Re: electrolytic capacitors

["Bernard G. Cleyet & Nancy Ann Seese" <georgeann@REDSHIFT.COM>]

High their!

When I was kid (6th grade). I made a battery charger using baking soda
electrolyte, carbon rod (from a #6 Le Clanche
cell) and a sheet of aluminum.  Soon after I read that I'd made half of an
electrolytic cap.

bc

P.s.  In an early foray into industry, I made a hiV. ramp generator to test
BaTiOx caps.  The charging rate (electron flow-flow) would give the cap., but
more important(ly?) would indicate when it saturated!  (memory, ca. '62)

P.p.s In addition to the informative post below:  The construction of
electrolytic caps. easily explains why their tolerance is + large value (e.g.
200%) - smaller value (e.g. 50%), and during still another foray (about the
same time at Raytheon) "blown" storage caps. for flash pumping of Cr  and Nd
were too expensive to toss, so we'd "heal" them by charging them overnight in
series with, say, ten megs. This way one may convert a polarized to a
non-polarized cap.

John Denker wrote:

> At 10:29 PM 1/29/01 -0700, Larry Smith wrote:
> >     Electrolytic capacitors ...
>  >  How do they get so much capacitance in such a small package?
>
> At one level you already know how to get a big capacitance in a small
> package:  you want the thinnest possible gap between the capacitor
> plates.  So the question is, how might we accomplish that.
>
> The parts of an electrolytic capacitor are arranged like this:
>          terminal wire
>          aluminum foil (plate 1)
>          liquid electrolyte (which is conducting and reactive)
>          oxide (which is insulating, and very thin)
>          aluminum foil (plate 2)
>          terminal wire
>
> We want the oxide to be very thin.  On the other hand, if it is too thin it
> will break down and allow DC current to flow.  The brilliant idea is that
> the breakdown current will cause an electrochemical reaction will take
> place, laying down more oxide.  The layer gets thicker until no more DC
> current flows, whereupon you have a capacitor with exactly the right amount
> of dielectric.
>
> This comes with a few drawbacks:
>    *) You never know for sure how thick the layer will be, so you never
> know for sure the capacitance value.
>    *) The required thickness depends on the operating voltage.  And the
> chemical reaction is reversible.  Consequently, if you operate one of these
> critters for a long time at a voltage less than its rated voltage, the
> capacitance will increase.
>    *) Note that only one plate has an oxide layer.  If you reverse the
> polarity, the capacitor doesn't work at all, and can easily be destroyed.
>    *) The liquid layer isn't the world's best conductor, so there is a
> nontrivial series resistance.  This degrades the usefulness at high
> frequencies.   A very common practice is to use a high-capacitance
> electrolytic capacitor in parallel with a lower-capacitance
> non-electrolytic capacitor;  the former can carry a large current at low
> frequencies, while the latter can carry a large current at high
> frequencies.  (A high-capacitance non-electrolytic capacitor would be much
> more bulky and expensive.)
>
> BTW you can make the plates out of materials other than aluminum.  Tantalum
> works particularly nicely.
>
> >  Is this "capattery"
> > <http://www.evanscap.com/Capattery.html > the same thing?
>
> That sounds similar, but they don't give enough detail about the product to
> allow me to say much more than that.  I have no idea what is the chemical
> nature of the dielectric layer in that product.
>
> The name is a blend of "capacitor" and "battery".  There's nothing special
> about that;  if we view things as "two-terminal black boxes", any battery
> can be viewed as a capacitor, with an enormous capacitance and terrible
> high-frequency performance.  But they have very different physics
> internally;  discharging a capacitor, even an electrolytic capacitor, does
> not require a chemical reaction.