From: dkevefnzs <dkevefnzs@yahoo.com>
Subject: [Fairlight-CMI] Re: Storage concerns for a Fairlight IIx & III - Re: Capacitors
To: Fairlight-CMI@yahoogroups.com
Date: Monday, September 20, 2010, 7:20 PM
Hi Greg & Steven,
the CMIs have many electrolytic capacitors. The question is which are just annoying when failing, and which might be dangerous, and what could be done for protection.
Details, as far as I know (I may be all mistaken - corrections highly welcome!):
Electrolytic caps consist of a cup with the fluid electrolyte, and the two electrodes - flat long thin metal sheets wound up in a spiral with very little distance between them. There is an insulation layer on one of the electrodes that's continuously re-built electrochemically while the cap is used.
If not used for a long time, the insulation layer slowly dissolves. (And using the cap in the wrong polarity, the insulation layer is actively destroyed.) When powered on again, with failing insulation, a short circuit may occur, destroying the cap, and possibly affecting the surrounding electronics. While the cap gets hot, its fluid may extend, and/or boil, and spill (through a security valve).
Alternatively, electrolytic caps can just dry out over time, and without the fluid electrolyte, they will lose their function as close-by charge-storage, -buffer or low-pass-filter.
More recently produced caps offer higher ratings in the same volume. But I don't know whether this translates to a robustness advantage for historic gear - it might have larger inter-electrode distances, but just as well, the modern parts might feature - instead of more filigrane elements - have a rougher=larger electrode surface, better electrolyte etc...?
--
How many electrolytic caps are there in the CMI II/III?
Hundreds.
Some very large, stressed ones in the power supplies: high capacities, high currents, high voltages, and high environment temperature.
Then, as computer ICs require relatively high currents on relatively short notice, many located closely to the ICs to provide the required charge without delay. And some more as filters in the audio path. All these little blue shining cylinders!
In modern PCs, failing motherboard CPU power supply caps are common; partially due to cheap/underrated hardware. The small circuitry buffering caps in the CMI are probably less critical (dried ones may impair reliability/function, but should not be deadly); I'd mostly be concerned with the power supply related ones (including those on the audio backplane/boards). A short circuit there *might* really affect surrounding parts, drying might result in more than just "ripple" on the PSU output.
I read somewhere that after years of storage, it would be good (old fashioned) practice, to connect an electrolytic cap to a low voltage first for hours or even days, before using it at its nominal voltage/load - in order to restore the insulation layer before applying real-life charges to the electrodes. I remember vaguely even having seen detailed suggestions as to cap ratings, recommended voltages, and durations in some source dealing with just the parts.
Sadly, that may be impractical for devices like the CMI, where e.g. the 68020 manual says: "the clock must never stop!" - (I forgot whether that would just impair operation or kill the IC?).
A logical compromise might be: After years of storage w/o use:
(a) Get more info; and (b) use a regulation transformer to first condition at least the CMI power supply caps by providing mains alimentation at a strongly reduced level - with all computer and audio boards disconnected, but possibly small dummy loads attached instead. I think they are all linear PSUs, so that should be feasible (?!).
But I don't know the true impact of ignoring such a procedure in real life (although after writing this posting, I remember vaguely having read some precaution recommendation related to antique computer gear due to the capacitor problem as well). And I certainly don't give a guarantee for the benefits of the compromise approach :-)
So the best thing may be to keep the gear in use?!
Kind regards, Joerg
--- In Fairlight-CMI@yahoogroups.com, "stefano_fonzarelli" <einstein@...> wrote:
> [...] I think that the Fairlightboards do not contain (m)any capacitors but mainly use integrated circuits. [...]