Lifetime of electrolytic capacitors vs film capacitors

[ocset]

Hello,
The MKP1848  film capacitor has a life quoted as 100000 hrs at 70 degC.
The life of a Rubycon BXC series wet electrolytic capacitor is quoted as 12000 hrs at 105degC.
..By way of the  “10 degree rule”, we can say that the lifetime of the Rubycon electrolytic capacitor at 70 degC would be 96000 hrs.
This puts the lifetime of the wet electrolytic on a par with the film capacitor.
This doesn’t sound right, please advise if they are talking about a different kind  of “end of life” failure mode in each case?

MKP1848 polypropylene fim capacitor
https://www.vishay.com/doc?28164

Rubycon BXC electrolytic capacitor
http://www.rubycon.co.jp/en/catalog/e_pdfs/aluminum/e_bxc.pdf

[agehall]

Does the "rule" really apply in that way? I've always assumed that you can only use it to estimate shortened lifespans at higher temps...

[Zero999]

Yes, electrolytic capacitors do last for longer, when not run at their maximum temperature ratings.

Voltage is also a factor: running near the rating will also shorten their life. Strangely enough the same can also be true at low voltages: leaving the capacitor idle, i.e. with no charge on it, for a long time carries the risk of the oxide layer breaking down over time.

My bet is the film capacitor will have a longer life than the electrolytic capacitor. If it's run within a fair safety margin of it's maximum ratings, then it should virtually last forever.

[Gyro]

There are other factors that might come into play. For instance, Electrolytics have much higher capacitance per unit volume. Given space constraints (which often seems to be an issue for you treez), having a higher capacitance means that you have a larger margin for capacitance loss without affecting operation.

Along the same lines, Film capacitors will loose capacitance quickly if exposed to spikes, due to self-healing. If you again consider the same unit volume, a higher value Electrolytic is probably much better at soaking up those spikes.

Both are factors that are worth taking into account as well as looking at the datasheet life specs.

[ocset]

Thanks, you’re all making excellent points.

I must admit, with the electrolytics, actually nailing that absolute proof that they last less time than film capacitors is impossible to do. No matter how many Electrolytic cap or film cap App Notes you wade through.

Strangely enough the same can also be true at low voltages: leaving the capacitor idle, i.e. with no charge on it, for a long time carries the risk of the oxide layer breaking down over time.

Thanks, I tend to agree, (though ive never found written proof of this) I’ve certainly read that electrolytics suffer high initial leakage current after being stored with no voltage on them for a period of more than 18 months…..but all the app notes tell you that this initial high leakage current soon dies down when the capacitor is charged up in a circuit…..then its right back to normal again and all recovered. I must admit I’ve never read that the oxide of electrolytics breaks down when stored  with no charge for long periods  but I am sure you are right.

Voltage is also a factor: running near the rating will also shorten their life.

I must admit I tend to agree but never found an official reference to back this up.

If it's run within a fair safety margin of it's maximum ratings, then it should virtually last forever.

Thanks, (regarding films) that’s what I thought, but clearly virtually all polypropylene film capacitor datasheets have a lifetime stated. –Given that this is the case, our investors simply will not fund our building of SMPS’s with more expensive film capacitors instead of electrolytics  as they say the basis that film capacitors have no wear out modes is wrong….they simply point to the lifetime quote in the film cap datasheet.

By the way, I am referring to wear out due to running at high temperature rather than overvoltage spike related things.

I  thought that wet electrolytics “dry out”, whereas film capacitors do not?
However, having reverse engineered many SMPS’s and seen electrolytics covered over with potting compound, I am wondering if this is a way that electrolytics can be prevented from drying out, and made to last longer?

[CatalinaWOW]

I agree with you, that film capacitors "last longer".  But it may be a difference that doesn't matter.  I have never seen a film capacitor die of old age.  Only from severe abuse.  And I have experienced large numbers of electrolytic capacitors that have survived many decades.  While lower quality electrolytics in stressful applications (switching supplies particularly) that involve thousands of high current exposures per second do regularly fail early, I would not worry overly about quality electrolytics failuring during normal product life cycles.  Few products experience continuous exposure to 70C and even if they do experience that kind of continuous high temperature exposure 10 years is a very respectable operating time.