> It is apparently possible for drafts to set off the PIR sensors
> used in security systems

In Marco's proposed application it seems as though to detect
events reliably you'd need to know their nature, as compared
with what else is going on. For example, a still warm body in a
still cold room will not cause a detectable output from the PIR,
because the output is the difference signal between the two bars,
and if their inputs are stable, so will be the output. It floats mid-
supply, very much like some of the Hall sensors. When the PIR's
input changes, the slope of the output change imparts some
information about what caused it. Again, much like a Hall sensor

The steeper the slope (ie higher frequency) the greater the difference
between the two bar's signals. This indicates both the speed and
temperature difference, wrt to the stable environment, of the event/
object. A high-pass filter should remove drafts, although PIR's have
a very narrow bandwidth and in a dynamic enviroment this might
not be enough. Depending on the application, digital processing
would be needed to augment the analogue
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist