Waterboarding and other enhanced interogation methods for capacitors

[DC1MC]

So after suffering some real pain with crappy elcos and tantalums, most of them bought via large distributors and not aliexpress, I'm thinking of stating to waterboard these little terrorists  , or somehow enhanced interrogate them, I mean of course I have some LCR for measuring the value, ESR and whatever this "loss" coefficient means at the couple of volts applied, but this is like a normal police inquiry, i want some Bush/Cheney level of interrogation, that is, applying voltage close to nominal with an adjustable current rate, pumping some current in and out and generally testing them NONDESTRUCTIVE to sort the best ones close to working conditions.
Of course, if some of them die during the investigation, then it was due to a pre-existing condition 

So first question is, of course: is there any king of capacitor exerciser/testing device that does this and could be afforded by a hobbyst ?

Second is for the experienced engineers and scientist here, how would you do this, are there any papers, I've seen some regarding artificial and accelerated aging, but I don't want to kill them, just to give them a bit of difficulties, you know, what doesn't kill you makes you stronger (or cripple) ?

 Cheers,
 DC1MC


 

[CJay]

Fill 'em up with half wave rectified DC at some high percentage of rated voltage and stick a load across them that draws some current then measure temperature and ripple.

[GreyWoolfe]

I don't want to kill them, just to give them a bit of difficulties, you know, what doesn't kill you makes you stronger (or cripple) ?

break out the electrodes! Muahahahahahahahahahaha!

[cvanc]

Fill 'em up with half wave rectified DC at some high percentage of rated voltage and stick a load across them that draws some current then measure temperature and ripple.

This sounds like a pretty good idea.  Set it up, leave it for an hour, and see how much it heats up.

Better to use a beefy transformer so the charging current spike is nice and tall, then a low value resistor to pull the current out fast.

[DC1MC]

Besides measuring the temperature increase, should I do some other measurements before/after ?
Any parameters that are going to significantly change, dI/dt limitation, different frequencies to try for different values (a 10uF at 25V may not care too much of 100Hz) and so on ?

Also, there is no such tester already available ?

 Cheers,
 DC1MC

[T3sl4co1l]

What types have you purchased, and what applications are you using them in (specifically, not just "oh this and that")?

There are many types available, you may've just picked the wrong one.

I haven't seen any particular reason to doubt the datasheets, at least among name brands from large distys.

Tim

[CJay]

What types have you purchased, and what applications are you using them in (specifically, not just "oh this and that")?

There are many types available, you may've just picked the wrong one.

I haven't seen any particular reason to doubt the datasheets, at least among name brands from large distys.

Tim

Mhmm, the application is king, if you're choosing capacitors for smoothing/filtering in PSU applications then checking the ripple current as I suggested above would be a useful test, temperature rise from self heating would be a potential indicator of quality and possible longevity.

if you want them another application, signal coupling perhaps, then

 A. WTF are you thinking?

 and

B. Ripple current rating is a fairly useless specification, there are other tests that would be more useful.

[David Hess]

For the tantalums you could surge test them.  If they are sold tantalums and you have no better options like voltage derating or using a different type of capacitor, then they could be surge tested after soldering if the circuit is designed properly.  Polymer tantalum capacitors avoid this problem.

Other than that without more information on the problems you are having, it is difficult to say what other tests would be appropriate.  Leakage testing is the easiest and might be combined with burn in testing but my result with this have been marginal.  Network analysis to measure the impedance up to and above the resonate frequency will reveal subtle problems that single frequency ESR measurements will not.

[ogden]

So after suffering some real pain with crappy elcos and tantalums, most of them bought via large distributors and not aliexpress, I'm thinking of stating to waterboard these little terrorists 

Before you torture potentially innocent subjects, you shall properly investigate previous crime scenes. Such way you can get additional clues how exactly you shall torture poor thingies or reconsider to not torture.