When should I toss old unused electrolytic capacitors?

[mcovington]

I'm reactivating my home lab, which hasn't been used heavily for several years.  I have a good stock of electrolytic capacitors ranging from 5 to 30 years in age (mostly over 15).  I know an electrolytic can't be sold as new if it's more than 2 or 3 years old.

If an old electrolytic tests OK for capacitance, leakage, and ESR (all after charging it through a resistor to re-form it), does it have a long life ahead of it, or is it still unreliable?

I am wondering what to throw out and how much to restock.  My goal is more to get reliability than to save money, since I'm (thankfully) no longer on a student's budget, and I do both repairs and breadboarding. 

Should I throw out everything over 10 years old?  20?  5? 

[TheMG]

At work we have quite a stock of capacitors for repair purposes, some well over 20 years old. I have not had any reliability issues with using such old stock.

Do take note that these are generally all name-brand good quality capacitors (Cornell Dubilier, Nichicon, Rubycon, etc).

When stored in a climate controlled environment, away from excessive heat, and the fact that they are not being subjected to any electrical or mechanical stresses, there really should be no issues at all for quite some time (decades), despite the manufacturer's suggested shelf life which obviously errs on the side of caution.

I also have never manually reformed capacitors, just take em out of the parts bins, solder them into circuit, and apply power. Never had any issues.

[coppice]

Most components need to be used within 2 or 3 years from manufacture, to meet the manufacturer's recommendations. Its not an issue of internal degradation. Its a matter of solderability. Even stored well you can't just trust the solderability of long stored parts on a production line. When hand soldering you naturally correct for harder to solder parts. Stored at room temperature it normally takes many years for the internals of an electrolytic capacitor to degrade. Its heat that kills them.

[mcovington]

That makes a lot of sense.  I had noticed that older components need a couple of swipes with sandpaper on the leads -- no problem when I'm assembling things one at a time.

[Shock]

If any component is still within specification and you are not manufacturing products or building spaceships etc, then there is no need to throw the component out.

[SparkyFX]

Certain packages attract moisture over time, which then leads to boiling and popping when exposed to e.g. hot air soldering. Those would need to be baked at ~50°C for some time (usually mentioned in the datasheet) to get them dry again. This mostly affects SMD, through hole parts that are only soldered on the leads might not require this.

[factory]

I binned a few new old stock capacitors at work where the seal had failed & electrolyte was leaking or corroding the leads, some were Rubycon & Panasonic (about 15-20 years old). This had also happened with some wet tantalum capacitors.

David

[tooki]

That makes a lot of sense.  I had noticed that older components need a couple of swipes with sandpaper on the leads -- no problem when I'm assembling things one at a time.

Try a fiberglass eraser "pen" instead -- you may never go back to sandpaper for this application!

[bsudbrink]

I'm going to put a "shameless plug" in here for anyone here or anyone in the future that might find this thread...

If you find old electrolytic capacitors with the following markings:

       +M+
2000 MF 10 V
   20-48923

working or not, in just about any shape (leaking, no leads, etc.), I will pay you something for them.
See attached picture.

[Seekonk]

I have probably a thousand 200uF 10V no name electrolytics I picked up at the Dayton Ham Fest in the mid 80's. So crappy they left them for free at a garbage can. Every once in a while I'll pull ten out and test them when bored. They still pass every test.

This is just a shameless post so I can find the above post later on.

[bsudbrink]

Note that the ones I'm after are _2000_ uF not _200_.  But thanks for looking.  They are from 1974 so they use the old MF notation.

[TheMG]

This had also happened with some wet tantalum capacitors.

Wet tantalum capacitors were notorious for that, the electrolyte was acidic and would eventually eat away at the leads and case.

I'm not sure if more recent wet tantalum capacitors have gotten better but nevertheless they have become extremely rare. Haven't seen anything made after the 1970s with wet tantalum caps. They were very common in military and aeronautical electronics prior to that.

[coppice]

I'm not sure if more recent wet tantalum capacitors have gotten better but nevertheless they have become extremely rare. Haven't seen anything made after the 1970s with wet tantalum caps. They were very common in military and aeronautical electronics prior to that.

Military grade wet tantalums were always very long life in terms of leakage, although they still suffered from the same tantalum whisker problems as any other tantalum capacitors. By the end of the 80s dry aluminium electrolytics had displaced them from most telecoms applications, mostly to avoid the intermittent short problems crashing telephone exchanges. I'm not sure about the military.

[TheMG]

Military grade wet tantalums were always very long life in terms of leakage, although they still suffered from the same tantalum whisker problems as any other tantalum capacitors. By the end of the 80s dry aluminium electrolytics had displaced them from most telecoms applications, mostly to avoid the intermittent short problems crashing telephone exchanges. I'm not sure about the military.

I have a few bits of gear from the 60s and 70s with mil-spec wet tantalum capacitors, as well as I used to have a box of NOS capacitors. Over 50% of the wet tantalums in the equipment, as well as the NOS ones, were failed. I'm sure they were very reliable during the service life of the equipment in question, but they certainly have not survived several decades of practically non-use. See attached image for the type of wet tantalum cap I'm referring to. The failed ones all end up having a brown band visible between the metal casing and the red epoxy, which is when the electrolyte has eaten through the case and is starting to leach out.

Newer stuff (80s and onward) that still use through-hole capacitors seem to use solid tantalum "Tantalex" style capacitors. Mil-spec M39003 to be precise. They are ridiculously expensive but surprisingly rugged and reliable, I maintain some equipment at work that uses a lot of these capacitors, many of which are exposed to high temperatures 24/7 as they are part of linear RF amplifier modules that run pretty hot. I have yet to find a bad one, let alone one that has gone out of spec!!!

[coppice]

I have a few bits of gear from the 60s and 70s with mil-spec wet tantalum capacitors, as well as I used to have a box of NOS capacitors. Over 50% of the wet tantalums in the equipment, as well as the NOS ones, were failed. I'm sure they were very reliable during the service life of the equipment in question, but they certainly have not survived several decades of practically non-use.

A lot of 60s and 70s military equipment hasn't reached the end of its service life, including a lot of things put in storage after delivery which have yet to be switched on.