Electrolytic capacitors

[timmy]

In a solid state circuit factory capacitors that are 46 years old test good 100uf tests at 136uf should they be changed or do they rise up in value with age or they could have been higher when new and still are.

[init4fun]

Originally Posted by timmy

In a solid state circuit factory capacitors that are 46 years old test good 100uf tests at 136uf should they be changed or do they rise up in value with age or they could have been higher when new and still are.

Hi Timmy,

Since it's said that electrolytic capacitors can be as much as +50% (from a quick internet search)

"What is the tolerance of electrolytic capacitors?
Electrolytics are often rated at +50 -20% or similar wide tolerances. They are rarely used where an exact value is important."

Your +36 over figure does fall within the acceptable tolerance.

As you likely know, there are two VERY sharply divided schools of thought, one being that ALL old electrolytics aren't to be trusted, and the other being that if the cap is still within tolerance, and isn't showing too much leakage current within itself (as shown by running too hot & loading down the circuit's voltage too much) then it's ok to run.

I personally do replace them just as a matter of not wanting to deal with their possible failure, and the potential for damage to other circuit components if they do fail.

[ppppenguin]

init4fun's comment is entirely correct but misses an important point.

If a capacitor reads much higher than its marked value then you are probably using a modern tester and the capacitor is leaky. This is simply due to the way modern testers do the measurement. Effectively they charge capacitor through a resistor (or with a constant current) and measure the rate of rise of voltage. Any leakage slows down the charging and gives an erroneously high reading.

With an electrolytic this may well not matter - they all leak to some extent. With any other sort of capacitor then a high reading means it's leaky and should be changed.

A traditional capacitance bridge will give a correct reading but not many of us use them for routine measurements.

[djski]

Originally Posted by timmy

In a solid state circuit factory capacitors that are 46 years old test good 100uf tests at 136uf should they be changed or do they rise up in value with age or they could have been higher when new and still are.

I wouldn't worry about a small increase in capacity. It probably measured that way on day one. Electrolytics decrese in value as they age until they just go open.

[Tube TV]

I don't really look at the UF too much as I have seen countless unusable caps in terms of leakage or high ESR that read dead on. What I'm always on the lookout for leakage or high ESR or dead shorts. If you're just testing capacitance on a multimeter and using it to determine if it's failed, you're pretty much wasting your time and you need at least a basic ESR meter for the low voltage stuff or a capacitance bridge like a Heathkit C3.

[Penthode]

Before condemning the electrolytics, I make the effort of at least first trying to reform the dielectric and thoroughly testing old electrolytics especially leakage. I have found before powering up and restoration, the electrolytic capacitors, having remained dormant for decades, will be leaky and the capacity risen. I attributed the rise in capacity to the diminished or thinning dielectric. Certainly after the capacitor dielectric has been reformed, the capacitor will measure more closely to their rated capacitance printed on the can.

I am surprised the high percentage of 60 and 70 year old capacitors which have survived. My RCA 1947 721TS l made a point of very slowly reforming each of the electrolytics over the course of a week or so. I feel because the set had not been powered on for decades and the set had for over 50 years stored in a heated attic, all of the chassis mount electrolytics survived.

I restored the set in 2020 and five years later I regularly power it on and the capacitors remain fine. It is nearing it's 80th birthday!