Replacing Electrolytics with NOS parts

In my experience duff electrolytics are usually obvious from the electrolyte ooze. I was fooled once by an old Leo3 computer power supply unit that had four 500uF electrolytics in parallel. It smoothed well until the last one failed.

It has been my experience that electrolytic caps of a certain age are as likely to fail just sitting on a shelf as in service. I suggest you purchase modern, fresh caps of the proper size, voltage and style from a reputable source made by a reputable manufacturer. Why risk it?

Peter Wieck Melrose Park, PA

Personally I've not had any problems with NOS electrolytics provided they were sourced through RS or suchlike originally. But YMMV.

On Monday, August 28, 2023 at 8:25:21 AM UTC-4, snipped-for-privacy@outlook.com wrote: I was fooled once by an old Leo3 computer power

The first time my Tek 7603 mainframe crapped out, one of the secondary supplies was quite low. This unit had like 6 huge Mallory caps in parallel and I figured it couldn't be the caps, but like yours, it turned out they were all open when I used my old B&K to troubleshoot it. All six. They must have been dying off one by one with absolutely no evidence seen on the trace until the last one croaked. I replaced just one of them to get me going and used it several years like that until another group of caps did the same in a different secondary of the power supply. At that point, I ordered all new electros and recapped the whole power supply.

Years ago, I tossed out my NOS electros because testing them was uncovering some leakage at voltage. I don't recall the brand names if you're wondering that but brand new quality caps are so cheap that it didn't make sense for me to use the old ones.

For temporary work or for something that only has to last another small handful of years at most? Go ahead as long as they don't show leakage or abnormal ESR.

Electrolytics tend to get leaky over time if stored with no voltage across them. When voltage is applied after a long time in storage, an unusually high leakage current will flow until the insulating film reforms. If they have been stored long enuogh, this re-forming process can draw a lot of current and make the capacitor get hot and go bang. For this reason it is best to do it with a current-limiting resistor. If you are intending to store electrolytic capacitors for a long time, it might be worthwhile making provision to charge them periodically to prevent them from becoming leaky. Similarly, it is worth powering up stored equipment on a regular basis to re-form the capacitors in the equipment so that it doesn't go bang when powered on after a long period of being off.

It's a transient leakage, probably. The manufacturer poles the electrolytics by deliberately feeding leakage current, to drive the chemistry to form the oxide layer, which... doesn't leak.

After storage for months to years, the oxide layer can thin or perforate, but biasing it up (and leaving it at voltage for a few minutes to hours) rebuilds the oxide. NOS items don't meet like-new specs straight off the shelf, but usually DO work normally, by day two under bias. I've also used a curve tracer to watch the leakage, and applied hot air to speed it up, and re-formed electrolytics back to good-as-new while watching the leakage current dwindle.

Note, for the record: When I purchase new caps, they still get tested for ESR and value prior to installation. All of them.

Now, here I am with a choice: $20 for a high-value, high voltage electrolytic cap. Or, $0 for a NOS one that, although never installed or used, has been sitting in my closet, or $8 for one that has been sitting on some dealer's shelf (properly conditioned, of course) for the last 20 years. I do all that has been suggested in this thread to bring that NOS cap up to proper behavior. Or so I think.

That cap goes *POOF*, and takes some unobtanium parts with it.

What have I saved?

Peter Wieck Melrose Park, PA

Sorry, it seems you've been out-voted. Re-formed & tested NOS wins. :-)

Majority voting is mob rule - doesn't make it the right decision. Even so, I think if you did an actual poll, you'd find going with new production would win.

These aren't $50 parts here. Even quality capacitors are cheap. It makes little sense to reform a cranky electrolytic.

Why would an old stock capacitor fail? If it was battered, or stored next to the fuming nitric acid, or poor quality to begin with, maybe. But, if it's just normal aging of non-biased electrolyte, it's NOT gonna fail, just will take an initial hour or three under bias to reform the oxide.

You've saved, in 98% of the probable cases, the time it takes to look up mechanically and electrically suitable replacements, and the overhead costs of shipping and receiving it.

The unobtanium parts are fictitious, I won't address nonissues.

If this were about a production run of 2000 units, with a deadline, and you couldn't take time to evaluate the whole batch, it'd be a different decision.

Electrolytic capacitors depend on the seal for their life.

Seals aren't perfect and eventually the electrolyte dries out. Then the capacitor fails permanently, reforming isn't possible.

Manufacturers recommend a shelf life of around ten years.

These days, due to what I'm seeing in vintage equipment, capacitors from the 1980s and earlier simply can't be trusted.

Caps from the 90s and 00's are suspects, but not yet time for blanket replacement. Except in monitors and switching power supplies...too warm an environment for the seals to last.

John :-#)#