wtf is with these caps?

[coppercone2]

I got bandolier with 220uF 25V axial caps from digikey today

I measure them at correct polarity on DE-5000 meter, I get readings like 3nF.

I charge em up on a power supply to 25V, and spark discharge them on a block of metal ( I see a substantial spark on charge and discharge (touch it to power supply terminals and steel block). I get a reading of similar low nF value when I measure it after discharge.

I try a regular cap, it reads fine in uF (others from the same order). Repeatedly, other capacitors.. are normal. I tried a new one, a ancient one from the junk box, the meter is fine.


But, if I take the axial one, and measure it even once on the LCR meter backwards (that is, can body on positive), it reads fine, and then it reads fine in both directions. So its like the LCR meter activates it, if you plug it in backwards.


I never saw this on a capacitor, they always measure fine when I put them into a meter right from the box.

What am I dealing with here? Is this like in any way acceptable or should I just chuck this shit out?

I guess their eSR is kind fo shit when activated too (0.7 ohms), its in spec but I expected less based on similar measurements
 

the magnitude of the 'spark' from tapping it on a steel block does not seem to change before or after 'activating' it with the LCR meter to make it measurable ....

[wraper]

My guess is you're using default 1kHz test frequency instead of selecting proper 100/120Hz, PEBKAC.

[coppercone2]

I said I tested other caps of the same value fine, the frequency is not the problem.

I got four left that I did not "trigger" with any experiments, they still read in the nF

pretty shitty for $4 capacitors

[wraper]

So what? Electrolytic capacitors are far from ideal and at 1kHz this meter is marginally able to measure such capacitance to begin with. Add non ideal nature of electrolytic caps, and here you go. Not to say capacitance of electrolytic caps is specified at 120 or 100Hz but ESR usually at 100kHz. All your measurements are garbage.

pretty shitty for $4 capacitors

Rather shitty to not learn how to use the meter properly

[coppercone2]

I am not measuring at 1KHz. I said frequency is not the problem.

whats shitty is your assumptions and advice. Its like useless spam

[coppercone2]

And your assumptions are nonsense too. I just tried it out for kicks, the meter will read a 1000uF capacitor just fine at 1KHz

 difference between 120Hz and 1KHz is 956uF vs 926uF, since you made this thread about 1KHz for some reason

maybe you should not be so quick with rules of thumb, accusations and assumptions so you can contribute something

you know what I even tried a 220uF cap at 10KHz. It still measures 170uF and the same ESR, if you were to measure it that way

[5U4GB]

whats shitty is your assumptions and advice. Its like useless spam

Somewhat off-topic but this is an ongoing problem here [X], in an average technical discussion 90% of people will contribute usefully and 10% will provide extremely condescending and mostly useless comments, "you're an idiot", "you're an amateur and don't know what you're doing", "maybe some day when you learn a bit you'll stop making stupid mistakes like this", and so on. I'm not sure if this is simply a trait of EE types but I've never noticed anything like this on other sites I'm on, HN, *Exchange, etc.

[X] Not meant as a criticism of the site but more an observation about why it seems so prevalent here compared to other sites.  Is it just EE personality types that naturally do this?

[timenutgoblin]

I got bandolier with 220uF 25V axial caps from digikey today

I measure them at correct polarity on DE-5000 meter, I get readings like 3nF.

I charge em up on a power supply to 25V, and spark discharge them on a block of metal ( I see a substantial spark on charge and discharge (touch it to power supply terminals and steel block). I get a reading of similar low nF value when I measure it after discharge.

Maybe you need to clean the leads with IPA and a tissue in case there is a thin film of contamination. Otherwise try scraping the leads with a scalpel or similar tool, maybe fine grade sandpaper if necessary.

[sw_guy]

I'm not sure if this is simply a trait of EE types but I've never noticed anything like this on other sites I'm on, HN, *Exchange, etc.

This is not something new and happens on other EE sites too. Especially one other forum seems to wait for new members to kick them down.
Years ago eevblog was much more disrespectful and the attitude was pretty bad. This was due to one hot topic back then (IMHO).
Sorry for the offtopic reply. My point here is that please respect the ones who ask questions and please respect them, who use their own time to share knowledge.

[Whales]

It's very strange that reverse charging the cap temporarily fixes the problem.

1. Do you have another ESR meter to test with, to rule out the caps being super strange?

2. Do you have an oscilloscope?  As long as you run the DE-5000 meter on batteries (and have no other wires/leads attached) it would be safe to attach an oscilloscope probe across the capacitor being tested.  The extra capacitance of the probe lead should be insignificant when measuring uF size caps.  Then you can compare how the tester behaves against your other (good measuring) caps.

[shapirus]

What does a DMM (i.e., DC measurement) show in each orientation?

[MariuszD]

Connect the RC circuit, generator, and oscilloscope. You will be able to see when and how the capacitor fails and how this depends on the voltage, which can be adjusted during the measurement. I suspect the connection between the pins and the foil is poor. Maybe it would be worth breaking open one capacitor to see what failed.

[magic]

Who said that there is anything wrong with the caps? If they failed open, they would be pico farads, not nano.

Looks like there is something unusual about them which confuses this meter.

