Older electrolytic capacitor testing

  • Thread starter artis
  • Start date
194
13
So some time ago I got a bunch of electrolytic caps from a friend of mine who works at a local university experimental laboratory,
the caps have been in good storage (temperature etc) they were sitting some while in my room and now I finally decided to see whether they are still any good.

Some are manufactured in 1983, some in 85 and some in 89, so all are older than 30 years.
8 pairs of 50V 22000uF , two pairs of 80V 15000uF, and two pairs of 50V 10000uF, there are two boxes full of other smaller values like 50v 1000uF and so on.
So what i did is this , I measured the capacitance with a multimeter , surprisingly no capacitors were below their rated capacitance value, some were much higher one was showing 38 000uF.
Then for some days I took a car battery and charged them and discharged them with a low value resistor, charged again and discharged again and so on.
Measured the capacitance values once more and they were all roughly the same as before.

So then I charged every capacitor to around 12.8 Volts (the voltage of my car battery) and left them overnight,
this morning I took my voltmeter and checked everyone of the capacitors.
Some of the capacitors have retained almost all the charging voltage (around 10,11 volts) while others are lower to around 3,4 volts, while a few had only mV left in them.

the question is what does this suggest about the capacitors ? Any commentary is welcomed.

I personally feel that the caps that still have almost all the charging voltage are really good while the ones that lost it overnight are probably still okay (they measured to rated capacitance) but have a higher "leakage" probably due to age?
 

Tom.G

Science Advisor
2,506
1,341
30 years is approximately the end-of-life for commonly available electrolytics in storage or light use. The failure mode is that due to imperfect sealing, the liquid electrolyte dries out. Any of them showing evidence of even minor leaking should be discarded.

I agree that the ones that held a charge overnight are candidates for being usable in non-critical applications, like filter caps in an experimenters power supply. You might as well trash the ones that did not hold an overnight charge or had a measured capacitance below 50% of marking. (typical tolerance for Aluminium electrolytics is -50% to +100% of value) Don't expect any of them to function well in more critical applications like interstage coupling, time delay circuit, or in a tuned circuit.

Note that electrolytics that have not been used for several years tend to both lose capacitance and increase leakage current due to internal chemical reactions. If they haven't dried out, they can sometimes be 'rejuvenated' by applying rated voltage to them for a few hours or days. It's usually not worth the trouble though. If you decide to try it, put a resistor in series with the cap to limit current to a few 10's of milliamps. That way if the cap shorts it won't get violent and smell the place up. They really stink when they blow!

Perhaps you can borrow a capacitance meter from someone to measure the dissipation factor and capacitance. Or if you have spare change in your pocket and want to build up your inventory of test equipment, good capacitance meters are available for around US $150. There are lots of lower cost kits available but I have no experience with them. Just make sure any meter you buy covers at least the range of a few hundred pF thru a few thousand uF, the wider the better... and more expensive.

Enjoy!

Cheers,
Tom
 

jim hardy

Science Advisor
Gold Member
2018 Award
9,750
4,750
Here is a good paper on electrolytic capacitors.
http://www.cde.com/resources/catalogs/AEappGUIDE.pdf

page 7 has instructions for testing old capacitors.

As that document describes
The dielectric in aluminum caps is a layer of aluminum oxide just molecules thick and that's why they have so much capacitance per unit volume.
C = εA/D and D is tiny.
In storage the dielectric can degrade and become too thin to hold against rated voltage. They'll leak off voltage then.
In the "old days" we'd "reform" the dielectric by applying rated voltage through say a megohm and measuring voltage.
As leakage drops off due to dielectric coming back voltage across the cap increases.
When leakage is down to spec the cap is probably okay.
Unless it's lost capacitance because the dielectric has dried out , as @Tom.G said.

download and print that CDE guide - it's a treasure.

old jim
 
194
13
Well first of all sure I'm keeping them just for experiments etc, from my past experience such caps work well enough for a makeshift power supply to test a amplifier board etc.

With regards to the dielectric, these are old soviet made in various factories in Russia, I know that from the markings on them. Although I haven't opened any one of them up I have done so in the past with similar dead caps and these use "paper" soaked in electrolyte as the dielectric not any oxides so as far as the paper itself goes it's almost eternal because the low capacitance high voltage caps that were just paper and no electrolyte can be good for 50 and more years unless damaged mechanically. But I can't be 100% sure because maybe these use oxide but the smaller ones I opened up years ago had simply aluminum foil wrapped with paper in a roll and the whole roll was wet and smelly which was the electrolyte so hard to tell, also can't find any documentation on these.


Anyway since I have a spare traffo and some diode bridges I will just hook them up and let them "work" for a while and see maybe they all come back to being good maybe not.


On a sidenote , even if some of the caps leak small amounts of current I assume they are still fine for a power supply filter because the leakage current is small and if they have rated capacitance they will still hold charge for the small timeframes involved between cycles of AC, so I suppose they are usable in this regard.

PS. Ebay sellers write their aluminum oxide.
 

jim hardy

Science Advisor
Gold Member
2018 Award
9,750
4,750
"paper" soaked in electrolyte as the dielectric not any oxides
That electrolyte soaked paper is the other "plate",.
Its wet solution provides conduction .between the cathode foil and the aluminum oxide dielectric on the anode foil..

see page 4.
 
