Bob Axtell <engineer <at> cotse.net> writes:
> Just incidentally: sources tell me that these PIR sensors 
> have been improved 10,000 fold, and some can pick up long 
> range IR from the "Big Bang". 

I don't think so. PIR chips have always worked very well against an 'ice'
cold background (and the +20K ?) temperature of the black sky is certainly
cold enough for that. But they do not really detect the sky's IR, instead
they 'see' the contrast between objects in their bandwidth and the absolute
black of the sky.

The problem with most PIRs made for commercial applications is noise. The
FET preamp and the crystal itself create enough noise that any use outside
the designated purpose is nearly impossible. The S/N ratio for the
designated area of application (detecting nearby passing people through a
cheap plastic segmented Fresnel lens, at a hot/cold target ratio of about
1-5 degrees K - the difference between a clothed person's surface
temperature vs. room ambient) is something like 10dB-ish. The crystal
itself drifts like crazy and shows bumps and peaks in the output that have
nothing to do with IR (maybe they detect some radiation particle or single
photon impact events). This is easy to test. Put a PIR in a thermostat
(empty thermos bottle Dewar works) and connect the output to a DVM and
analog scope (the preamp analog output). You will be surprised at what
comes out of there. The FET that has a lot of 1/sqrt(Hz) noise at LF to dc
is also not very helpful in this situation. Chopping is limited to <10Hz
before the gain decreases sharply (as specified by the datasheet, likely
due to the crystaland die capacitance), a frequency that does not pull the
signal out from the horrible low frequency noise domain of the built-in FET
transistor.

I have made experiments with commercial PIR sensor chopping and other
similar devices and the results were not good. I believe that using a raw
(sans FET) sensor and a better quality preamplifier may yield better
results, if the PIR crystal is of a low noise type (I don't think that the
types used in commercial PIRs are low noise at all - they seem to make
noises reminiscent of damaged carbon microphones in an otherwise quiet
environment).

Peter P.


-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist