i have a fluke 117 DMM

i have a capacitor with these markings. its ceramic capacitor out of power supply.

CNR
1 0D471K

therefore it is a 470pF capacitor according to this

http://www.marvac.com/fun/ceramic_capacitor_codes.aspx

my DMM reads 1 nF
and every other setting like in μF shows 0

so does the capacitor work????? or is there a more accurate way to test it???

 

Your meter must be junk. Send it to me.

Actually, the lowest reportable value is 1 nF. See here: http://www.testequipmentdepot.com/fluke/pdf/117.pdf

John

 

You need a meter with a lower range.

It is not practical to test such small values of capacitance using a general multimeter that has leads that are a foot or two long, as the parasitic capacitance/inductance of the wires will goof up the readings.

If you want to test small values of capacitance reasonably accurately, you will need a measurement instrument dedicated to that function.

 

alright. thanks for your help.

if the capacitance on DMM shows the right value as the capacitor. does that mean the capacitor works?

 

no.

the amount of capacitance is only one parameter. For instance, The capacitor could be perfectly good when suffering the tiny voltage a meter uses, and short out at 2 volts. It could have increased its ability to absorb the energy going through it. It could change value with changes in temperature, time, and voltage applied.

 

no.

the amount of capacitance is only one parameter. For instance, The capacitor could be perfectly good when suffering the tiny voltage a meter uses, and short out at 2 volts. It could have increased its ability to absorb the energy going through it. It could change value with changes in temperature, time, and voltage applied.

so how would you go about testing it?

 

Bychon's essentially correct, but he didn't go into enough detail.

Frequency response is a major factor. If you're only talking about a few kHz, no big deal.

If you are talking about frequency response in the MHz range, that's a whole different ballgame.

A cap measured at low frequencies might measure in the 50pF range (if your equipment is sensitive enough to recognize that to begin with). However, when you get into the 500MHz range, yo may measure 150pF with laboratory equipment. This is due to the parasitic qualities of the cap.

Don't mean to confuse the issue. Your application is probably low frequency. If your meter is limited to 1nF response, and the cap shows as 1nF, it's probably good for a low frequency range application.

 

so how do you test the capacitor if your saying the capacitance test on a DMM isn't good enough because it isn't at full voltage or high frequency?

 

You use a meter designed for the purpose. Here are some examples, in order of increasing capability and price:

http://cgi.ebay.com/DM4070-3-1-2-Di...emQQptZLH_DefaultDomain_0?hash=item27a9f792b4

http://cgi.ebay.com/10Khz-Precision...emQQptZLH_DefaultDomain_0?hash=item3efdbb1433

http://cgi.ebay.com/AGILENT-E4980A-...emQQptZLH_DefaultDomain_0?hash=item25597a5e69

http://www.waynekerrtest.com/global/html/products/LCR/6500P.htm

http://www.home.agilent.com/agilent/product.jspx?nid=-33831.536879654.00&cc=US&lc=eng

 

.............................

If you are talking about frequency response in the MHz range, that's a whole different ballgame.

A cap measured at low frequencies might measure in the 50pF range (if your equipment is sensitive enough to recognize that to begin with). However, when you get into the 500MHz range, yo may measure 150pF with laboratory equipment. This is due to the parasitic qualities of the cap.

.......................

Sgt, are you sure about this?
By what mechanism would a capacitor increase in value with an increase of frequency?
Its true that the effects of stray circuit capacitance become much more significant as you increase your operating frequency but the actual values of fixed capacitors do not change with frequency to any marked degree.
After all the value of a capacitor is only dependent on the plate area, distance between plates and the dielectric constant of the dielectric.
The only property that could change with frequency is the dielectric constant but generally capacitors employ dielectrics that work well into the GHz range of frequencies.
The effects of any stray capacitance or inductance become more significant as you increase frequency but the actual values of stray capacitance and inductance do not change.

 

To add to the Electrician's list, here are some other pieces of equipment:

B&K sells some LCR meters. A few years ago I got the model 886 and have found it useful when needed. I got it with the optional 4-wire probe which lets it make quick Kelvin resistance measurements down to 1 mΩ. It also makes ESR measurements and can measure the complex impedance, showing the result in ohms magnitude and angle in degrees. Normally, there's no way I'd spend that much money on a piece of hobbyist equipment, but I was able to get a significant discount on it and trade it for something else.

An impedance bridge like the GR 1650 will measure the capacitance and a series or parallel resistance. You can also bias the capacitor while testing it. You can find them used on ebay (a properly working one should be on the order of $100 or so).

Alas, most of this test equipment is priced beyond the budget of the hobbyist.

 

This is bothering me. So a cap can test fine out of circuit, and fail under load? How am I supposed to diagnose that?
Ceramic caps seem easy - they usually crack, or fail shorted. I'm still trying to learn how to test tantalum caps - I guess a similar test would work, but I have yet to find a confirmed bad tantalum cap.
I too just bought a Fluke 117. How could I use this meter to test these caps?
I have an idea. Tell me if it's viable or not..
I use a lab power supply to inject say, 10V into a cap while it's in circuit, being careful to check the polarity of the circuit so I don't break anything. Then, use the DC voltage part of my meter to see if I get 10V off of each cap on that particular rail. My idea is that if one of those caps is intermittently failing and shorting under load, it will cause a short in that segment, and when I connect the fluke, I'd see zero volts. Yes? No? Stupid idea?