Capacitor failure from disuse ?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Script

Well-known member
Joined
Nov 27, 2008
Messages
2,393
Location
Japan
Just found following in wikipedia on "Capacitor". I pretty much knew about limited lifespan and what shortens it. But failure from disuse was new to me -- like so many other things.

"Electrolytic failure from disuse
Aluminum electrolytic capacitors are conditioned when manufactured by applying a voltage sufficient to initiate the proper internal chemical state. This state is maintained by regular use of the equipment. If a system using electrolytic capacitors is unused for a long period of time it can lose its conditioning. Sometimes they fail with a short circuit when next operated."

But this applies only to older caps, right ? Not new ones, I suppose. -- new as in cap bought ten years ago for stock but never used ?
 
Most likely. Nowadays, even the lowliest caps you can buy (from the handful of reputable manufacturers) are rated for 2000h and up. That lifetime is at the rated temperature (85 or 105c), and it doubles for every 10c under that.

So even a 2000h 85c cap should be totally fine even after 128000h of just sitting at 25c (5333 days or 14.6 years).
 
It's called "forming in" and very old caps can be reformed. It is kinder to them using a series resistor to current limit while applying rated forming voltage to them.

While working inside Peavey we would annually review slow moving raw material inventory. One time I had a pile of very old capacitors to consider. IIRC I disposed of them rather than investing more time and effort to only end up with maybe OK old parts.

JR
 
[Off-topic ]
I guess this is what happened with at least one nuclear reactor in Japan. Turned off in 2010, and when they finally got approval to start up again after years, they flipped the switch and immediately had to turn off again -- news reported on fire in a control console.
[/off-topic end]
 
[Off-topic ]
I guess this is what happened with at least one nuclear reactor in Japan. Turned off in 2010, and when they finally got approval to start up again after years, they flipped the switch and immediately had to turn off again -- news reported on fire in a control console.
[/off-topic end]
I would not bet heavy money on capacitor forming causing a fire after sitting only ten years. Lots of other things could go wrong with that complex technology.

JR
 
very true.

[more off-topic] TThings like that seem very little forgiving when it comes to neglected maintenance -- definitely was the case with those reactors in F., which dated from the mid-60s and had a remaining lifespan in 2010 of only a few months before desaster. They were scheduled for decommissioning lin second half of 2010 Meaning, the quake alone was enough to rip apart tubing etc, as had been reported by some staff back then. The rest, well...[/completely off-topic]

Will keep in mind possible need for 'conditioning' when grabbing for NOS caps in my stash. Although rather recent hobby for me. And I will turn on gear that sees little use more often from now on, promise ; )
 
Most likely. Nowadays, even the lowliest caps you can buy (from the handful of reputable manufacturers) are rated for 2000h and up. That lifetime is at the rated temperature (85 or 105c), and it doubles for every 10c under that.

So even a 2000h 85c cap should be totally fine even after 128000h of just sitting at 25c (5333 days or 14.6 years).

I suggest that other mechanisms are likely to come into play when a cap' is just 'left on a shelf' for an extended period ?
 
[Off-topic ]
I guess this is what happened with at least one nuclear reactor in Japan. Turned off in 2010, and when they finally got approval to start up again after years, they flipped the switch and immediately had to turn off again -- news reported on fire in a control console.
[/off-topic end]

Bit of a stretch. Not sure what mechanism being envisaged there but I'd hope that, in any case, there would be enough 'housekeeping' monitoring involved to avoid a simple component failure causing a catastrophic system failure.
 
Just found following in wikipedia on "Capacitor". I pretty much knew about limited lifespan and what shortens it. But failure from disuse was new to me -- like so many other things.

"Electrolytic failure from disuse
Aluminum electrolytic capacitors are conditioned when manufactured by applying a voltage sufficient to initiate the proper internal chemical state. This state is maintained by regular use of the equipment. If a system using electrolytic capacitors is unused for a long period of time it can lose its conditioning. Sometimes they fail with a short circuit when next operated."

But this applies only to older caps, right ? Not new ones, I suppose. -- new as in cap bought ten years ago for stock but never used ?
Just putting it a different way, but see if this helps: As you know, capacitors are two conductive surfaces separated by an insulator. The main advantage of an electrolytic capacitor is high capacitance for a given size. It does this by having essentially one metal (foil) plate—what serves as the other "plate" is an electrolyte (from which we get the name of the capacitor), usually liquid, the cathode. But it would be a dead short with nothing else, so an oxide layer is grown on the foil to separate them. This is done by limiting voltage during forming during manufacture. Obviously (maybe), running the voltage in the other direction would undo, the oxide. So, the oxide is a dielectric, the capacitor is inherently polarized.

