??? 02/21/12 15:47 Read: times |
#186110 - It's not just ESD, it's "the works" Responding to: ???'s previous message |
Kai Klaas said:
Up to now Christoph wasn't talking about such events but about ESD: It's basically about everything Per mentioned and a bit more. Not producing any output, or any meaningful output, is acceptable while the devices inputs are subjected to pulses of a couple of hundred volts, crazy RF interference and similar scenarios, as no one expects a useful measurement in these circumstances (quantity being measured are voltages in the sub-millivolt range). Examples of expected external events include: http://en.wikipedia.org/wiki/Elec..._waveforms and http://en.wikipedia.org/wiki/Defibrillator However, the device should resume operation in less than ten seconds after the interference stops, and without the user having to cycle power. "Playing dead" is not acceptable. The watchdog isn't there to cure bad electronics, it's there to cure bad external events. If electronics could be designed to handle any external events you throw at them, watchdogs wouldn't be necessary at all, would they? (Well, maybe for firmware bugs, but in a world with perfect electronics, perfect firmware shouldn't be that hard either). Of course, you could encase a satellite in a few tons of lead and water to protect it from the odd cosmic ray ... but then it wouldn't get off the ground. :) |