22 Jul
2015
22 Jul
'15
8:33 a.m.
On Wed, 22 Jul 2015, tony duell wrote:
On Wed, 22 Jul 2015, Noel Chiappa wrote:
I am way out of my knowledge range in this
discussion, but here's
something I wanted to ask about: how do you reconcile this observation
(assertion?) with the observations from several people (e.g. the PDP-1
people) that they _have_ measured the electrolytics in their power
supplies, and despite being N decades old (where N ~= 5), they are
_still_ within specs? If the very nature of electrolytics mandates that
they degrade, how are these still meeting specs?
Well, at least 2 possibilities...
Firstly, the tolerance of the capacitance of an electrolytic capacitor
is very wide -- -20% to +80% is not uncommon. So it's quite possible
they started off at the top end of that range, have deteriorated over
the years, and are still within spec. Of course nobody can prove that
(unless there are records of the values meaured 50 years ago) and nobody
really knows how they will continue to change (if indeed they do).
Secondly, I have no idea what was measured. The capacitance value is not
the whole story by any means. In fact the most important thing most of
the time is ESR (Effective Series Resistance) which increases as the
electrolyte dries up. The ESR of these components could well still be OK
after 50 years, but again nobody knows what it was when they were new.
That said, I keep on with the comment that the important thing is 'does
the circuit behave as required', If so, then the capacitor is almost by
definition OK in that circuit. From what I remember from an earlier discussion about
that PDP-1, after
reforming, those large capacitors were leak tested at or slightly
above
their rated voltage.
As for meeting spec, unless DEC documented what their original design
criteria was, there really is no way to know with 100% certainty. The only
thing we could know today is if the capacitor passes industry standard
tests and if the power supply those capacitors are a part of functions
correctly when fully loaded.