PIR tracker prototype

This is a PIR based tracking device. It's very simple, and the circuit is pretty well out of the datasheet.

The main changes are:

* outputs of the comparators drive the LEDs and the two output signals: LEFT and RIGHT

* The PIR doesn't have a Fresnel lens

* I divide the PIR's window into 2 sections with a piece of cardboard. This eliminates the common mode signal and makes it much more sensitive - to both noise and heat sources.


Range is about 3m.

parts:
* D203B PIR sensor from Futurlec

* LM324 quad op amp

* resistors, capacitors, trimpot

* diodes

* LEDs

* 9V battery

* proto board


total cost is about $10

Most of the schematics for PIR sensors are almost the same. Just search Google for "PIR schematic" and you get stuff like this:

http://www.bitsbox.co.uk/data/circuits/pirdetect.pdf
http://www.glolab.com/pirparts/appckt.pdf

Here is the schematic for the unit shown in the video:
http://i39.tinypic.com/2e0u4a8.png

The schematic is from a simulator. So V1 represents the output from the PIR, R10 is just a dummy load so I can measure the output of the diodes. Replace the diodes with resistors and LEDs and you have the circuit from the video.

Also, R5 is a 2MEG pot.

Here is the schematic for a new one I'm working on:
http://i39.tinypic.com/2z3w3lc.jpg

Notice how it differentiates the signal first, adds a heap of gain, and provides a V/2 offset to begin with. Then it compares the signal and provides a right or left digital output.

Category:

Science & Technology

Tags:

• PIR
• track
• robot
• sensor
• electronics

[RoteKampfSocke]

I have a question : how quick do these sensors react to heat changes? could you use them as a heat sensitive element in a mechanical scanner ?

[ArtistEngineer]

@RoteKampfSocke No idea sorry.

It would be a function of the thermal mass of the element, but I don't have any figures.

For a mechanical scanner, I'd probably buy a TPA81.

[sagar29s]

hey artist engineer, iam also making motion tracker using D203B

I made the circuit in the datasheet but it didn't seem to work. the output was continuously low and went high only on touching the window.

Then i started on circuit given by you, but this time it went continuously high and low on touching window.

i was thinking if sensor was damaged. could you tell me how to test it

thanks in advance :)

[ArtistEngineer]

@sagar29s First thing to check is the window splitter, without this part it won't work.

You can always try another D203B, but I doubt it's broken.

Do you have a CRO? Try checking the signals as you go from the sensor to the final output. The output of the sensor is very low, so set your cro at maximum V/dic and look at the output pin, move your hand in front and see what happens.

What you mean by "went continuously high and low". Did it oscillate? At 50Hz? 60Hz? Could be ground issue.

[melsaidy]

Thanks ArtustEngineer for your video but i made the circuit (as in the data Sheet )but it does not work , i guess it comes from the PIR sensor , So please could i know the output signal from the PIR sensor to compare this signal with my output signal, i noticed that the signal is about 100 mv and does not varied

So please could you send me the output signal from the PIR sensor.

Thanks for advance

[ArtistEngineer]

This signal straight from the PIR sensor is very small. The PIR I am using has a 750mV DC offset and the actual signal is only about 30mV peak-to-peak. You need to amplify this greatly to make it useful.

Try setting your scope to AC coupling, Bandwidth limited and set the vertical scale to about 10mV/division.

Also, make sure that you divide the PIR window into 2 parts, this make a BIG difference.

[RoteKampfSocke]

@ArtistEngineer

Thanks or the info, but i am looking for a rather more long range solution.2 meters detection range is not that great. Too bad, no datasheet of the pir sensors i can buy in area show how quick the sensor reacts. I assume the reaction speed is rather slow, so it won't detect somebody quickly waving a hand past the sensors field of view.

[patrickikis]

I made the circuit from the data sheet as well. I made a few modifications to it, but now I've got a good feel for the PIR. There are so many passive components in the data sheet circuit. Make sure you haven't shorted anything. I made that mistake several times. As well, when apply power to the circuit, I find that the PIR needs about 10 seconds to settle. The output of the second amplifier saturates (depending on the gain), until about 10 seconds after the intitial power up. I have vid.

[ArtistEngineer]

Also, C2 in the first schematic should be about 10uF. This might be limiting your response at low frequencies. That could be the problem.

[ArtistEngineer]

Also, because of the high gains involved, your opamp will have a significant offset error. This means that it will trigger more easily on the positive comparator than the negative one. Check the average level going into the comparators, is it equal to V/2?