NOS Capacitors and aging

[Paul Ebert]

I'm looking to purchase a set of electrolytic capacitors for an amplifier power supply. One of the options is to purchase a set of NOS caps. These might be as old as 7 or 8 years old. Is this a problem? If so, how big a deal is it?

Thanks.

Paul Ebert

[jacco vermeulen]

It's like military rations: some say they still taste good 10 years after the expiration date, others view them as bad to begin with.
Should be ok, if you make the effort of reformatting them.
I'd do that even if the capacitors had been on the shelf for 7 weeks.
(reformatted 45 year old high voltage army capacitors and they came out 100%)
Capacitors dry out, you can shake them to check if and how much they are loose.
It's like The Mummy: if he isn't dried out completely he'll regenerate, otherwise you can huff and puff and it will still remain a corpse. Reformat electrolytic capacitors gently and they are able to swell like the Schwarz, just a few drops of steroids at a time.

[KBK]

Most folks don't have a clue of what you are speaking of. I've done a few old mallory sectoined caps before, but it's been a long time, and maybe I didn't do it right.

Youy'll have to explain it to them.

[senderj]

jacco,

This is the second time I heard about reformatting an old cap. Can you briefly tell us how to do it?

What I understand about electrolytic cap is it has + and - terminals. But I saw some has 5 leads, marked 1 to 4 and a -. The "label" only mentioned one cap value and one voltage. Is it 4 cap inside with same values, or 1 cap that can use any one of the leads to connect?

[SY]

It's usually called "reforming." Basically, you charge the cap up slowly to reform the insulating electrolytic layer. It's done with a power supply set to the cap's rated voltage and a series resistor. You want to limit the current to just a few milliamps; if a cap is leaky, the series resistor will keep it from getting too hot and blowing up. Hopefully. You still want to take precautions (goggles and distance).

[jacco vermeulen]

Originally posted by senderj
But I saw some has 5 leads

Besides regular 2-pole capacitors there are types that house 2 capacitors in 1 shell. 2 poles belong to the first cap, the other 2 to the second.
A long, long time ago there were 4-pole Frako electrolytics, 1 capacitor with 4 attachments. Currently Jensen in Danmark is producing simular 4-pole capacitors.
Some electrolytic capacitors have additional legs for more secure mounting. Roederstein EY series is a wellknown one with a third leg, Siemens B43306 series was an example of one with 2 additional mounting legs.

Reformatting was the term someone used a couple of days ago as a joke, thought it was funny.
The electrolyte in capacitors is moist in new condition, aging dries out the electrolyte. The capacitor shrinks because of it and becomes smaller than the outer shell.
Tube folks can generally be recognised by the shaking of caps next to their ear.
By slowly charging a capacitor through a resistor the electrolyte regenerates and increases its volume. With increasing volume the capacity of the cap also increases. I've found that it takes time for the electrolyte to reform. With really old capacitors my routine is to only charge the capacitor to half of the nominal voltage, slowly uncharge it after some time, do that a couple of times before charging to full voltage.
As mentioned by Stuart, slow charging is also safer, especially with high voltage capacitors. With faulty capacitors the outcome is less spectacular.
10 mA charging current is the value i use, a variac and a different resistor value switchboard make reforming very easy. (way back that was a piece of wood with nails hammered in, and crocodile clips)

[AndrewT]

Hi,
reforming at 10mA sounds quite fast. This is reforming NOT slow charging.

I use 1M0 in series with a cap. Initial voltage about half cap rating then ramp up the voltage as the leakage proves to be low.

You can monitor leakage by measuring the voltage across the resistor once the cap has charged up (at a few uA the cap will never reach the charging voltage). It will take half a day or so the reach half voltage and a couple of days for the whole process, if you do it slowly.

Multiple resistor + caps in parallel and you can spot the leaky ones.

There are a few sites that detail the process and at least two are in the forum.

[jacco vermeulen]

Originally posted by AndrewT
reforming NOT slow charging

The 10mA was not a reference to the reforming but for checking which are the heroes and which are fonies. I don't see the point in reforming a capacitor for a day to find out it's a goner.
Charging a 1000uF/25V with 10mA takes 12.5 seconds, that is not even slow charging imo.
The last years of my diyA-ing i've been using big cans mostly, 10mA at half the voltage rating is slow and safe enough not to kill potential candidates and it's fast enough to do a selection of the spitters.

I recall answering your question on the time it takes to charge a capacitor.
Fulling charging a 10.000uF capacitor through a 1Meg resistor takes more than half a day. Using a 1M on the 78.000uF Mallory's i have would take almost a week. (time to charge a capacitor = 5*R*C)

I gathered that when the electrolyte of capacitors dries out and becomes thinner, the voltage the cap withstands becomes lower. Through C=A*S/V, the capacity of the capacitor decreases with lower max voltage. This happens with old capacitors but also with new ones if they are not regularly charged to the voltage rating.

For the reforming job, i prefer to take a reforming period of 12 hours to 100% nominal voltage load. This timeframe and the capacitor value leads to a charging current and resistor value.
For the 9" big can Mallory's i have a few boxes of that gives a 100k resistor.