Maker Pro
Maker Pro

IR proximity sensor

varun1991

Jun 5, 2014
4
Joined
Jun 5, 2014
Messages
4
hello,

I have to design a low cost proximity sensor for an automatic sense and spray device. here is the link for a product similar to my task which is already in market

http://www.glade.com/en/products/sprays/sense-and-spray-automatic-freshener
http://tronixstuff.com/tag/glade/

1. As far, I have modulated the Ir emitter at 56khz using a pic mcu and used a tsop 4856 sensor in the reciever part. But i could acheive only a detection range of upto 4- 5 cms. the range is better(10 cms) when i use a 555 timer at the transmitter side. I dont know why.
how to acheive a range of 1 meter ( any low cost solution)??

2.I have also tried with high intensity Ir emitter ex: 700mw/sr, 180mw/sr to increase the detection range, but could not acheive it. shouzld i change any components while using these high intensity emitters??

3.I see that tsop sensor uses an active bpf and an amplifier to allow 56khz signals. And it allows only discrete signals modulated at 56 khz.

I want to create my own reciever circuit to allow only discrete on off signals modulated at 56 khz freq. how can i do that. what should i consider in building a circuit. I have attached a picture of a BPF that i tried in filter pro. but it doesnt seem to be correct when i tried to simulate it in TINA- TI.
Please let me know if this filter is suitable for 56 khz.
If not, what values should i change for my freq and why

4. In the second link that i have attached, is it a motion sensor that is used or a normal photodiode.

I am a newbie in this field. I have just started learning by doing. I dont want to give up yet, please help me out with my queries

Thanks and Regards
 

Attachments

  • 0243_bpf.jpg
    0243_bpf.jpg
    34 KB · Views: 126
Last edited:

davpankhur

Jun 5, 2014
25
Joined
Jun 5, 2014
Messages
25
I had the one with the S-shaped button, it only had 1 hole and so im sure it does not use IR. I think it uses a simple light sensor, and i remember reading in the manual that it needs to be placed carefully, like out of sun light. I think that one gets ready when it first detects light, then when your shadow casts over the little hole, it activates and does the spray action, then resets.

I plan to modify one soon, as the new one i have is motion sensored too but is also programmed to spray every 36mins no matter what - just so i continue to buy more refills!
 

varun1991

Jun 5, 2014
4
Joined
Jun 5, 2014
Messages
4
I have opened the product and had a look on the circuit. the sensor seems to look like a photodiode. Is it a special PIR sensor or any other kind of sensor (or) just a photodiode working as a motion detector?? If its just a photodiode. how does it work this way
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I'll only address a couple of issues

  1. You get more range using a 555, probably because it can supply more current to the LED
  2. I would be considering using an ultrasonic sensor rather than IR
  3. The tsop device should respond to a continuous 56kHz signal with a constant output, so I'm not sure what you're on about with the on/off question.
 

varun1991

Jun 5, 2014
4
Joined
Jun 5, 2014
Messages
4
I'll only address a couple of issues

The tsop device should respond to a continuous 56kHz signal with a constant output, so I'm not sure what you're on about with the on/off question.

I want to build a circuit with normal phototransistor, bpf for 56khz and amplifier instead of using a tsop sensor
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Have a look at this brief teardown of a Glade sense-and-spray unit:
http://www.jcopro.net/2011/02/09/glade-sense-and-spray-teardown/

He says "I don’t think the “proximity sensor” on the Sense and Spray works as one would expect. After reading the back of the box, it seems to work on sensing shadows in a room and not by actual proximity (how this is supposed to work in the dark, I don’t know)." This agrees with davpankhur's comment in post #2 and your comment in post #3.

If that's true (and I suspect it is), the sense-and-spray doesn't actually transmit anything. It detects changes in the light level falling on a photodetector and uses that to guess that something is moving around in the room.

I would guess that it would ignore brightness changes that are very sudden (e.g. a room light being turned ON or OFF), and changes that are very gradual (e.g. variations in daylight).

This control philosophy is not very foolproof and I suspect the reason the false positives aren't a "problem" is the enforced delay between sprays, which makes the detection scheme hard to test in a timely way.

You seem to want to detect actual movement, rather than changes in light level. That's an admirable aim, but it makes the circuitry a lot more complicated. You could transmit an infra-red beam and detect its reflections. In this case you need to use a modulated beam because you need your receiver to be able to distinguish between reflections of your transmitted beam, and variations in the ambient infra-red light level which could be caused by other factors. Most kinds of visible light, including sunlight, contain at least some infra-red light, but it's either steady, or modulated at twice the mains frequency, never at 56 kHz.

For that scheme you would need a transmitter (a 555 is a reasonable option, though the frequency is not very accurate) and a receiver followed by a bandpass filter to extract the modulated light signal. You would then have to track the amplitude of that signal, and detect changes. The simplest algorithm would be to simply detect any significant change as the movement of a person. You could get smarter and regard a sudden change as a probable indication that the unit has been moved, and ignore changes for a while until the signal settles down again. There would be other smart logic you could probably add. These are just my own random ideas.

This scheme using modulated infra-red light has the advantage that it detects actual changes within the environment of the detector, such as people moving around, things in the room moving around, and the detector moving around, and it would be fairly immune to changes in ambient light, such as room lights being switched ON and OFF and changes in natural light. Changes in natural light could be quite rapid - for example, a low cloud moving in front of the sun - and this could be enough to falsely trigger the simple shadow detector, which again fits with davpankhur's comment in post #2 about the need to place it out of direct sunlight.

Ultrasonic (using the Doppler effect) is a common way to detect movement. It is totally immune to lighting changes. It would be falsely triggered by any non-human movement in the monitored area, and movement of the detector itself.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
If Kris is right then a PIR sensor is another option.
 

gorgon

Jun 6, 2011
603
Joined
Jun 6, 2011
Messages
603
All active transmitter/sensor systems will require a higher power consumption than a PIR solution. If this is based on a battery powered unit, I would think that the best is to make it passive, and some form of IR sensor would come to mind. A PIR sensor is based on a 'picture' of points, and the change of temperature relative to the background. The least effective temperature is around 25C, where the outer body temp is equal to the ambient, and the difference is more or less not exsisting.

A range of 1m is hard to achieve for a capacitve sensor system.
 

varun1991

Jun 5, 2014
4
Joined
Jun 5, 2014
Messages
4
I have opened this device and It seems to look like a photodiode. As far as i understood, the photodiode is calibrated according to surrounding light using opamp and when there is no light falling on the diode, voltage diff occurs in opamp and the output is triggered. If it works on this principle, does photodiode work effectively in detecting fast moving obstacles . If not could anyone explain in detail how does passive sensing work effectively with normal photodiode.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
It simply detects changing shadows. As a person moves near the photodiode, the amount of light falling on the photodiode changes. Typically, it will decrease then increase. The change may not be great. But a change is what your circuit needs to detect. It doesn't need to be "calibrated according to surrounding light" because it only needs to detect changes.

The circuit needs to ignore gradual changes that might happen due to variations in daylight. It should also ignore small variations. Other ways to make it more resistant to false triggering, I imagine, would be (a) ignore very fast changes because these might be caused by room lights being turned ON and OFF, and (b) expect a drop in lighting followed by an increase, within a few seconds; in other words, ignore an increase in light level by itself, and ignore a decrease unless there's an increase within a certain amount of time.
 
Top