Shop OBEX P1 Docs P2 Docs Learn Events
Options for IR light curtain — Parallax Forums

Options for IR light curtain

T ChapT Chap Posts: 4,223
edited 2012-12-10 13:20 in General Discussion
Hey guys, question. I have a project I am working on and part of the requirement is to have a light curtain, which is bank of IR LEDs every 6 inches that will detect a break in the curtain. The curtain covers 6' vertically. There are easy options online for the strip of emitters and detectors, and built in processor with relay. They are all pretty crude looking, typical elevator stuff that is not designed to sit out visible in an exposed way. So I am thinking about options to do something cosmetically much nicer. A quick example: just simulate the strips with a nice piece of hardwood, drop the LEDs on one side, drill through and have only about a 1/16" every 4" or so. Same for the detectors, so the result is not some metal gadget with an ugly plastic lens(full length on most products, top to bottom, 7' long). The hole pattern still is a bit obnoxious, this is a high end home, an entrance to bedroom.

I wanted to see if anyone had any other suggestions. An obvious 'invisible' choice might be the Xband. I have those, and they can be mounted behind wood or drywall easily, so no cosmetic detail. But, since there is a door that is closing in the path, the Xband would likely trigger off of it. The IR beam is logical, as it is a very narrow line of sight that must be broken, it ignores the door.

Any thoughts or suggestions would be really appreciated.

Comments

  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2012-04-05 10:14
    Since you'll probably be using modulated sensors anyway, could you use emitter/detectors on one side, or even from the top of the jamb only, looking for a change. The emitters would have to be driven pretty hard for this to work. The disadvantage to this method is that not all objects are reflective to IR, so if the aim is to provide security (as opposed to casual entry/exit monitoring) I guess it's not the best choice.

    Otherwise, I guess clear emitters and detectors would be less intrusive looking on a white background.

    Does it really need to be every 6"? That seems a bit extreme, even for commercial use.

    Personally, I'd just hang a set of 60s hippie beads connected to some piezo sensors to detect movement. Add a black light or two, some incense, and sitar music, and you'd be all set!

    -- Gordon
  • idbruceidbruce Posts: 6,197
    edited 2012-04-05 10:22
    T Chap

    What is the ultimate goal?

    As gordon questioned, security, occupancy,...
  • ElectricAyeElectricAye Posts: 4,561
    edited 2012-04-05 10:28
    Is it possible to use a plastic material that is opaque to visible light but is transparent to IR? In that case, your system could be virtually unseen.
  • T ChapT Chap Posts: 4,223
    edited 2012-04-05 10:29
    The goal is that if the set of motorized pocket doors are closing (dual doors), and a pet or child steps between the doors, it will retract. The standard elevator strips are 6' tall, spaced every 6" typically. The commercial units are all one side is emitter, the other strip is detector. I have not seen one strip for emit/detect using reflection.

    Any plastic lense(strip) will be offensive to the cosmetic detail of hardwood and drywall. But as a last resort...
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-04-05 10:49
    The dark, smoked plastic used on TV remotes is nearly opaque to visible light, yet passes IR readily. A narrow pinstripe of the stuff embedded bottom-to-top in the door frame to hide the IR components, set off against a woodgrain surround, could easily be made to look attractive.

    -Phil
  • T ChapT Chap Posts: 4,223
    edited 2012-04-05 11:00
    Agreed Phil. I am also thinking, recess the lens strip, whether I build the array or use store bought.
    446 x 409 - 30K
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-04-05 11:05
    I'd be afraid that a narrow recess will attract dust, cobwebs, etc., be impossible to keep clean and, ultimately, will get blocked. Better, IMO, to keep everything flush with the surface.

    -Phil
  • T ChapT Chap Posts: 4,223
    edited 2012-04-05 11:10
    Not to mention the need to replace any bad elements.
  • idbruceidbruce Posts: 6,197
    edited 2012-04-05 11:13
    Is there any casing going around the perimeter of the jamb?
  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2012-04-05 11:21
    If that's the case how about placing the strips recessed in the doors themselves, rather than the jamb perimeter? As it's a breakbeam system it shouldn't matter if the light intensity changes, and regardless you'd be using modulated beams anyway.

    Don't know what state this is for, but by law there might need to be a secondary fail safe in the motor door closure as well. You can see how the Star Trek fans have tackled this. Some of the designs are pretty ingenious (though some look institutional). I'm sure you thought of it, but fire laws prohibit automatic doors in residences that cannot be also manually operated. Occupants must be able to freely and easily open the door by feel.

    -- Gordon
  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2012-04-05 11:50
    Also, I assume you want to put light strips on either side of the door, for people/animals coming through from either direction? The drawing shows it on one side only, but the scale is hard to gauge, to know how close the sensors will be to the doors. I can imagine if they are too close the light could reflect off the doors, effectively missing the fact that someone is in the direct path. If the sensors are too far from the door, then there's an added risk of the doors closing on someone walking through the opposite way.

    -- Gordon
  • T ChapT Chap Posts: 4,223
    edited 2012-04-07 18:54
    Hey guys, here is a video of the IR Safety Light Curtain mock up. Thanks for the suggestions on this and the transistors to drive it. This will be a hardwood strip that gets mounted on each side of a door jamb(actual will be part of the complete jamb, at 72" apart). The LEDs will shoot through .20" holes, and receive on a TSOP 31238 38k detector on the other side. They will be mounted every 4" starting at the floor for 4' up. I only wired up one LED to test. I am using the Sony IR remote Send Obj for the starting point. The code will ultimately turn on one LED at at time, check the TSOP, and move to the next LED in some type of pattern. If one of the detectors get's stuck, the system will eventually cancel it out of the scan loop until reboot. This way, if one LED or detector ever dies, the whole system does not go down.


    This was a lot of fun to figure out, amazing how tiny the beam is. This is a very shaky iphone vid and the narration sucks.

    http://www.youtube.com/watch?v=2l7-UM3D39I
  • jmgjmg Posts: 15,157
    edited 2012-04-07 20:03
    Interesting project,
    To minimize wiring, you could chain these with 1G79 / 2G79, and if you resistor-couple the IR Rx, you can wait in one state with all on, then if you sense a change, send a scan wave via the FF, and the CLK(s) where no Load decrease is seen, is the one that is off.
  • T ChapT Chap Posts: 4,223
    edited 2012-04-14 15:17
    Progress on the system. 12 emitters, detectors. LEDs can be placed in any pattern. The basic concept is every 4".
    640 x 478 - 93K
  • evo803evo803 Posts: 1
    edited 2012-12-10 13:20
    What are you using for led's and do they have to be pulsed ? I am assuming they have to be based on the receiver but what are the advantages of using pulsed style receivers.

    Thanks
Sign In or Register to comment.