Most and Least Reliable Electrolytic Capacitors

[dfuller]

Sort of a branch off the madness in the Apollon NC2K teardown thread:

Which brands of caps have you had the most and least trouble with? I've had nothing but trouble with Illinois/IC TTA and LBA/LBB series . They seem to consistently be a problem in every single amp I encounter them in. In contrast, I've yet to have a problem with Nichicon, Rubycon, and TDK/Epcos caps of any variety.

How about you?

 

[sq225917]

Only ever had issue with Lelon organic polymer caps in an Audiolab mdac. Never had issue with any other caps. I do stick to the big three for self build projects, nichicon rubycon and panasonic.

 

[JeffS7444]

There was a period in the early aughties when computers in particular were plagued with capacitor problems, and I saw some name-brand parts affected along with unfamiliar Chinese brands, in Dell and Apple products. The old maroon Roederstein ("ERO") electrolytic capacitors as used in Bang & Olufsen and other European makes from the 1970s can also be a pain.

I've used brands like Xicon with good results, but often the cost of the capacitors is pretty trivial compared to the labor I'm investing into the project, and there, I'll want to see some datasheets, paying some attention to the Mean Time Before Failure (MTBF) ratings. And all other specs being fairly equal, I can't imagine why I'd want to choose a part with lower rated MTBF. For a recent preamplifier build, I used mostly United Chemicon parts.

Although tempted to replace aluminum electrolytic capacitors with polymer types, leakage current ratings of the latter are typically higher, and if I'm not sure that's going to be a problem, I'll just stick with the former.

 

[Cbdb2]

Horses for courses. Just replaced some lytics in an LG monitor. They were obviously underspeced for the the application, (theres a youtube vid showing how to replace them its so common). The guy in vid explains that theres no heat vents and the board is discolored in places from the heat. So why did they use 105c caps instead of the 125c caps? Bad design. So I think its not always a bad cap but the wrong cap. If your dealing with high temp, high ripple current, get the right cap, and if your worried, over specing the temp and voltage will increase the cap life. The compromise is larger more expensive caps.

 

[dfuller]

Only ever had issue with Lelon organic polymer caps in an Audiolab mdac. Never had issue with any other caps. I do stick to the big three for self build projects, nichicon rubycon and panasonic.

Because I usually work on tube amps, I often need high voltage axial caps - so that limits me to either IC (crap) or F&T (expensive, but damn near indestructible). Well... Technically, JJ Electronic makes them too, but they're also expensive, their temp ratings aren't amazing (70C on the axials...), and the wire leads are kinda chintzy.

Horses for courses. Just replaced some lytics in an LG monitor. They were obviously underspeced for the the application, (theres a youtube vid showing how to replace them its so common). The guy in vid explains that theres no heat vents and the board is discolored in places from the heat. So why did they use 105c caps instead of the 125c caps? Bad design. So I think its not always a bad cap but the wrong cap. If your dealing with high temp, high ripple current, get the right cap, and if your worried, over specing the temp and voltage will increase the cap life. The compromise is larger more expensive caps.

The issue I encounter with bad caps most of the time is premature failure, even in circuits where they should be totally fine. I'm usually lucky enough to catch it at early stages where it's just causing motorboating or some kind of oscillation, but occasionally I get an amp where it's just constant 120Hz hum and the cap has actually leaked.

 

[valerianf]

"Nichicon, Rubycon, and TDK/Epcos caps "
Designing industrial and military products since decades I can tell that these are good brands.
However in the electronics for audio, as manufacturing plans are in Asia, you could easily find unknown local brands mounted onboard.
Then it is a reliability issue.
The gamble is easy to accept for a cheap product, but I am more reluctant for a high end one.
Anyway you never know before buying the product or you need to buy the service manual before.
Nobody does that.