And the basic problem is that the oxide goes back in solution with the electrolyte over time, the capacitor turns into a dead short, or close enough that the insulator burns up when you power it up, or the chemical reaction causes the capacitor to burst (most often seen with power supply caps). If you suspect this in advance, you can reform the capacitors. That's usually only worthwhile for large caps, where you either want the vintage look for the device, or it's a matter of fit (though modern electrolytics are typically smaller), or just plain hard to find.

So, given enough time, it's likely that it doesn't matter whether the cap is more modern or not. It's just the nature of electrolytics. (There's "capacitor plague" from ~20 years ago, but that usually resulted in the capacitors rupturing from gas build-up within a couple of years I think. Believed to be from a mis-copied stolen formula, the electrolyte caused corrosion.) Big caps seem to last longer, particularly big power supply caps, but I don't know for sure, worth powering up an old amp with a dim bulb tester or variac and check current.

I've recapped a number of devices over the past year or two, stuff I bought in the '80s that were no longer working. Basically every one of them worked after replacing the electrolytics (cheaper than troubleshooting which were bad). Lexicon PCM-70 (had a clearly "blown" cap, 5v supply), NAD integrated amp, Oberheim OB-8 (AC line filter also smoked, also battery dead), PS Audio IVH preamp...
 
Last edited:
replacing the electrolytics (cheaper than troubleshooting which were bad
That's what I ve been doing with (old) malfunctioning gear ever since I started . In 80 % of cases it's faulty caps here -- followed by pots etc

At prices for caps I think it's a no-brainer and usually I do not fancy having to dis/reassemble gear again a year later just to fix that next cap, which happened.

One leaked cap in a non-heat area and all same type/maker caps have to go ! Wth common gear, anything else I'd say is a field experiment.

//
Will keep an eye on my stock of big power supply caps.
 
Such as.....?
Well if I could answer that in detail, then I'd likely be working for Cornell Dubilier, Vishay or some other Capacitor manufacturer rather than specifying and designing with them. But basically it's Shelf Life rather than active hours.
But since you asked the question I see that the issue has been addressed in n a more comprehensive manner.
 
I often notice with tube amps which havent been powered up in years , that noise levels are high initially , you here hum come and go and often other frying type sounds as things reform , it can take a couple of days with power applied for everything to settle back down to normal again . With very old stuff the lamp limiter or variac can help ,but in any case where the seals are bulging or ruptured the cap will probably have dried out long before , so they are best changed out for new . I notice the old LRC 50+50 500v cap like used in Marshalls is no longer available ,
modern dual can caps can be quite costly , some people empty the original can and build in smaller modern components .

There does seem to be a few modern Chinese made dual caps on Ebay , Id be interested to hear from anyone whos tried them as I have a room full of vintage amps that I'll need to fix up at some point .

https://www.ebay.com/sch/i.html?_fr....l1313&_nkw=50+50+uf+capacitor+500v+&_sacat=0
 
CE Distribution bought the Mallory production line for multi-section can caps. CE Distribution You need to have an account to buy from them, so more for folks that run a business than one-off DIY'ers. My dealings with them have been 100% satisfactory.
If you're in Canada, and need some of their product, I can probably help you out.
 
CE Distribution bought the Mallory production line for multi-section can caps. CE Distribution You need to have an account to buy from them, so more for folks that run a business than one-off DIY'ers. My dealings with them have been 100% satisfactory.
If you're in Canada, and need some of their product, I can probably help you out.
You can get single pieces through the tubesandmore.com store
 
I looked at the tubes and more site , they have a few of the CE caps, quite costly , but they do have upto 4 sections with various values of capacitance , handy to know as some older tube gear use three and four section caps with different values , always nice to be able to keep things close to original rather than having to bodge in extra caps here and there .
 
Multi-section caps like you see at Tubes and More always seem to have lousy specs (if they are even provided) - and they are expensive. Isn't re-stuffing with quality caps a much better option?
 
A bit off topic but back in the early '70s the public TV station I volunteered at kept racks of old tube gear powered up 24/7 on the theory that if they ever needed to patch over to the backup it would be a lot faster to replace the hundreds of tubes than thousands of caps. I suggested the electric bill would have justified buying new gear, but then what did I know, I was just a kid.
Caps have gotten a lot better over the years and I haven't worried about aging of caps since about 1980. Maybe I just don't care anymore. I did trash my last piece of tube gear last year though, due to a 2 section cap failure. It just wasn't worth the cost of the cap.
 
Back
Top