Last edited:

Borek

Mentor
27,963
2,470
Chemist remark: any piece of aluminum foil (unless made in controlled conditions in an inert atmosphere) is covered with the layer of oxide. This layer is thin (can be made thick through anodizing), hard, durable and protects the metal below from further oxidation. If you machine the aluminum bare metal reacts with the air oxygen faster than you say "no oxide here" :wink:
 
4,236
39
These are the kind of caps we use in audio amplifiers. I talked in forums before, consensus are electrolytic caps do age even if you just store it away. 30 years is a long time. My advice to people there is if it is in the amp, the amp is 30years old, just power up the amp and use it, if you don't see anything leaking, the top don't have any bulge and it doesn't smoke, don't change it, use until it blows. But I definitely will not put it in new builds.

These kind of caps are mostly used in power supply after the rectifiers to smooth out the voltage. YOu usually see a saw tooth wave on it. You look at the amount of over shoot at the leading edge. Too much over shoot implies the ESL or ESR are high. But I did test with brand new caps ( new, not old new caps), they do show over shoot. So I am not sure this is a reliable test either.

Bottom line, I won't use them for my new amps. For experiment to play around, sure.
 

jim hardy

Science Advisor
Gold Member
2018 Award
9,750
4,750
from the CDE catalog linked above


upload_2019-2-22_17-5-16.png



that's why they have so much more capacitance per unit volume than a dry paper&foil type.
 

Attachments

194
13
I think I get it , electrolyte being a liquid can come in contact much closer and get into all the countless pores that have formed on the anode oxide dielectric surface, any solid surface could never come into such close contact microscopically so the capacitance would be much lower, so when the electrolyte dries up the capacitor may still work and have its voltage rating but will have decreased capacitance rating.
yesterday I did some measurements and discarded for recycling some smaller caps that showed signs of being "dry".

As for the large caps , I made a quick setup of a traffo and bridge rectifier to charge them to rated voltage. most did fine and achieved that voltage after a few charge/discharge series, but some seems like have a large leakage current and can't get up to the max voltage, after a while they get a bit warm to touch while others remain dead cold so I think that is a suspect that those might be unusable.?

Although doing the test today the caps now come up to voltage fast enough and are not warm anymore, maybe there was some reaction going on and now they "healed"
 

jim hardy

Science Advisor
Gold Member
2018 Award
9,750
4,750
but some seems like have a large leakage current and can't get up to the max voltage, after a while they get a bit warm to touch while others remain dead cold so I think that is a suspect that those might be unusable.?

Although doing the test today the caps now come up to voltagyou might have 'reformed'e fast enough and are not warm anymore, maybe there was some reaction going on and now they "healed"
You might have "reformed" the dielectric .
It'd be best to do that through a high series resistance so it proceeds slowly and give them a day or two,. monitoring voltage across the resistor to measure leakage ..
I've had them arc over inside if done too fast. Sounds like a firecracker in there.
 
194
13
Well I don't know about arcing inside but the few caps were just barely warmer than the others which were cold, its not like they were hot to the touch.
I'm using an under powered traffo , now it seems they aren't getting warmer any more , also the voltage holds better but all in all its not that important for me to have a few bad ones, this is more of a test and learning for me than trying to salvage each and every one old cap.
out of the 10 capacitors , 2 or 3 seem suspicious the others are fine.
 
194
13
By the way , how long does an average used or older electrolytic has to store charge for it to be considered ok or usable? Because I think that if it loses charge quickly that is a sign of large leakage over the dielectric or otherwise.
So say I charge them to half the rated max voltage or something how long would it be ok for me to still measure the same voltage, hours , days?

I know on good caps I have measured after a week or longer and they still hold the same voltage that I charged them with or a bit less but still a considerable amount.
 

jim hardy

Science Advisor
Gold Member
2018 Award
9,750
4,750
I would reform them for about a week
through a megohm of resistance at rated voltage or perhaps 10% more
and expect them to discharge with a time constant equal to Capacitance X ## Rated insulation resistance ##
clearly a 22,000 microfarad will bleed down slower than a 220 microfarad.

practically speaking though if leakage current is in the microamp range they're plenty good for power supply filters.
If they hold overnight i'd declare victory.
 
4,236
39
As I said before, if it is for experiment or some bench power supply, just use it. I think you test enough already. I don't think you will get anyone to tell you they are as good as brand new. These are not very expensive caps, just use it. I bought those 10,000uF 63VDC caps by the 100s, they are $3 each. Caps have improved and getting smaller now, I won't even use those that are 30 years old even if they were brand new now.

I would not put it in my amplifier, I don't want to have any long term reliability problem. But for just experimenting, it's plenty good. No need to keep beating on it.

FYI, I bought extra 6 of the 33,000uF 100V screw top caps 3 years ago when I recap one of the old power amp I have. It was a good price of like $30/each( those are very expensive). But later I decided that using 4 10,000uF smaller caps work a lot better because I can get much lower ESL and ESR with 4 smaller ones in parallel. I never design using those 33,000uF caps anymore. They are still sitting in the garage. In another 10 years, if I still don't find any use, I will just dump it. It's getting old after 7 or 8 years and it's not worth keeping anymore. I don't do experiment that can put it to use, and it's too old to put it in a real design. You talk to people that repair amps, when people want to use an amp that's been sitting around for years, first thing first is to change out all the old electrolytic caps. Even if they are not being used, they age.
 
Last edited:

Want to reply to this thread?

"Older electrolytic capacitor testing" You must log in or register to reply here.

Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving

Hot Threads

Top