On Sun, 1 Nov 2009 01:07:09 -0800 (PST), NT
wrote:

On Oct 31, 10:14*pm, Peter Parry wrote:

Your notion that resistors almost never fail is naive. In real life
they're subject to conditions that do cause failures at times, they're
not kept within all the conditions of the data sheet tests. Vibration,
voltage overloads, physical shock, mechanical stress, these thing
happen irl, and failures follow.

I haven't come across too many domestic microwave ovens operating in
hostile environments (unless of course you count children as a hostile
environment) but that is why MIL-HDBK-217F is so much more useful
than data sheets as it models such excursions from the norm. Using
the ground mobile model - the microwave in a cross country vehicle
used cross country - the reliability does, as you say, drop. The MTBF
goes down to merely 4,000,000 hours

What do you think you pay them for?

That doesnt even address the question of how they collect stats on
microwave deaths, and whether it covers all cases.

They gather information on all industrial and commercial accidents.
Until about two years ago they were also funded to collect domestic
accident information from fire, hospital and ambulance service
reports. While it is always possible for things to slip through it
remains a fact that screwdrivers have killed more people in accidents
than microwave ovens.