Paralleling caps to meet ripple current requirements?

No. Just the lowest rating they are available in.

No. But I think you will find that going up to a 25 volt rated electrolytic (for the input side) will give you a lower esr and higher ripple rating for the same capacitance. Most of that comes from the one step larger can.

The output generally has a lower ripple current because the converter includes an inductor that turns the output voltage pulses into a current triangle, upstream of the output capacitor. ^.3 volt units are fine for this low voltage. You may find polymer electrolytics rated for even lower voltage with very low esr to give a low ripple output voltage.

e.g.

Cool, I've got plenty of room for the larger cans.

I've dug a bit further into the app notes and the input caps need to have less than 20mohm total ESR and the output caps less than 12mohm. With the guidance you and Graham have given me, I feel good about selecting the caps that will work well for this app.

Thanks for all your help!

John

-- remove SPAMMENOT for e-mail responses --

Thanks for all your help with this Graham!

John

-- remove SPAMMENOT for e-mail responses --

More correctly, ripple current goes up as ESR drops. The ripple current is a function of how much heat the cap can stand, and if you decrease the ESR, the heat generation drops. What you're seeing is that as capacitance goes up, the ESR typically drops as well. The same is true when you raise the voltage of a capacitor.

So, to answer your question: As long as you're meeting the ripple requirements to keep the caps from exploding, and the ESR requirements to keep the regulation stable, it doesn't matter if the cap says "Low ESR" or not.

It's not how fast your load changes, it's how fast the pulses from the switcher are coming down the line. If your ESR is too high, not only does your cap heat up, you can end up with more ripple going down the line. On the flip side, some converters also have limits as to how *low* of ESR you can use without affecting the regulation.

To see how ESR affects ripple, think in terms of individual bursts coming down the line. For the very, very brief duration of the pulse, there's a pretty significant amount of power being transmitted - and a very good amount of current has to flow into or out of the capacitor in a very short amount of time in order to keep things smooth. The more resistance, the less current can flow in and out, and the less effectively the cap is able to accomplish its job.

steve

Is the temperature really the ambient temperature? I was under the impression that it was the temperature in the capacitor, which will be higher than ambient under high ripple, and hence, you can also increase the lifespan by not trying to hit the max rated ripple current for the cap. (Thereby keeping the cap from heating so much.)

You see 105 C caps all of the time, but how often do you actually see them running in 105 C ambient - or even close to it? =)

steve

Great explanations Steve, thanks!

John

-- remove SPAMMENOT for e-mail responses --