how quick to electrolytic capacitors take damage?

[coppercone2]

I accidentally soldered a bunch of 35 v capacitors on a rail that is ~45 volts, the components were different from device to device.

They are brand new and I want to install the proper caps, but are these ones DC-biased at 45V ok to use for other stuff? They were ran for like 2 minutes only.

I figure its safe to say they have 95% of their life left for use in other prototypes? I don't want to throw them away because its 4 capacitors and they are a little pricey. ($15)

I did hear a 'whine' from the system for about 0.5 seconds after turn on, and I am not sure what that was from, because I replaced every capacitor in it, but these 4 are improper. Its a sorensen linear supply. Everything looked good and it operated fine for the short test (short circuit current 1A for 30 seconds and some time without load playing with the knobs).

[wraper]

They should be fine. 35V is working voltage, withstanding voltage is higher.

[Ian.M]

How hot did they get?  If they didn't overheat to the point where any of their electrolyte boiled, they'll be fine for any application except the most stringent safety critical stuff.

[coppercone2]

yea lol for safety critical I would only install capacitors once anyway, it would not be heated twice with iron..

they did not get hot or anything, I only noticed because I decided to double check my work, and it needed to be taken apart again anyway because i ran it without thermal grease just to make sure I did not break it during the repair.

The graph says their life at that voltage is 20% only. I assume if you use the percent life vs over voltage graph, you need to operate it in the 'several hours' down from 5000 to quickly damage them? They are over sized also, I replaced 3000 uF with 10k because it was convenient to repair both power supplies with the same parts (I thought the 3300uF was just old & big capacitor, but it turns out they have a higher voltage rating).

Mainly I want to put them in to the other supply now, but I want that to last because I would need to re-do thermal grease to fit new ones in if these break

[Gyro]

On the up side, they probably have really low leakage at 35V now! 

[coppercone2]

I will compare the ESR to fresh ones and install them in the other circuit ( i expect its fine) then

[Hydron]

Would be worth having a look at the capacitance too.

Gyro mentions leakage as it may have increased the oxide layer thickness by being re-formed at 45V. Would be interesting to see if that has changed the capacitance too (easier to measure than leakage, which would hopefully be in the low nano-amps range or less, though using a DMM in a mV range as a 10M current shunt is an option).

You're probably fine though. I did something worse the other day - connected up a 5600uF 450V cap (worth far more than $15 new, though I got it free) the wrong way around at 10V bias or so, though thankfully with only a 10mA current limit. Turns out the stripe on the label is for the POSTIVE terminal, opposite to ~every other electrolytic . When I fixed the error it still measured fine though (capacitance just over 10% down from nominal, i.e. still within spec, and it was potentially already well-used), with leakage fine up to 400V (max I can easily test until I upgrade my Keithley 236 SMU to a 237).

[Conrad Hoffman]

There should be a peak or short time spec. They likely saw that voltage during manufacture. If it was only a couple minutes, and nothing overheated or shorted, I wouldn't worry at all... unless the project is going into space.

[coppercone2]

it wont go into space because I don't want to put silastic around it anyway, enough trouble to scrape it off