I need help with PIR Sensor Camera Trap....

Thread Starter

GenX

Joined Mar 24, 2007
1
Hi,
I am a final year engg student and we i am doing athe above said project. I am using a LHI958 PIR sensor. The sensor will basically detect the variation in the temperature in it's range and then if a warm body is detected then it will send a trigger signal to a digital camera uaing the PIC microcontroller and thus a picture will be taken. This project can be used in capturing images of wildlife in remote places. The problem i am facing is that the datasheets that i found are not much helpful. And the professors of my college have not worked with this sensor before. So currenly i don't have any guidance and hence my work is lagging. Could anyone please direct me in the proper direction. I have been able able to secure the circuit for the PIR sensor and amplification stage but i am still not comfortable with the working of the PIR Sensor itself. When i was testing the ckt. there was a variation in the voltage wen i placed my hand in front of the sensor. But it is very unstable. It fluctuates often. So if any of you has any prior knowlegde about this ckt. then i'll really appreciate it. Looking forward for a correspondence. Thank you. Have a nice day.:)
 

beenthere

Joined Apr 20, 2004
15,819
What you are seeing is the basic nature of the beast. I have an outdoor flood that uses a PIR detector. It is sensitive enough to operate when one of the cats is in range. Also when the wind blows hard enough.

It may be that you will have to take the sensor output and integrate it using soomething like an R-C filter, with a comparator to trigger when the activity level has risen enough. I have no direct experience with the actual PIR sensor output, so this is a guess, based on experience.

Such devices are commercially available. One appraoch is to get one and do some investigation on that circuit. I think the only difference between that and an outside light is that the PIR sensor has a more limited field of view.
 

thingmaker3

Joined May 16, 2005
5,083
PIR does not detect warm bodies, but rather detects changes of IR levels in a matrix of rays in front of it. (Those would be rays in the geometric sense, not radiated.) Angle on your PIR is likely wide enough that wildlife will be in the edge of any picture taken. I suggest making a "blinder" out of a soup-can.

Instead of using your hand at close range, which will drastically alter all of the rays, aim the thing at your girlfriend from across the room and have her change pose. Such will give a much more valid indication of o/p levels.
 

mrmeval

Joined Jun 30, 2006
833
http://www.quasarelectronics.com/3030.htm

+1 thingmaker
The older PIR motion detectors did not detect a warm body per-se they detect the transition from one reflective surface to another. If you tear one apart you'll see an IR device similar to the one you list pointed at a mult-surface mirror. Some newer ones can use a freznel lens as the link above shows.

It may be possible to use several of those devices to get a vague sense of movement but at that point you'd want to look into machine vision. Someone did build a device that could easily discriminate a human from background and shoot an airsoft at them. Now Samsung says they have a 'product' for 'automated defense' And so it goes.
 
Top