Al Kossow wrote: What data did you use to figure out that this process was worth doing? I think Will's implying that the need to reform caps that new is not necessary, because there's no data available to say that it actually does any good... Presumably someone somewhere's done some real tests on old caps of different vintages and come to the conclusion that it does indeed have some benefit. I don't think anyone can say that it just won't happen - so it all comes down to a question of how much risk there is. cheers Jules

I should add that offlist with another listmember, I pointed out that when I power up a rare item, I will disconnect as much as possible, so if there is a big failure, the effects will be localized. I am not so reckless. Basically, unplug the power supply from the circuitry... -- Will

Al Kossow <aek at bitsavers.org> wrote: Was that the PDP-1 at the Capacitor History Museum, Al? :-) I am certainly the guilty party in advocating the position that equipment expected to operate under adverse condition should be margined to available extremes during the PM cycle. At my day job margining the logic (be it relay, solid-state, SSI, or microprocessor) up and down also happens to take care of margining the capacitors up and down too. Margining isn't a procedure to preserve components, it's a procedure to make sure that equipment will work to specification. That's not to say that margining has a goal of blowing bad parts up, either. After 30 years we rewrote one margining procedure for track circuits after we had a loss-of-shunt and two trains nearly collided at rush hour in a tunnel under the Potomac. We found out that the procedure could result in adjustment to pass the margining but not be safe. So margining is to be done with thought and care towards important function, not just to blow the bad parts up! I don't say that babying the tenderest components cannot yield a working result, just that it's not my personal style nor the standards of the industries I've worked in, and being responsible for the logic that runs a railroad 24x7 and has to run even when it's -20F or 110F, it's more likely than possible that my experience doesn't apply to something in a museum environment. Tim.