Is there a manufacturers tolerance for ripple?

[Ady2017]

Hi all

I need to replace a 680uF 35V SamYoung NXB cap which has got a ripple of 1820. Whilst passing my local little electronics shop I picked up what they had in stock which was a 680uF 35V Panasonic FC with a ripple of 1655.

As the ripple rating is only 9% below what the original was I am wondering whether to risk it and fit it.

The catalogues specify a capacitance tolerance of +/-20% but state nothing about the ripple. If there is a similar tolerance for ripple then I might be worrying about nothing. Does anybody know?

Thanks

[stj]

where did you buy them?
i didnt know london had any electronic parts shops left - other than cricklewood electronics.

[TheLaw]

The ripple current rating is generally used as a point of guidance in selecting a capacitor. It's not a hard limit, unlike the voltage rating, for instance.

Each switch-mode topology has a set of equations which govern the (theoretical) RMS ripple current in the input or output capacitors. It is generally good practice to select a capacitor with ripple current rating higher than the actual RMS ripple current. But, it is not strictly necessary.

You can run a capacitor at 2x its ripple current rating (for instance), and even get good reliability, as long as you can keep it cool enough. When you operate near or above the ripple current rating of a capacitor, you can expect a somewhat significant temperature rise at the center-most part of the capacitor (a.ka. "the core"). At 2x the ripple current rating, you might expect a 20-30C rise compared to the capacitor's case temperature. But, if you can keep the case temperature at 30-40C (with a fan for instance), then you might be okay!

The ripple current rating goes together with the endurance rating. The endurance rating: 2,000; 6,000; 10,000 hours (for instance) is the expected life when the capacitor is run "fully maxed out". That is: At it's max temperature rating (105C, usually), max ripple current, and max DC voltage.

In this case, if everything else was equal, 9% lower ripple current rating doesn't mean the thing will explode, just means it may have slightly shorter life. But then again, it appears SamYoung has a 4,000 hour rating and Panasonic has a 3,000 hour rating, so it might just even out.

[stj]

thats not true,
if the ripple is higher than the cap can handle then it wont smooth it out.

[TheLaw]

No. A capacitor's ability to smooth out voltage ripple is related to its capacitance and ESR. Whether a given capacitor can "smooth out voltage ripple" is totally dependent on the requirements of the application. I've designed power supplies running polymer caps at 1.5-1.8x the rated ripple current rating (with forced air cooling) and the voltage ripple is plenty acceptable.

The ripple current rating is derived from the temperature rise of the capacitor core, that is the self-heating due to I(ripple-rms)^2 * ESR. It's important to remember that the ripple current rating is defined as:

"...the RMS ripple current which, when operated at the maximum category temperature (i.e. 85 or 105C) will result in the stated capacitor life (i.e. 1,000 or 10,000 hours)."

Ripple current rating is good to gauge at a capacitor at first glance. And it is useful to understand the ripple current ratio (Ia/Ir), where Ia = actual ripple current, and Ir = ripple current rating @ max temperature. Ia/Ir helps you determine the extent of the self-heating of the capacitor, which helps you understand what temperature the case of the capacitor needs to be to meet the product lifetime requirement.

http://jianghai-america.com/uploads/...mation_AAL.pdf

[AleXis6]

Originally Posted by TheLaw

Ia/Ir helps you determine the extent of the self-heating of the capacitor, which helps you understand what temperature the case of the capacitor needs to be to meet the product lifetime requirement.

thank you!

[momaka]

Thanks for posting this, Law! Really useful information there.

In regards to the OP's question (which the above info probably answered already), I think that Panny FC will do the job just fine, if not better. It may be rated at slightly lower RC (ripple current) than the SamYoung NXB, but I think the Panny FC will actually hold true to its specs... whereas the NXB was probably a bit over-rated, just like most non-Japanese capacitors are (though probably nowhere near as bad as the cheap Chinese/Taiwanese crap brands.) So I think the FC will do fine and probably outlive the NXB by many years.