When to toss unused old electrolytic capacitors

--------------------

** Seen ones go very bad in less than that if not used, but likely a manufacturing fault in those cases.

Siemens brand 40,000uF @ 100V turned into a 10V cap after two years in storage - and could not be reformed in any sensible time.

** There is no particular age, but large physical size examples seem to be long survivors - 30 years is not unusual. ** The answer is a very definite maybe...

... Phil

My "plan" is to warehouse them for another 6 months and just repurpose them as fuses. :)

Yeah, but the manufacturers just recommend 'conditioning' before use. It's not great if you want to feed an assembly line, but for a 'got it in stock' event to occur and make you happy, it helps to keep the stock around.

I've recovered shelf-worn capacitors with a milliameter, current limit resistor, and variable voltage source. Watch the leakage on the meter, and maybe blow some hot air at the capacitor. When/if the leakage goes to zero, if the capacitance reads good, you're golden. Success is more likely than a dead capacitor, and it's a lot quicker than e-maiing in an order (or driving across town).

I recently tested a range of capacitors for leakage current under bias and over temperature -there is a write up of the results on E14:

One of the types I tested was a couple of ancient but unused 10uF 16V specimens (probably Sprague) that are about 25 years old (I bought them at an auction 19 years ago), They seem OK. (leakage current about 30nA

10V bias, 80C, less than 1nA at -20C).

I might use them in a quick test on the bench but I certainly wouldn't inflict them on a customer.

MK

--
This email has been checked for viruses by AVG. 
https://www.avg.com

On a sunny day (Wed, 2 Oct 2019 10:06:25 +0100) it happened Michael Kellett wrote in :

Not sure it is relevant, I have electrolytic caps (Philips) from about 50 years old (from the seventies). Those are still fine. My test is: just scope the ripple if in a supply circuit.

Elcos are the parts that most often fail in my experience. Funny, just yesterday a security camera failed (at 5 o'clock at night), turns out its wallwart died, and a 1000uF 16V cap was sort of looking puffed up. Replaced it with 2 470uF 16V from some other scrap board, now works fine to this moment. Reason is probably high temperatures that cause the Schottky diodes to leak more, causing a large AC component in the filter caps... I measure outside housing of that wallwart to be around 50 degrees C, inside must be a lot higher... Problem with fixing wallwarts is those are all glued, they saved on screws, a law should be made to force them to use screws so you can open those and repair those. There is some political push here now in the Netherlands towards making things 'repairable'. Using screws would be a good start. Almost every wallwart that delivers some power here and is on 24/7 (about 8 or 10 I think) has now had its electrolytic filter capacitors changed,

Unquestionably true, but OTOH: if a 1000uF/16V electrolytic were replaced with 300 pieces of 1206 10uF/50V caps (3000uF nominally, but one needs to include the C(V) loss, which is huge for many parts), would the resulting capacitor battery be more reliable? One MLCC would for sure be better, but for hundreds of them the probability of failure is merciless -- one shorted, the entire unit is dead. The electrolytics often have a limited level of self-healing, which MLCCs lack.

Best regards, Piotr

Drilling holes in the case can drop the temp a lot. Assorted caveats apply of course.

If you do have a no-better replacement ready, a hammer sometimes opens ultrasonic welds.

Re caps... all depends on use. Last time I checked them the majority of my few 1930s lytics were still ok.

NT

Huh I've got several ~5-10 year old Al electro's in my parts drawers. These are just used for prototyping, is there some easy test to check if they are still good? They seem to work OK, but lots of times I wouldn't notice the difference between 1 and 100 uF.

(Oh never mind... not 'my' parts box anymore, someone else's problem :^)

George H.

On a sunny day (Wed, 2 Oct 2019 04:54:54 -0700 (PDT)) it happened snipped-for-privacy@gmail.com wrote in :

Yes, interesting idea :-)

In this case I used a vice to hold the bottom part, and the biggest pliers I have to wiggle the top part lose. Case opened that way without damage. Cable ties now keep it closed (and easy to remove).

On a sunny day (Wed, 2 Oct 2019 13:47:53 +0200) it happened Piotr Wyderski wrote in :

A hypothetical question indeed, I would not want to do that in practice, I bought some Nichicon LF 16V 470UF low ESR High ripple current aluminum solid capacitors on ebay for my drone systems, those so far work OK, several in parallel, what was it? 10A ripple at 100 kHz and 7.5 V or something like that:

uses air cooling from the propellers for the rectifiers and caps.

Used to keep it flying indefinitely via a thin coax and a few hundred volts transformed down (in that ring-core).

one turn secondary hehe :-)

My experience too. My test is using a proprietory ESR meter and I have found provided the caps concerned come from quality manufacturers, they're usually fine, whether they've been in service for years and years on end or not. Sometimes I do find the odd duff electro in a 40-50yr old piece of equipment, but it's really not *that* common. So to answer the OP's question, I would *never* throw an old electro out

*unless* it tested faulty.

--
This message may be freely reproduced without limit or charge only via  
the Usenet protocol. Reproduction in whole or part through other  
protocols, whether for profit or not, is conditional upon a charge of  
GBP10.00 per reproduction. Publication in this manner via non-Usenet  
protocols constitutes acceptance of this condition.

These cases are molded with an outer and an inner part with overlapping lips. A series of sharp raps delivered to the inner side adjacent to the weld via a flat screwdriver blade breaks the weld but leaves the lip intact. Go right around and the case falls open, but can easily be glued shut again.

Clifford Heath.