Capacitor failure

[Studiot]

For those EE that have heard of, but not seen, here is a photo of capacitor failure on a motherboard.

I have labelled the relevant material.

[berkeman]

For those EE that have heard of, but not seen, here is a photo of capacitor failure on a motherboard.

I have labelled the relevant material.

Were you fixing it at the time? :biggrin: (see your footer)


Did the supply overvoltage to cause that?

[Studiot]

Were you fixing it at the time?

Rumbled!

:rofl:

Did the supply overvoltage to cause that?

No, this mobo (probably) died of heatstroke.

[Jiggy-Ninja]

I've got one too. Maybe I'll post a pic of that later today.

[Skaperen]

One aspect of capacitors I never see described is current capacity. They only have a voltage rating. Could it be that for a given capacitance, voltage, and frequency, they can always handle the current that would be involved continuously? Or is there really a limit?

If these capacitors overheat, isn't that more directly an issue of current flow? If the metal inside is too thin for the specified usage, they would get too hot, right? Then there should be either an ambient temperature limit, or a heat rise rating. But that can also depend on air flow, etc. I so often seen gangs of little capacitors all tucked together. That doesn't let the heat early and we end up with the magic smoke being let out later.

[Studiot]

Not in this case.
T'was a seriously abused pc, running 24/7 and the environment inside the case just too hot for too long.

[berkeman]

One aspect of capacitors I never see described is current capacity. They only have a voltage rating. Could it be that for a given capacitance, voltage, and frequency, they can always handle the current that would be involved continuously? Or is there really a limit?

Look for the "ripple current" specification in the datasheets...

[dlgoff]

I was wondering about Power Factor compensation capacitors for inductive loads and what causes their failure and came across this about "Harmonic Overloading of Capacitors". It's from the Galco Industrial Electronics, Inc. site.
In certain circumstances, harmonic currents can exceed the value of the fundamental (60 Hz) capacitor current. These harmonic problems can also cause an increased voltage across the dielectric of the capacitor which could exceed the maximum voltage rating of the capacitor, resulting in premature capacitor failure.
http://www.galco.com/circuit/PFCC_har.htm

[Bassalisk]

Not relevant but:

http://www.youtube.com/watch?v=NQQMK1Rvq0A

I get chills every time i watch this.

[mdjensen22]

Look for the "ripple current" specification in the datasheets...
Also the ESR of a capacitor - lower ESR yields lower heat, lower aging, etc.

http://en.wikipedia.org/wiki/Equivalent_series_resistance#Capacitors

[Averagesupernova]

Not relevant but:

http://www.youtube.com/watch?v=NQQMK1Rvq0A

I get chills every time i watch this.

It's been a while since I've seen that one. A bit scary. I don't recall where and there may be plenty others linked on youtube but there are alot of other vids on the net that show similar things. My favorite was a transformer that went up in what appears to be someone back yard.

[MATLABdude]

A lot of really cheap capacitors came on the market supposedly after a little industrial espionage wherein a disgruntled scientist at one of the Japanese components firms stole the formula for electrolyte and sold it to manufacturers in Taiwan and China:
http://spectrum.ieee.org/computing/hardware/leaking-capacitors-muck-up-motherboards

Since it wasn't the complete / correct formula, this led to a lot of faulty motherboards, from both low and high-end manufacturers (the capacitor plague):
http://en.wikipedia.org/wiki/Capacitor_plague

I thought it was basically a late 90s, early 2000s sort of problem, but I had to replace cheap bulged caps in the power supply board of an LCD (Samsung 205BW) from 2007 with high-temperature, low ESR, and high lifetime ones (Nichion).

http://i55.tinypic.com/4grywg.jpg

I generally like Samsung products, but I hear this is quite an endemic problem with them.

For those of you in a similar predicament, this site was quite helpful:
http://www.badcaps.net/forum/showthread.php?t=11439

[Antiphon]

A capacitor made for pulsed service often has a maximum dV/dt rating. This can be related to a maximum pulsed current through the capacitance.

Ripple specs are important for internal heating of the cap. The dV/dt spec is a matter of instantaneous peak internal current above which you damage the electrodes.

[Bassalisk]

I actually saw these electrolytic capacitors not exploding, but smoking compared to the others

http://www.youtube.com/watch?v=3b7mjukhTyQ&feature=player_embedded

[aravind.vlb]

hey what about the condition of that 1st one from left in the first pic?

[MATLABdude]

hey what about the condition of that 1st one from left in the first pic?

Same thing as in the others: exuded electrolyte (or a really dirty camera lens)