Multi-section caps last longer?

[Axtman]

Is it my imagination or does it seem like multi-section electrolytic capacitors seem to age better than individual electrolytic capacitors?

[Rhodesplyr]

Is it my imagination or does it seem like multi-section electrolytic capacitors seem to age better than individual electrolytic capacitors?

I don't think they age better overall. I've replaced plenty of them that have shorted, lost capacitance, or developed high ESR.

The real enemy of electrolytic capacitors is *heat.* It's possible that being located in a metal can somewhat isolated from external heat sources helps some of them, but in other amps, they are situated right next to rectifier tubes. This is the case in Scott 299 integrated amps--can capacitor right next to a 5AR4. In every one I've run across, this capacitor is ruined.

It also doesn't help when builders attach 2-5 Watt voltage dropping resistors directly to the can terminals!

David

[Dave Curtis, dB AudioTech]

It also doesn't help when builders attach 2-5 Watt voltage dropping resistors directly to the can terminals!

You mean like Ampeg did with my VT22? (and thousands of other V-series amps) The one that's still going strong after 37 years? With the OEM can caps? That's what those terminals are for, man! If those resistors are that hot, there's something wrong with the circuit...

(IMHO)

[melvin]

multi-sectionalism

I buy the heat business, but I also think they just used better materials back then, at least at Mallory.

[Dave Curtis, dB AudioTech]

True, but materials science has come a long way since then. But then we're talking modern/vs "back then". Mallory didn't have the market cornered; that's for sure. There's Sprague and Cornell-Dubiler to name two.

Comparing multi-section to discreet, I don't think there's much difference AFA failure rates go. Most 'lytic failures I see are from improper storage (lengthy and/or hot). They need a charge every now and then.

[Diablo]

You mean like Ampeg did with my VT22? (and thousands of other V-series amps) The one that's still going strong after 37 years? With the OEM can caps? That's what those terminals are for, man! If those resistors are that hot, there's something wrong with the circuit...

(IMHO)

I read too much on the internet about how E-caps should be replaced after 10-15 years. I replaced my original E-caps in my 37 year old V4 last year with new can caps. I don't think there was anything wrong with my original caps. I hope the new ones last another 37 years....but I doubt it.

[Dave Curtis, dB AudioTech]

Well, I'm not saying that 37 year-old caps shouldn't be replaced (I do have new bias caps in it), for all I know, the next time I fire it up might be the last time for those old caps. I just haven't had the chance to do it yet (cobbler's kids syndrome). I believe the "10 year" replacement schedule is a military spec.

[tedmich]

CDE 381LX series caps have a life span rated at 12,000h, or ~33 years at 1h a day or ~230 years at 1h a week.

From the handy Illinois cap lifetime calc (I avoid their caps though!):Illinois Capacitor, Inc. - aluminum electrolytic, metalized film, igbt, power film, organic semiconductor

the temp factor if an exponential, higher voltages are just geometric...

Replace whats broken with same value.
Replacing with different values is designing; possibly better, likely worse.

[R.G.]

If you're wanting to compare expected life of caps for multisections versus single cans, you run immediately into a problem - what generation of caps are on each side?

Modern electros have made many process changes that make it hard to compare against older caps. Not necessarily super improvements, just different.

The aluminum foils are now of much higher purity than caps that were 30-50 years old. High purity helps with good oxide coatings, and that contributes to lower failure rates. The foils are now almost universally etched to make for more surface area, and hence higher capacitance per unit volume.

Smaller caps for the same other ratings have better thermal characteristics, because although the same heat must be dissipated, the heat source is closer to the outside, reducing hot spots. Of course, the surface area is smaller now, which hurts. But multisection caps by definition have less surface area per cap, so the dissipation might be worse.

The electrolytes are now quite sophisticated chemically, for both low cost manufacture, low ESR and life.

The forming of oxides is now almost completely automated, which helps with consistency a lot.

Electrolytic cap life also depends on being charged up to some extent. The insulating oxide eventually decomposes and will not hold rated voltage. This happens faster if it is not charged. Modest use helps the oxide not deteriorate, and may actually rebuild weak places. Not being used forces the cap back into shelf life, which may be shorter than life under use with modest temperatures.

Finally, you are seeing the "survivor effect". Noting a really long lived cap as an example and thinking you can generalize to old caps lasting longer is not always good. You might also make the same generalization about caps printed with black inks (as most of the older ones were) instead of sleeved and printed. Older specimens are old and reliable because their weaker peers have long ago died and are not here to bear witness. As they say, the survivors write the history books.

[darrellcheng]

I've always wondered the same thing, maybe people were more willing to change out axial and radial caps for general maintenance purposes (like a perfectly working all original sf/bf fender) and cap cans stayed in unless someone was forced to change it.

[Dave Curtis, dB AudioTech]

All you ever wanted to know about aluminum electrolytic capacitors:

Chemi-Con.com: About Our Technical Aspects

[tedmich]

Finally, you are seeing the "survivor effect". Noting a really long lived cap as an example and thinking you can generalize to old caps lasting longer is not always good. You might also make the same generalization about caps printed with black inks (as most of the older ones were) instead of sleeved and printed. Older specimens are old and reliable because their weaker peers have long ago died and are not here to bear witness. As they say, the survivors write the history books.

Yes! and if you replace the old guy once it does die with a NOS made the same year you are compounding this fallacy.