability.
So I'm guessing that heat is the culprit,
PeteS, Gonzo and I have been conjecturing for a while regarding aqueous
yes I have been following the comments on that and it is a topic of concern and should be followed up with feedback from the recaps out there with FM's.
Thanks linuxguru
was just a passing thought on the supply thing, (just a long possibility really)
It will be interesting to see what KC8 comes up with When he examines the system. At lest he will be able to line the ducks up and knock them down.
This one is a real concern as its only 6 months to failure in something I would have thought would have had a slim chance of failure like this in that time.
Strange
cheers
You step into the Road, and if you don't keep your feet, there is no knowing where you may be swept off to." Bilbo Baggins ...
Even if i am aware, that any aqueous cap bears a inherent risk for going bad, i am a little surprised about this dramatic case (regarding the time frame of the failure).
So i think, that there must be a failure either in the air flow or in the electronic design of this system or otherwise some problems.
I have used FM`s in probably far more hazardous environments and i even put a lot of them in a regularly used TV set, which is a pretty hazardous environment for caps (heat & ripple).
Up to now no problems at all.
I have had a MSI GFX6600GT which was equipped with a few UCC KZG caps, and even they have not developed any sings of failure (which we all know, is quite possible if heat is involved).
From the pic, i think there could be an linear voltage regulator on this card, which certainly would generate a lot of heat.
If put in a small box without airflow, i think this could really be the cause.
Any way, i think we will soon know more about this issue.
I've seen many AGP video cards with the SP series OSCON caps and KMG chemicons that work very well. That card should not run very hot. What is it? A 6200? I'm skeptical that heat alone was the problem here. It is more likely that the ripple current specs were exceeded by a large margin or that it may be a bad batch of FM's.
well that was my passing thought but then I thought wouldn't it be showing signs else where? or shouldn't it be showing signs video wise before this
Unless possibly very localized maybe bad (High) contact resistance pci card/slot? and the video card wasn't showing signs or perhaps nobody was picking them up.
(perhaps even a faulty supply rail track on the card its self?)
the 3 that are spewing , this would make sense if they were I/P type caps and the other were o/p caps (on a regulator, reg acting like a shield for the O/P caps)
The 2 3 term devices look like regulators or possibly Fets..but I got no idea.
Faulty batch I suppose is possible, but I think KC8 buys them by the bucket load
(so guess he would have a bucket of duds if that was the case)
Heat is a definite contender and linear regs can get hot but looking at the card I cant overly see it so guess its back to the box as a total environment and what is the environment it works in.
So really I don't know
I know you guys work with this stuff a lot more then me so you guys would be more aware as to possible causes.
Anyway just my ramblings, thoughts and somewhat uneducated guesses.
Post some photos of the box if you can and don't mind KC8.
Whats the internal temp of your PC? Download a program like Speedfan which will tap into your mobos built in temp sensors and monitor your system temps. Preferably while using your graphics card to its potential by playing a game.
Excess heat can cause ANY cap made by anyone to fail prematurely.
Also the caps that bit the dust are very close to those MOSFETs or VRs can't tell which. Its very likely that their heat cooked the caps. Which points back to the need for better ventilation.
That's very interesting for this case. I want to send you Samxon GC free of charge, and test it on same display card. Would you want to get a try?
i am going to look at the box very closely.
if i dont find a reason for the failure we will try your caps.should be a good test in this cap killer.
btw the first batch is still going good in the emachines box i put them in.
if you look at the empty one it appears it has been designed to take both
(or either I should say)
and yeah ditto the FM positions look the same
(in 2 possible spacing by the looks of it too, through hole)
I have done a little math, which would make this scenario possibly if the caps are exposed to an ambient temperature of about 75°c for 6months 24/7 use. This would sum up to about 30.000h.
This could explain a failure, but the obvious bulging and leaking is probably not the standard failure mechanism.
Most GFX PCB`s does have pads for both, SMD and radial caps. this is not an indication for which cap technology this pcb was meant for.
Comment