Ripple Current ratings for electrolytic caps - what's good - more or less?

[Marco17]

When I look at filter caps on mouser, the ripple current ratings is usually listed. The cheaper priced caps seem to have a higher ripple current rating so does that mean that they are passing that amount or is it that they cap can handle up to that value? And how much of a rating is good for amps? thanks

[R.G.]

Ripple current is the AC current going through the capacitor as it "swallows" voltage fluctuations to filter the pulsed currents into and out of it.
Ripple current rating is the maximum ripple current it can accept without being damaged. So higher is better, all other things being equal.
v
v
... but as we know, all other things are never equal.
The ripple current rating is a sideways method of saying " don't run too much AC current through this thing or it will overheat and degrade its insulation and electrolyte. This can be either slow (i.e. the capacitor ages faster) or spectacular in that it spews its internal fluids out the vent/burst opening. I once had a 3inch by 8inch cap spew itself onto the ceiling.
ACK.
The capacitor part of the capacitor doesn't generate heat. The resistor part of the wires, internal foils, electrolyte solution, etc. does generate heat, and unless the outer skin of the cap can dissipate the heat, something inside gets too hot. The ripple current rating is a statement of how much AC current the cap runs through its internal resistance and how easily it can move it to the outside skin and still stay cool enough. All this would be easy if all you had to do was to measure the ripple current and the ESR and compute the heating. Unfortunately, that's not simple to do, even for a professional. I've tried, and it's a bear.

So yes, bigger ripple current rating is good. Bigger maximum temperature rating is even better, as the 105C rated caps get to there by having better means of getting rid of internal heat. They're putting life ratings on caps these days, in either years (at some unstated duty cycle/conditions) or thousands of power-on-hours for some kind of condition.

Bottom line - it is quite difficult to pick the best cap, and sometimes only possible to get "enough". For a given capacitor position, pick the best combination of physical size, highest temperature rating, and longest life rating from the capacitors that are in stock that you can actually buy. If there is more than a couple of choices after that, favor higher ripple current rating or lowest ESR.

[Marco17]

Wow, great explanation! Thank you, that is very helpful. I can't imagine seeing a cap that big expel it's innards onto the ceiling! Yikes!

I guess I'lll stick with NIchicon or Panasonic

https://www.mouser.com/ProductDetail/647-UCY2H330MHD - up to 500V, but only? 350ma of ripple current. I mostly restore vintage PA heads so the B+ is well under 450V. These are $3.56 each

https://www.mouser.com/ProductDetail/667-EEU-EE2W330S. These are 450V, but 1A of ripple current and only $2.42 each or $1.68 each for 10.


What is an acceptable amount of ripple current to be able to handle?

[R.G.]

I guess I'lll stick with NIchicon or Panasonic
[...]
What is an acceptable amount of ripple current to be able to handle?

Let me attack the problem a slightly different way.
Is your power supply a semi-standard center tapped high voltage secondary output type?
Does it use a rectifier tube or solid state diodes?
Do you know the DC current out at full power, or failing that, what output tubes are you using and what is the expected audio output power?

Instead of ferreting my way through the equations for a few hours, I'll model the amp in a circuit simulator and see what it says the RMS current in the filter cap is. This will give a quick and probably good-enough estimate of the ripple current in your application. My expectation is that the ripple current will be well under 350ma, but I like to check out my guesses.

Edit:
I did a quick simulation of a tube power supply with transformer parameters taken from the Hammond web site and an output of 490Vdc at 250ma. A 50uF first filter cap carries 340ma of RMS ripple current in this one, crude simulation. That at least represents a good guess point for other situations. Given that, the Nichicon looks marginal, and of the two I'd prefer the Panasonic. More as new results come in.

[Marco17]

Yes, very basic power supplies. Mostly 5U4 or 5Y3 rectifier tubes and two 6V6, EL84, 6L6 or EL34 power tubes. Just basic rock'n'roll amp circuits.

I will have to measure the DC current when I get home. All amps are cathode biased 2x 6V6 or EL84 power tubes for the most part.

Thanks!