On 10/29/07, M. Adam Davis wrote: > Even nice electrolytics have lifetimes measured only in the thousands > of hours at reasonable temperatures. The typical power supply is much > warmer, and the lifetime is cut considerably. Some badly designed > products leave the entire power supply turned on even when the machine > is off, rather than have another more efficient standby power supply. One thing I discovered at Muzak, was that failure rates of equipment that was turned on and off a lot was MUCH higher than the same equipment left on all the time. Later, I put this knowlege into a power cycler to shake out weak components in repaired credit card terminals. It used an SSR to turn on for a random amount of time, as short as 1/2 cycle, and as long as 30 seconds, and pretty much everything inbetween. The repair guys hated it, because many repaired terminals didn't make it through cycling, but once we shook out the weak parts, the terminals that got delivered had very low failure rates. Normally, those were left on all the time, or turned off once a day when the business closed. The 2000 hour (or whavever the number) doubles for every 10C below rated temperature, and also has ripple current and frequency determinants. Jim Williams had a really good formula for predicting cap lifetime, but I've lost track of which book it's in. -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist