[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Electrolytic caps and Apple analog circuits (semi on topic)

jsnospam@cimmeri.com wrote:

Forming can be maintained by occasionally applying
working voltage to a capacitor.



So for the on-topic generalization for Apple computers, how often at a
minimum do you think folks should be powering them up, and for how
long?

Kinda would have to know the rate of oxide breakdown to hazard that
guess, I guess!


Exactly.  Which depends on capacitor manufacturer and particular
production run!

But for all practical purposes, if a device is powered up every
3-5 ("several") years for "several" minutes, the capacitors should
not fail from lack of forming.


If an electrolytic is properly sealed, evaporation is not a
problem--of couse, not all *are* properly sealed...



I'm reading in a few places that the electrolytes are now more paste
like or even solid, and are designed to absorb moisture from the air
to stay functional longer... so if that claim is true, the canisters
must be breathing.  Does that seem plausible?  Doesn't to me,
otherwise, why would the top of the can be scored?  You've already got
some other vent which would rupture.


That's correct.  True liquid electrolytes haven't been used for
many years (1940s?), and all electrolytes are now either very
viscous or even solid.  But all the same chemistry applies.

Absorbing moisture from the air would be a serious problem, since
then they could absorb other things that would poison the electrolyte.

The scoring is an overpressure relief system, designed to make
an explosion of a capacitor more predictable and at a less dangerous
pressure.


The easiest test for leakage is to lift one end of the cap and
see if the DC voltage on the other end shows up on the free
end.  Digital VOMs will usually have megohm-range
input impedances on DC ranges, so it only takes several
microamps of leakage current to read as a significant
voltage. 


Hm.  Are you sure, Michael?  I know that works for non-electrolytics,
but I thought that all electrolytics leak at least some -- eg, I
quote: "A forming current of 100 microamps to 1 milliamp is a typical
maximum allowable current."   So you would always see the DC volts on
the other side.


I generally find that "good" electrolytics leak far less than that.

If a <5000uF cap leaks 1mA (after forming), then I regard it as failed.


The other effective "test" is replacing the capacitor and
confirming that the symptom of leakage is gone.  ;-)



With the Mac 512k running, I put a scope on that horizontal sweep
blocking cap, and at least noted that DC was on one side and not the
other.   But then I put my finger on the cap and noticed it was very
hot.    Displeased, what I ended up doing was replacing it with 4 1uf
monolithic caps in parallel.   Much tinier, no electrolytes, and
totally cool to touch now.    Makes me wonder why they put that elec.
cap in there to begin with.  End result?  No change in video. lol.
So, yoke slip rings, it became!


A cap that gets hot should be replaced.  Excessive leakage and/or
too high an ESR is causing too much (DC or AC, respectively) power
to be dissipated in the capacitor.

Horizontal displacement of the raster is, though, a sign of leakage
current of several mA.

-michael

NadaNet file server for Apple II computers!
Home page:  http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it's seriously underused."