Ceramic cap (MLCC) Voltage ratings

[pavithra_uk]

I have lot of Multi layer SMD Ceramic capacitors salvaged from some circuit boards like motherboards, laptops, HDD, VGA cards etc.

I can measure their capacity from my DMM. But I don't know about their Voltage rating.

I have 0.1MFD to 10MFD in capacity & sizes 603 to 1210

Are there ways to "determine" rough idea about Voltage rating..
I mean size vs capacity etc.. some 1205 size has 1MFD, 4MFD, 10MFD lower value in same type -> higher voltage ??

How about voltage tolerance ? can I use 10V cap on 12V or higher supply ?

Thanks.

 

[D.A.(Tony)Stewart]

X7R has higher ratings than X5R and Y5V for the same size and value.
Voltage drops from 50V to 6.3V as you go to 10uF in small packages.
You should have recognized the voltage before removing them.. i.e. 5V or 12V bus.

Testing with 1Mohm for leakage when hot ( <85'C ) with a hot plate may reveal if the device will fail....
When they fail, they can burn out from shorted metallic layers once metallic migration starts from over voltage.. pop.

Running them in over-voltage may cause secondary failures depending on circuit.

 

[mtwieg]

The capacitance of ceramics will change depending on the applied DC bias voltage. For many dielectrics, the capacitance will decrease severely as you approach their rated voltage (we're talking at least 50% of the nominal capacitance). Beyond the rated voltage, you're looking significant risk of failure, and a greatly reduced capacitance.

If I were you I would get a test setup that allows you to measure capacitance under a variable DC bias voltage. Change the bias until the capacitance drops to some acceptable percentage of it's 0V bias capacitance, and call that its usable voltage rating.

 

[sky_123]

It's hard to say, but they can probably handle 25V, at least up to 10nF or so. Up to 100nF or so, maybe up to 20V.
However, even these are uncertain. There is no real way to determine this.
However, some cheap SMD caps are available for $0.01 (maybe in quantities of 50), so it is possible to buy (say) 20 different popular values for less than $10, and you could probably sell half of them, so for $5 you can have a set of SMD caps that will last you a very long time, and you will know their full specs.
A general rule is to to double the rating, e.g. 20V cap on a 10V supply, for a safety margin. For tant, maybe have even more of a margin.

 

[pavithra_uk]

Thanks for replies

- - - Updated - - -

The capacitance of ceramics will change depending on the applied DC bias voltage. For many dielectrics, the capacitance will decrease severely as you approach their rated voltage (we're talking at least 50% of the nominal capacitance). Beyond the rated voltage, you're looking significant risk of failure, and a greatly reduced capacitance.

If I were you I would get a test setup that allows you to measure capacitance under a variable DC bias voltage. Change the bias until the capacitance drops to some acceptable percentage of it's 0V bias capacitance, and call that its usable voltage rating.

Can I measure capacity with my DMM (its 10$ meter) while DC bias applied ?

 

[mtwieg]

Can I measure capacity with my DMM (its 10$ meter) while DC bias applied ?

It should be doable, but the DMM can not be directly connected to the capacitor under test. You'll want to connect the DMM to the capacitor under test through another DC blocking capacitor (whose capacitance should be much larger than the test capacitor). That way the DMM will still see the small signal impedance of the capacitor under test, but will not see its DC bias.

 

[D.A.(Tony)Stewart]

40% of C at rated voltage for some suppliers and materials as I recall and for Class 2 parts 50mA DC leakage at 2.5x rated voltage for 1 second test. So if you test suspect 6.3V part at 15.7V with 1K series resistor and measure voltage drop is one way to determine derating or similar method to measure leakage threshold then rate at 50% of V where leakage rises quickly using 1K ohm in series. This threshold drops as part gets hot.

 

[FvM]

The capacitance of ceramics will change depending on the applied DC bias voltage. For many dielectrics, the capacitance will decrease severely as you approach their rated voltage (we're talking at least 50% of the nominal capacitance). Beyond the rated voltage, you're looking significant risk of failure, and a greatly reduced capacitance.

Unfortunately, there's no clear rule of thumb how much the nominal capacitance (at 0 V DC) will be reduced at rated voltage. With high capacitance exemplars I see reductions ranging between about 30 % and 75 % (70 % downto 25 % of nominal capacitance remaining at rated voltage).