PDA

View Full Version : Door entry detection(with direction) URGENT



hitman47
04-11-2009, 01:04 PM
I want to check whether a person goes outside or inside a room and count them. So I need to use some kind of laserbeam/IRbeam/Lightbeam sensor and place them across the door frame. Please tell me how can I do it?

My idea:
place a bright Led on one side of doorframe and place two LDR(CdS photocells) on the other side of the frame. Bcoz i would be using 2 Ldr, the ambient light would not have any effect. Only difference between them would be taken as output. When first LDR is triggered(i mean light is cut off) before second, direction is in OR when second LDR is triggered before first the direction would be out.

I could also use two photodiodes/phototransistors/IRdetectors and IR led, but they get saturated easly if direct light or sunlight falls on them? However we can cover the LDRs with dark palstic.

Plz help

See image:
http://img15.imageshack.us/img15/4019/60432381.th.jpg (http://img15.imageshack.us/my.php?image=60432381.jpg)

macx75
04-13-2009, 11:49 AM
Hello U can use a pir sensor,
ldr and photo sensors can be used if u want a cheap version which will not be reliable ,
a small pir sensor is been used which will tringer and then can be controlled by a microcontroller . so when a human enters it is know.

emb_sys_dev
05-04-2009, 11:40 PM
PIR sensors are quite susceptible to environment,sunlight,air draft,ambient heat sources etc etc.So you have to very careful during installation,proper selction of fresnel lens etc.
Also it will give two counts one when person approaches and another when he leaves.
Why not look for those automatic door openers used in malls.airports etc and open one and see how they do it ??

allbits
05-06-2009, 10:28 AM
you cannot directly connect an LDR and forget about ambient light. YOu need a difference amp at some stage to cut that off. So you may need a comparator and some components to get a clean ttl output. Its possible to make a good system with LDRs.

If you want to make it simple, and cheap, buy two lasers and photo transistors, and bias the transistors properly. That will work, better than the LDRs. AND, place the LDRs/IR sensors in proper tubing(IMP !!!), which will improve the responce a lot.

suhasm
05-07-2009, 01:25 PM
I have tried what allbits says. Its not as easy as you say.

First of all , aligning the laser and the LDR is very tough. Second , You dont get laser diodes anywhere as far as i know. So , you will have to mod a cheap laser to work from an external power source.

I tried this some time back. All these pointer lasers (even the costly ones) are all crap. They burn out after just 15 mins. I used to bias them with 4.5V with a series resistor.

The simplest solution would be a Ir transmitter and detector coupled together.
The Ir would reflect off the person when he comes in front of the emitter and would fall on the sensor.

I have tried that. It works pretty well.

If you are adamant about using LDRs , then it will be better if you connect it to a uC and write code to compensate for the ambient light.

allbits
05-07-2009, 08:25 PM
Yes, aligning is tough. tricky. But it has solutions - like fixing the tubes permanent, and using small screws to fine tune - like you adjust the height of a DLP projector. Once you fix it permanent - it can work for ever.

Cheap lasers dont wear out n 15 minutes, if you give the right current !!! All you need is a good regulated supply. I have one running for more than a year at the local science center gallery.


The simplest solution would be a Ir transmitter and detector coupled together.
The Ir would reflect off the person when he comes in front of the emitter and would fall on the sensor.

yes, you can use the popular obstacle sensor - a TSOP and IR LED pair, but be careful with the distance between the sensor pairs. And, sometimes, even the colour of the shirt can change the sensitivity - So if the user will be walking at a longer distance from the sensors, there are chances that this will fail. You can see the difference with a black paper and white paper. This is because of the reflectivity of different surfaces, and the more the distance, the more the chances for failure.

suhasm
05-08-2009, 08:20 AM
Cheap lasers dont wear out n 15 minutes, if you give the right current !!! All you need is a good regulated supply. I have one running for more than a year at the local science center gallery.

@ allbits , what voltage and what current did you use?
I remember giving it 3*AA cells along with a resistor.
The resistor was a little bit on the higher side , so the current was kind of limited. So , the Laser did not glow as bright as it normally does - but it was good enough for my purpose.

But , it gradually grew dimmer and finally stopped working. The funny thing was that if i waited for a few hours and powered it up again , it used to glow very dimly for a few mins and then used to stop again.

I later tried it with a LM317 adjusting the voltage to 4.5V , only with the same results at the end.

I used the cheap lasers available in the market for 40rs like this one :
http://www.scitoyscatalog.com/Merchant2/graphics/00000001/small_5_head_laser_pointer.jpg

I removed the batteries and soldered one wire to the spring inside the laser body(GND) and another to the aluminium casing (Vcc).

Perhaps i damaged it while soldering?
I tried this with not only 1 , but over 5 lasers, and the results were the same for all.
I tried it with another costlier 200rs laser , but that never lit up at all , so i dropped the project.

Which laser did you use?
What current at what voltage did you give it?
And how did you connect the Laser to an external power source?