 

[restorer-john]

Most reliable in all my years in electronics are the following:
Elna
Nippon Chemicon
Panasonic
Sanyo
Nichicon
Rubycon
BHC Aerovox (now owned by Rifa)
Matsushita (Panasonic)
Sprague
Jamicon (Taiwanese)

Least reliable:
Towa (Japanese)
Rifa
Philips
Samxon
CapXon
Lelon
Teapo
Su'scon
Suntan
Acon
Samwha

There's a ton of bottom tier brands I simply cannot recall right now, but suffice it to say, they get removed and replaced with the top tier brands and I never see or hear from them again...

Here's two of five Acons I removed only the other day (they were still in my bin). They look pretty with the black and gold lettering. Notice the bulging vent top. They were about to go nuclear. SMPS secondary 12V rails.

This series of Samwhas is terrible for reliability too. Yes, I took a closeup, but Samwha, spend some money on better screen printing sleeves?

 

[TomJ]

Most reliable in all my years in electronics are the following...

How about Jianghai? Reason I ask is that the NAD C 298 SMPS uses them, rated for 3000 hrs at 85C. The M33 SMPS looks similar but can't see whether the caps are the same. Thanks!

 

[restorer-john]

How about Jianghai? Reason I ask is that the NAD C 298 SMPS uses them, rated for 3000 hrs at 85C. The M33 SMPS looks similar but can't see whether the caps are the same.

There's a lot of capacitors either coming out of a few factories and being labelled with multiple brands, or there's more capacitor factories coming on line all the time. A huge number of electros are clearly coming from Chinese factories, particularly the flood of fakes. In my opinion, the fake capacitors are one of the biggest contributors to the overall failures on cheap gear (not so much HiFi, but consumer electronics in general).

When it comes to audiophile approved capacitors, I can safely say all (except genuine old pulls from gear) Rubycon Black Gates being sold now are fake. There is one little thing the counterfeiters didn't take into account and it's obvious to us techs.

I posted a very useful graphic last year on identifying fakes from real. I won't repost it here, but have a search. Suffice it to say, a huge proportion of Nichicons sold across Asia are fakes. They are capacitors and they work, but internally (when you cut them open) they are not Nichicon. Exterior also gives away clues they are fake.

So along with unreliable 3rd tier brands, we have fakes of tier 1 caps. A bit of a nightmare, unless you source the best caps from a known and verified supplier/supply chain.

At least no-one is faking the 3rd tier brands- or are they?

 

[restorer-john]

See Post #16 in this thread.

https://www.audiosciencereview.com/...p-use-alex-cavalli-responded-on-page-2.10071/

 

[March Audio]

Only ever had issue with Lelon organic polymer caps in an Audiolab mdac. Never had issue with any other caps. I do stick to the big three for self build projects, nichicon rubycon and panasonic.

Me too ( not the actual one shown)

 

[March Audio]

Another major PITA for me was the PSU caps in my tag Mclaren AV32RDP. Started failing to boot the DSP section. Turned out to be the PSU caps and a common problem.

They were Elna replaced with equivalent Nichicon or vice versa, well major brand anyway cant remember - no doubt @restorer-john could identify.

Bottom line it was bad design, all the caps were running too hot crammed around the psu heatsinks.

 

[dfuller]

Another major PITA for me was the PSU caps in my tag Mclaren AV32RDP. Started failing to boot the DSP section. Turned out to be the PSU caps and a common problem.

They were Elna replaced with equivalent Nichicon or vice versa, well major brand anyway cant remember - no doubt @restorer-john could identify.

Bottom line it was bad design, all the caps were running too hot crammed around the psu heatsinks.

It's always annoying when caps get placed near a heat source when they don't need to be. One has to wonder if that was an engineering oversight or somewhat intentional to guarantee an upgrade at some point...

 

[levimax]

I just completed an amp build and the only PS caps I could find for the specs I was looking for from Mouser were "Kemet" brand. They say "made in EU" but I don't see them mentioned here as either good or bad... they have a nice web site. What is the deal with "Kemet"?

