* Doug <dmcgarr...@optonline.net> [101226 22:56]: > Maybe it's time to buy tantalum capacitors. More expensive, > slightly smaller, and (I believe) less likely to blow up. Available > with parallel wires or in surface mount configurations. Military > equipment has been using tantalum caps for years, so they must be > reliable.
Tantalums are good, but manufacturers consider them a little too expensive for mass-produced motherboards. Tantalums are valuable when you need high capacitance in a high-frequency application; the capacitance of electrolytic diminishes rapidly with increasing frequency. This is why you often see two or more capacitors in parallel; typically a tiny, low-value ceramic (which has excellent high-frequency performance) is paralleled with a high-value electrolytic -- and the combination still is less expensive than a single tantalum. You can make almost anything explode (that is, fragment) if you apply enough voltage and current. Years ago I had several clones of the LM317 three-terminal regulator explode when the output was shorted; this despite the fact that the data sheet claims that the device withstands a short of infinite duration. I phoned National Semiconductor and it was Bob Pease who picked up the telephone. I began by saying, "I have some LM317s manufactured by one of your competitors..." But before I could say another word, Bob interrupted to ask, "Was anyone hurt when they exploded?" Bob went on to say that National short-circuit tested every LM317, and "the ones that explode don't get shipped." It is episodes such as this that have made Pease a living legend among electrical engineers. But even if you manage to blow up a tantalum, there is no electrolyte to spill. %%% If I recall correctly, the problem which I cited was caused by manufacturing changes regarding the chemistry of electrolytics. The problem eventually was solved by further manufacturing changes in the chemistry of the electrolytics, but not before a great many short-lived motherboards were manufactured and sold. I remember that Tyan in particular received much bad publicity from the matter, and that some motherboards failed within three to six months of being placed in service. (Something similar happened with alkaline cells when the "get rid of the mercury" mandate came out several years ago. It turns out that mercury reduces gassing, and mercury-free cells gassed so badly that they leaked electrolyte.) Finally, I was in error regarding the P5Q-EM; it employs solid capacitors only in the critical power supply circuitry surrounding the processor; other capacitors on the board are electrolytic. This is typical of the garden-variety motherboards which I see on display at the local electronics emporium. RLH -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101227002514.gb3...@rlharris.org