It's very strange that reverse charging the cap temporarily fixes the problem.

Applying reverse voltage should "form" the caps in reverse and, for example, may reduce leakage under reverse bias.

[robzy]

I'm not going to pretend to be an expert - but do you have a 'scope you could hook them up to with a power source and resistor? How do they actually charge?

[PA0PBZ]

I got bandolier with 220uF 25V axial caps from digikey today

I measure them at correct polarity on DE-5000 meter, I get readings like 3nF.

I charge em up on a power supply to 25V, and spark discharge them on a block of metal ( I see a substantial spark on charge and discharge (touch it to power supply terminals and steel block). I get a reading of similar low nF value when I measure it after discharge.

a 3 nF cap does not produce a substantial spark on charge or discharge so clearly something is confusing your meter.

[PGPG]

I got four left that I did not "trigger" with any experiments, they still read in the nF

Could you check one of them with voltmeter?

[Zero999]

Have you tried building a low frequency oscilaltor (555, Schmitt trigger (74HC14), CMOS, dual BJT etc.) with them? Do you get close to the expected frequency output or is it much higher?

[coppercone2]

I feel sick today but a small update, the 'activated' caps still read OK and the ones that I did not reverse polarity into the de5000 read very low like pF or nF.

I had 10, 4 of them are still in the high Z condition, where there is no capacitance, and the other 6... measure like capacitors. Its not the leads, this stuff looks clean and measures fine with a DMM

So I think it will stay this way but I am not gonna be doing anything electronics today. It looks like every other cap that came in measures fine on first try

I can try other experiments tommorow

it this
https://www.vishay.com/docs/28334/118aht.pdf


I need to take my latest failed repair apart and check, because I put two of these caps from the last order into it, maybe3 thats why its not fixed. I did not happen to measure those before installing. maybe they are in circuit in high Z mode preventing operation and causing problems


I think its something with passivation but that is a guess

[mawyatt]

Send one over here if you want, we'll check with our DER5000, TH2830 and IM3536. Also could do a voltage vs capacitance graph(s), forward and reverse.

Best

[coppercone2]

hmm not something I normally use but I could not help it and tried to measure one with a fluke 289, it turns on when you measure it with caps (and it does measure a value),  and then it works on the De5000

I was not expecting that, anyway I did a experiment with two meters connected, in the same polarity, with DE5000 in parallel with fluke 289. It reads open (WIth de5000 connected and fluke connected on negative lead), and as soon as I momentarily touch touch the cap with the 289 positive, the DE5000 reads the correct value, but it does not work with both meters connected, but after triggering with 289, both meters read the correct cap value if one lead is disconnected


so I only have 2 'untriggered ones" left and I will think of what to do.


I think I saw something similar happen to a battery before where it was acting dumb



The interesting thing will be to find out which card the 4191A has this 220uF cap installed, and see what it measures in circuit with the de-5000. !!
I think I put a few of those in there during my retrofit!!!


What I am thinking is to plug it into oscilloscope / function generator / phase shift LCR "meter" (altoid) thing I made.......... a long time ago.... and make sure that FG output is floating properly. maybe the multifuction synth with the 600 ohm transformer could work, i contemplate this while visiting porcelain temple

[free_electron]

https://www.we-online.com/catalog/media/o467325v410%20SN019_EN_d.pdf?srsltid=AfmBOorOMkICn8IwvU5szVYcFaYdxBrHrOx7tgw7kew8jsnodXSFjnP1

[coppercone2]

Well I just plugged in 20 panasonic electro caps of various values into the DE5000 from my OLLLLDDD cap variety box (at least 40 years, possibly 50), and they all measure what I would expect

[UnijunctionTransistor]

Does your LCR tester have a “DC bias on” switch?

This is Deja Vu to me. Many, and I mean many, moons ago I had a batch of weird readings from an LCR meter when measuring electro caps; that is, until I turned the DC bias on.

Could never explain, and still can’t, the reason for this behavior, and had forgotten about it completely. Being so long ago, I no longer have the meter nor its model. It was a Leader Instruments brand if I recall correctly.

[aargee]

Occam's Razor

My 5000 does weird things like this when the battery gets a bit flakey, even with a healthy battery display.

[coppercone2]

I wonder if its a little bit of water in the electrolyte and electrocleaning from reverse polarity.

I am scared that even if you fix them, it will slowly revert. Say over 1 year. I don't see any changes yet, I think I will leave em quarantined in a junk box for a year or whenever I remember and test the 'activated' ones.

I am just gonna buy a diff part series because its looking like there are no good theories or any ways to prove anything. I cut one apart to look inside and it looks normal, pretty dry, but caps usually are pretty dry.

I bought a different type of cap to use instead, these will go to storage area

And I forgot to list the ESR for the 'unactivated' caps, its around 49 mega ohms (the meter full scale is 49.9 IIRC)


What might make sense is to measure it under incorrect bias and see when it 'activates' since if its electrochemical reaction it will be inert up to some voltage. If I try to measure the DE5000 DCV when its measuring C, its only 3.1 mili volts, which does not seem high enough to do electrochemistry, but what the hell do I know about that voodoo, I just know what happens in a beaker with coins