A Question on Aging Effects on Electrolytic Capacitors

It's not unusual for a capacitor to have a tolerance of plus 50%. Do a parametric search from a distributor and you'll see for aluminum electrolytics that -10% to 50% is a common tolerance.

Interesting trying to find what the tolerance is for these.

I did find this about pre-conditioning here https://en.wikipedia.org/wiki/Electrolytic_capacitor

For antique radio equipment or for electrolytics built in the 1970s or earlier, "pre-conditioning" may be appropriate. For this purpose, the rated voltage is applied to the capacitor via a series resistance of approximately 1 kΩ for one hour. Applying a voltage via a safety resistor repairs the oxide layer by self-healing. Capacitors that fail leakage current requirements after preconditioning, may have experienced mechanical damage.^[91]

It also talks about leakage current v time.

Sadly the markings are buried in a standard ... you know the one where someone wants you to pay for soemthing that should be free.

IEC/EN 60384-1 - Fixed capacitors for use in electronic equipment

My guess is that it will be 20% tolerance which only gets you to 396 uF ....

Mark

Hi John,

Interesting measurement! Agree, it could be tolerance related, also it might be very leaky, and if the capacitor meter is charging it, and waiting for the voltage across it to reach a certain value, it will take longer, so the meter could think it has a higher capacitance. Also interesting that the value that was measured changed after the first time. I wonder if fresh/new electrolytics do the same thing... plastic caps have dielectric relaxation effects, but I don't know if this applies to electrolytics too.

Nice post John.

I think you will find that the material separating the two plates of the capacitor has changed over time.

I suspect that the increased capacitance value was caused by the material shrinking slightly bringing the two plates closer together.

Please be careful with these capacitors because the increased capacitance also means that the electrolytic breakdown voltage has been reduced, so you can expect these capacitors to fail at lower than rated voltages.

Have fun testing these old capacitors, just do not be surprised when one fails.

DAB

Hi DAB,

Thank you for the insight. I had thought about the possibility of the metal crystalizing or fracturing thus increasing the surface area but I had not considered the distance between the plates decreasing. I will have to study a little more about the structure of electrolytics as I believe that the electrolyte itself serves as one plate and the metal the other with the insulating oxide layer as the dielectric. This would mean that the oxide layer itself may have devreased.

John

Hi John

When the electrolyte evaporates it can leave salt deposits on the two plates. These salt crystals can for conductive paths which greatly reduce the capacitance and voltage tolerance.

There has to be two conductive plates in every capacitor. The electrolyte is just a buffer fluid used to balance the insulating material to maintain a constant value of capacitance over the plate area.

DAB

I've just had a look around and found some old electrolytics.

From left to right (marked/measured)

1uF 2.034uF
33uF 46.08uF
64uF 90.39uF
10uF 23.26uF
1000uF 1198uF

So they all read consistantly high. The only one with a date code is the one on the right which dates from 1969. Curiously, that's the

most accurate, with a value that is almost certainly still within the original tolerance.

Hi Jon,

This is exactly the trend I have noticed. The readings on your first three capacitors are what motivated me to ask this question. Unfortunately there is no way to go back to see what the original values were so we do not really know if the original real values have gone up over the years or not. If we can determine a physical mechanism for this trend towards being much higher than nominal then we can understand it and capacitors better. Thanks for adding some weight to my observation.

John

An interesting read on the subject.

https://www.digikey.com/en/articles/techzone/2016/jun/what-is-the-capacitance-of-this-capacitor?WT.z_sm_link=twitter_Tzc…