Capacitor Failure caused by heat ???

[Tfeix]

hello community,

i already read some nice article's here at the forums but i didn't found out if a capacitor can also died in case of a "overheat". please, can somebody give me information if that might can be a problem? here is a short example:

i have problems with caps from a fic mainboard (the caps are labeld with "S. I.", don't know the manufacturer). the caps are specified at 105°C and located near the CPU. in my case the CPU (a celeron 633) is specified at 82°C (means, that cpu turns power of at a higher temp). if the cpu do not reach the 82°C barrier, the caps near by the cpu would also never get hotter than 82°C. do you think this could be a "heat"-related problem?

thanks for your help

[MD Willington]

Re: Capacitor Failure caused by heat ???

Yes, the life of the cap is degraded the longer it stays in a high temperature environment. Some boards have the caps very close to where the heatsink is, as a result of the heatsink fan blowing down and outward, the caps are bathed in hot air...

[arneson]

Re: Capacitor Failure caused by heat ???

A bathing Cap?

[MD Willington]

Re: Capacitor Failure caused by heat ???

LOL...

bathe  bathed, bath‧ing

6. to cover or surround

[yanz]

Re: Capacitor Failure caused by heat ???

haha..

i do simple calculation. yes heat is a great enemy for a capacitor, even the good one. but you'll get by with the 7k or 10k hours rate caps, it will last for years.

[PeteS in CA]

Re: Capacitor Failure caused by heat ???

The key temperature is that of the core. If at 25C ambient, the core temperature is 75C, at 65C ambient, the core will be 115C. Locating caps near hot components and/or in areas with minimal airflow can bite the designer, or user, if the problem gets shipped.

[yanz]

Re: Capacitor Failure caused by heat ???

i dont really understand that. you mean that if the body of a cap is 65C, then its core is 115C? isn't that too high? i tought the delta will probably be 10C...

well, it would be nice if all the mobo have "otes" feature like abit's or ecs's.

[PeteS in CA]

Re: Capacitor Failure caused by heat ???

yanz, I wasn't commenting on the case temperature, only pointing out that the core (the foils, electrolyte, and filler paper) temperature rise above ambient air temperature is the same, regardless of the ambient air temperature.

[yanz]

Re: Capacitor Failure caused by heat ???

ah, ok. sorry i've misunderstood what you said. so that the delta temp in ambient temperature will be the same with the delta temp for the core of the caps.

[Tom41]

Re: Capacitor Failure caused by heat ???

If the temperature of a cap rises high enough, the electrolyte will actually start to boil and change into vapor. If the high temperature continues, the pressure will eventually pop the vents at the top of the cap (or blow the bung out), with a loud bang and a cloud of smoke. If you have electrolysis occurring inside the cap as well - such as with many 'bad' makes - the explosion will be more violent.

Most caps have a vent temperature on the sheathing, if the internal temperature gets above that, it's likely to vent.

[UraBahn]

Re: Capacitor Failure caused by heat ???

Also S.I. is a bad brand. I had a PNY video card which developed bulging S.I. branded lytics. Replacing the caps might help the problem.

[yanz]

Re: Capacitor Failure caused by heat ???

the only sure thing i know is that the caps won't be any cooler than the ambient temperature. so that if the ambient temperature in your system (inside the case) is about 30-35C, than it won't be too far to expect that your caps are hotter than 30-35C, or even more hotter.

the lower the esr rate of the caps, the cooler it will runs.

the cooler the ambient temp or pcb/trace where the cap is placed, the cooler the cap will be.

the greater the ripple voltage come from the psu to caps on the mobo, the hotter those caps will be. add to that the vrm cpu will be less efficient so that the mosfet will generate more heat, and the heat will transfer to the ambient and trace of pcb it will warm the cap even more.


so that in summary there are three important factor on your system to be healthy so that caps can live peacefully:

- the quality of the caps itself. they should be low enough of esr. not necesarily to be the lowest, but good enough for the design of the mobo.

- airflow in your case. if you use watercooling, don't forget that the components on the mobo itself still need good heat transfer to the air/ambient. dead airflow sounds noiseless, but so does the sound of death systems because of dead/bad capasitors.

- good/healthy psu than can supply clean power to the system. don't give your system junk foods, or they'll operate less efficient and degrade faster.


as always cmiiw.