Re: Does a ][ power supply ever take out its mobo?

["Michael J. Mahon" <mjmahon@aol.com>]

jsnospam@cimmeri.com wrote:
> > Note that 85C is 185F, and I've never found a power supply case
> > too hot to hold my hand on (which would be about 140F)...
> >
> > We always like to have things run cool, but they are well within
> > their design specs.
> >
> > -michael
>
>
> True, but I presume the hotter the average temp over time, the shorter
> the average life of the electro caps on some logarithmic scale.

This is true.  In fact, reaction rates typically increase exponentially
with (absolute) temperature.

> Question: if you use an 85c cap at 86c, what happens?  Is there a mean
> failure rate at this temp?

Presumeably, if maximum ratings are exceeded, the rated life will
not be achieved.  In the case of capacitor voltage ratings, the
failure could be quite prompt, but I would only expect a shortened
life for a small overtemp condition.

> A detail I forgot to mention on this 1978 ][ power supply is that it
> was used only until Apple came out with a more RFI friendly version
> sometime after.  So it's been sitting unused for at least 20 years.
> This sit period really puts the whole "how often do the caps need a
> reforming charge" question up in the air.. as all these caps so far
> (save the one) test perfectly good though I've not tested for leakage
> rate yet.

Loss of formation will manifest itself as increased leakage, even well
below the rated voltage.  Slowly increasing the working voltage (hard to
do while installed in a switching supply), waiting until leakage drops
to increase it in steps, until working voltage +10% is reached is the
best way to re-form one.

> Electro caps are certainly statistical, unpredictable little shits,
> aren't they...

That's chemistry for you...  ;-)

-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."