 

[StefaanE]

How about Jianghai? Reason I ask is that the NAD C 298 SMPS uses them, rated for 3000 hrs at 85C. The M33 SMPS looks similar but can't see whether the caps are the same. Thanks!

FWIW, a German forum (Igorslab.de) had the following to say about Jianghai:

"Und wer jetzt beim Hersteller auf einen Chinesischen Ramschhersteller tippt, liegt schon mal grundverkehrt. Jianghai ist einer der ältesten chinesischen Hersteller und aus einem ehemaligen Lieferanten der Volksarmee hervorgegangen. Außerdem hat man mittlerweile die komplette Kondensatorproduktion von Hitachi gekauft. Also hat man im Endeffekt Hitachi-Qualität zum fairen Preis. Die Teile müssen sich hinter denen von Chemicon wirklich nicht verstecken."

Anyone who'd think the supplier is one of those Chinese junk manufacturers would be completely wrong. Jianghai is one of the oldest Chinese manufacturers and emerged from a former supplier to the People's Army. In addition they purchased the capacitor production line from Hitachi, so all-in-all, you get Hitachi quality at a fair price. The components don't fear comparison with those from Chemicon.

 

[dfuller]

I just completed an amp build and the only PS caps I could find for the specs I was looking for from Mouser were "Kemet" brand. They say "made in EU" but I don't see them mentioned here as either good or bad... they have a nice web site. What is the deal with "Kemet"?

Kemet are good quality parts. Probably roughly on par with most of the "good" brands @restorer-john mentioned, or others like CDE.

 

[DS23MAN]

It's always annoying when caps get placed near a heat source when they don't need to be. One has to wonder if that was an engineering oversight or somewhat intentional to guarantee an upgrade at some point...

Design wise the caps should be close as possible to the regulator ( short pcb tracks).

 

[JeffS7444]

Because I usually work on tube amps ...

The issue I encounter with bad caps most of the time is premature failure, even in circuits where they should be totally fine. I'm usually lucky enough to catch it at early stages where it's just causing motorboating or some kind of oscillation, but occasionally I get an amp where it's just constant 120Hz hum and the cap has actually leaked.

I noticed something interesting about an McIntosh MR71 tuner I owned for many years: Larger chassis than typical Dynaco meant heat was better distributed, and I saw more thoughtful little touches such as first filtering section of the power supply as a separate axial lead capacitor rather than a costly multi-section can, so that the cheap part gets hit with most of the ripple current. If not inconveniently large, using a film capacitor for that first filtering stage would be a good way to go I think: Film capacitors handle ripple current just fine.

 

[dfuller]

Design wise the caps should be close as possible to the regulator ( short pcb tracks).

That's true. There's something of a compromise there.

I noticed something interesting about an McIntosh MR71 tuner I owned for many years: Larger chassis than typical Dynaco meant heat was better distributed, and I saw more thoughtful little touches such as first filtering section of the power supply as a separate axial lead capacitor rather than a costly multi-section can, so that the cheap part gets hit with most of the ripple current. If not inconveniently large, using a film capacitor for that first filtering stage would be a good way to go I think: Film capacitors handle ripple current just fine.

Yeah, if low hundreds of uF, high voltage film caps weren't obnoxiously huge and expensive I'd prefer them. But considering they are (150uF/400V Solens are $30 each!)..

 

[March Audio]

It's always annoying when caps get placed near a heat source when they don't need to be. One has to wonder if that was an engineering oversight or somewhat intentional to guarantee an upgrade at some point...

I really don't think so in this case and it really doesn't fit with whatcI know and experienced with Tag.

If you look at the board it is rammed from one end of the case to the other. There really is very little space to do anything else. The designer(s) were probably constrained by "it has to fit in this size of case". Those premium (Elna?) caps failed in about 5 years.

It was however an excellent AVR, not like what we see these days. Actual hifi performance and measurements.