[Noisebridge-discuss] PIR sensor docs

[Christoph Maier]

On Sun, 2009-04-19 at 11:00 -0700, Jonathan Foote wrote:
> I fooled with the passive IR sensor board a little. Managed to get a
> little more range out of it by putting a 500 ohm trimpot in parallel
> with R107 to lower the comparator threshold. However this only
> increases the range to about 18 inches or so before it starts to get
> noisy.
> 
> The first gain stage is some sort of differentiating amplifier, but
> not of a topology I'm familiar with, and I'm not sure how to hack it.
> Increasing noninverting feedback R should increase the gain, but it's
> already >1M, and more gain will at some point result in more false
> alarms. Any analog heads care to comment?
> https://www.noisebridge.net/wiki/Image:PIR_skeem_Sensor.jpg
> 
> But here are some lenses designed exactly for the two-element sensor.
> At 35 cents each they won't break the bank.
> http://www.futurlec.com/PIR_Sensors.shtml

It's a little hard to play with actual hardware from a remote location,
so I can't do it myself ... 
Do you have a sufficiently accurate instrument (oscilloscope and/or
multimeter) you could hook up directly to the sensor? 
The piece of the puzzle missing in the schematic (and the SPICE derived
from it) is the output characteristic of the sensor, i.e., how fast does
it respond, what's the amplitude, and what's its output resistance?
If that is known, you can systematically design the response you want;
in particular, you can trade off response time against noise immunity. 

Christoph <- should enjoy the weather in San Diego instead of sitting in
front of a computer