# Looking to see if the ripple current specification for capacitor is being met

I am choosing input capacitors for a buck regulator. According to the regulator and based on my design, the maximum ripple current will be 4.7A, which is based on the input and output voltage as well as the output current (taken directly from the datasheet.)

I chose an AVX capacitor, part number M3253508E2Z226JRTB.

I went to their simulation tool and looked at the ripple current graph as shown below:

I know the board PCB temperature will be 70 degC and the capacitor's maximum temperature is 85 degC, so I can tolerate a rise of 15 degC, for which the chart shows a maximum ripple current at 5.5A so this capacitor will meet my specification since 5.5A rating is > 4.7A maximum ripple.

Is that a correct interpretation of this specification?

• The temperature rating doesn't mean at 86C the cap blows up. What it means is, the capacitor will work for the rated lifetime if operated at that temperature. If it gets warmer, lifetime goes down. Note most EL caps have a 2000 hour lifetimes (yes --- About 1/2 a year). I don't think you have enough headroom to be honest... I'd seek out a higher temp rated capacitor if I were you. They are readily available up to 125C rated if you look for them. Commented Oct 14, 2022 at 19:13
• @Kyle B . What if I add two in parallel? Commented Oct 15, 2022 at 16:45
• Two smaller caps in parallel is better than one big one. The thing that warms up EL caps is called "Internal resistance". It's gonna be some fraction of an ohm usually. It's literally a resistor, and will drop power from the AC waveform into heat. The watts is straight out of ohms law P=I^2*R By using 2 caps in parallel you'll cut the effective "R" in half so the power dropped in half also. All that would help keep the temperature rise down. Still... 15 degrees isn't alotta headroom no matter how you slice it. Commented Oct 15, 2022 at 22:08

You appear to have done the sums correctly. However ...

It's a rating.

Running things just below their maximum specification, though apparently OK (it's within the spec, isn't it?), often causes a reliability problem.

What lifetime do AVX rate at 85 C maximum temperature?