On Thu, 07 May 2009 13:59:30 -0500, Lostgallifreyan
wrote:

Spehro Pefhany wrote in
:

Panasonic and Nichicon make about 20 different grades and "types" of
electrolytics. Does one just look for high operating temp and low ESR, or
are there other factors (or combinations thereof) to consider?

Dimensions (which affects life reduction due to self heating as well
as simply fitting the available space), lifetime (at temperature),
cost, availability, low operating temperature, maximum ripple current,
leakage current, tolerance, working voltage, surge voltage, etc.


Those are interesting, but I found that it's often easier to focus on a
subset. Dimensions are often dictated by the space, so I just go for higher
voltage caps which will be bigger anyway, that solves the self-heating and
the voltage overhead for safety.

Maybe for replacement use, where someone else has already crunched the
numbers.

Availaibility is interesting, but often the best parts get neglected till
people cry over spilt unobtainium.

I always try to use the most bog-standard, inexpensive, available and
reliable parts. There's more than enough excitement to be had in the
parts that really ought to be oddball.

Although they cost a bit more, it might be
best to go for them anyway, that way suppliers will take care of
availability. The best way to make unobtainium is the slow fission of
amnesium...

I'm not eager to take dusty stock off the hands of distributors under
most conditions.

One thing I never remember well is the correlation between ripple current and
ESR, though I imagine low ESR types tolerate ripple currents better because
they will dissipate less power per amp of ripple.

Yes, all other things being equal. One thing that's not obvious unless
you study the data is that self-heating, even when within the
allowable ripple current rating, can dramatically reduce the lifetime
of the capacitors (and often also the product).