Weird capacitor readings

[Nick's Tvs]

I just bought a lot of capacitors all ranging from 680uf 10 volt all the way up to 2200uf 50volt.
Turned out to be about 100 capacitors or so in total. I decided to test a few of each value to make sure all were good.
I found out something strange, the capacitance on most of them was accurate within the 5% range. However it seemed like the higher the voltage rating on the capacitor, the higher above the farad value I would get. Once I got to 35 volt ratings, it would be above the 5% accuracy. For example the 1000uf caps would have a rating of 1100uf for the 35 volt caps, and the 1000uf caps for 50 volt would be all the way up to 1150uf some at 1200uf. Is that my multimeter screwing up or is this actually normal?
I bought a $100 digital meter specifically for the farad meter on it that is supposed to go all the way up to 5000uf. Did I get screwed over on the meter or on the caps I bought.
If I do install some of those caps that are well above the 10% margin of error, will it be the same as if I installed 1200uf caps?

Just trying to get more specific knowledge about components.
Thanks
Thanks

[mariushm]

Re: Weird capacitor readings

Read the specifications of your multimeter.
It varies from multimeter to multimeter, for example it could be that in some ranges the multimeter's accuracy could be anything between +/- 2% to +/- 5%.

This combined with capacitor's initial +/- 20% specification... basically, nobody needs/expects exact capacitance values when you go over let's say 100uF or so, every circuit designer is aware of capacitance variations.

[stj]

Re: Weird capacitor readings

wrong meter for the job anyway, you need capacitance and ESR.

one of these or better.
http://www.ebay.com/itm/181573126240

[R_J]

Re: Weird capacitor readings

This from another website:
All capacitors have a tolerance rating that can range from -20% to as high as +80% for aluminium electrolytic’s affecting its actual or real value. The choice of capacitance is determined by the circuit configuration but the value read on the side of a capacitor may not necessarily be its actual value.

So don't take it as gospel, but I know +-20% is close

[momaka]

Re: Weird capacitor readings

This from another website:
All capacitors have a tolerance rating that can range from -20% to as high as +80% for aluminium electrolytic's affecting its actual or real value.

I don't know about the +80%. Anything much above +20% + error of meter could mean the capacitor has high internal leakage.

Nichicon HM and HN from the affected period (date codes roughly between 2003 [H03xx] and 2005 [H05xx]) often fail like that. They will trick even an ESR meter, making you think the cap has low ESR when in reality it is high internal leakage.

[Wester547]

Re: Weird capacitor readings

^ Nichicon HMs have failed like that since 2001, the year they went into production (and as far as I know, HNs since 2002, the year they went into production). The fact that they read so high in capacitance actually means that the dielectric is very likely consumed by the aggressiveness of the large amount of H2O in the electrolytic solvent. It seems like either Nichicon didn't use very good inhibitors for their ultra low ESR lytics back then or they used a water-based solvent that had poor long term characteristics. A thinner dielectric (lower maximum voltage) means higher capacitance and lower ESR so while the internal leakage is indeed present, I'm not sure if they're actually tricking the meter in that case, they're just partially shorted internally. Capacitance depends on the thickness of the dielectric, the surface area of the plates (how etched the aluminum foil is), and the electrolyte. ESR depends on the length of the leads, the thickness of the dielectric, and the conductivity of the electrolyte.

The dielectric on the cathode (the cathode foil, what makes the connection to the electrolyte) is what controls the reverse (bias) voltage the capacitor can handle (why capacitors bloat when reverse polarity comes into play - the reverse bias voltage exceeds what reverse voltage they can handle - around 1.5V or more) and the dielectric on the anode foil is directly proportionate to the positive voltage (why they bloat when you apply too high a voltage to their plates - it destroys the dielectric on the anode). Also, the aluminum material is amorphous, that's why poor quality aluminum foil is an issue for crap brands.

Electrolytics fail in four manners:

Partially shorted (leaky)
Partially open (leaky)
Open circuit
Short circuit

And 80% is definitely wrong. +-30% would be the absolute maximum IMO (for capacitance).

[mariushm]

Re: Weird capacitor readings

@momaka : Electrolytic capacitors are +/-20%

Ceramic capacitors can vary much more, depending on several factors: temperature, voltage, chemical composition and how they're manufactured... see http://en.wikipedia.org/wiki/Ceramic...mic_capacitors and https://www.maximintegrated.com/en/a...ex.mvp/id/5527

The most often used (x5r or x7r) are usually +/- 15% or thereabouts, if properly used.

[R_J]

Re: Weird capacitor readings

I just copied a sentence from here: http://www.electronics-tutorials.ws/...tor/cap_3.html

[budm]

Re: Weird capacitor readings

"The tolerance value is the extent to which the actual capacitance is allowed to vary from its nominal value and can range anywhere from -20% to +80%."
Too bad it does not give more details, those tolerance are for CERAMIC type caps. I myself never run into lytis caps from manufacturers that have -20% /+80% tolerance and will never spec that into the designs.

[delaware74b]

Re: Weird capacitor readings

I deal with capacitors which are required to be within + or - 3% of their rated uF. If not, the HID light fixtures will destroy the lamps. For example, 24uF 480-volt cap gets replaced at 23uF. We don't worry if it's slightly over 24uF as it will decrease over time.

[momaka]

Re: Weird capacitor readings

@momaka : Electrolytic capacitors are +/-20%

Correct... unless otherwise stated on the capacitor .

I have some 50 year old axial electrolytic caps from an HP lab supply on my bench, and some of them are specced for +/- 10%. They are huge compared to modern electrolytics, though. But they still hold a charge and don't seem to have gone high ESR after all these years sitting unused. Just a bit more leaky internally, though.

[Nick's Tvs]

Re: Weird capacitor readings

Thank you for all of the answers, this has been a good amount of info.