Light curtain

Discussion in 'The Projects Forum' started by DenisC, Oct 15, 2010.

  1. DenisC

    Thread Starter New Member

    Oct 15, 2010
    I need to make a light curtain for a high speed photo project I am working. I looked at commercial products and what I found is either too expensive and / or too slow (10ms response time). Here are some of the specs:

    • Response time < 1 ms (from any beam interruption to output trigger)
    • Min object size < 1 cm (metric dimensions here :) - this relates to how close the sensors will need to be to each other)
    • Curtain size: 50 cm (which mean I'll need 50+ beams)
    • Curtain width: 10 m (distance between emitter and receivers).
    • Power: battery powered (needs to be field transportable and usable)
    • Ouput: TTL and solid state relay to trigger a custom made, processor based delay box.
    • Other considerations:
      • Needs to operate outdoors - i.e. the light emitters will probably need to be modulated and the receivers be filtered with a high band pass filter to eliminate interference from ambient light.
      • I might be necessary to have either / both a adjustment on the light output and / or a sensitivity adjustment on the receivers.
      • Since there are a lot of receivers (50+) a simple design on the receiver side would be great.
      • Each receiver should have an associated LED to give a visual feedback that the system is correctly aligned and that breaking the beams work.
    So, here is what I am thinking about:

    Pretty simple, just a bunch of IR diodes that are fed a amplified signal coming from an oscillator. With a potentiometer to control the amplifier gain that should pretty much do it.
    What I am thinking about for the basic "cell" (one beam receiver) is an IR sensor connected to a low pass filter, then into a comparator. The comparator is fed a voltage which is common to all cells that will serve as sensitivity adjustment.
    The output of the comparator is then fed into a jumbo NAND gate that would NAND all the inputs from all the receiver cells.

    The output of each comparator is also connected to an LED that indicates the status of the receiver.

    The output of this NAND would then drive an solid state relay... (maybe a pair, each would work in opposite polarity just in case...).
    So... here is where I need some help...

    First, some critique / comments on the overall concept. I think that speed performance wise, it should be a no brainer. Distance wise, it will probably depends on the emitters power and the mechanical design.

    If the overall concept makes sense, I'd need some help with the detail circuits and component selection. I can do all the soldering and testing no problem. My design skills are just... dated!


  2. retched

    AAC Fanatic!

    Dec 5, 2009
    Wow, that is pretty ambitious.

    I cant wrap my head around the need, but Im sure you have one.

    About using modulated beams: Each beam should be modulated at a different frequency so the paired receiver can tell it apart from the others, unless you are using laser..

    Am guessing you are going to want to use laser diodes to keep a condensed light beam.. There was or is a thread that a member just posted about laser modulation. That is a good start. He has the PCB layout in the thread. once done, you can use that design, or similar, on your project.

    It is a trivial task to make a "tripwire" circuit with a laser beam. But propagation times are going to be your Achilles heel. A CDS or photo-transistor likely will be too slow to satisfy the 1ms goal.

    This is not my area of expertise... actually, I dont have one ;)

    Others have done projects with pieces of what you want to accomplish. Hopefully the gaps get filled in.
  3. arunsjoshi

    New Member

    Oct 22, 2010
    Your curtain width of 10 metres is very large. So, you will have to use very high power laser emitters. Alternately, you can use pulse the laser diode emitters at a frequency of 2 kHz or more to meet your response time requirements. It looks as if you are into ballistic testing and want to design a light screen. Right ??
  4. marshallf3

    Well-Known Member

    Jul 26, 2010
    Phototransistors can be fairly fast but if I recall you can get photodiodes that, when operated in the biased mode, can be extremely fast.
  5. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    Look at the design for a ballistic chronograph. It is essentially what you are building.

    They use 2 or 3 phototransistors, reading diffused ambient light from a translucent arc shaped "shield" above each slot. When the light drops by a tiny amount in a very, very short time, the timer starts, when it crosses the second photodiode, the timer stops, the cpu then computes and displays velocity of projectile.

    Taking a photo of one is about the same thing, but simpler for a small detection/pre-focused area.

    I assume you want to use IR LEDs for lighting, so the shutter can be left open, and when the screen is "broken", a flash triggers for the exposure?

    If the latter case, powering all LEDs the same would be best for continual brightness, especially with the indoor/outdoor constraint. Then it is a matter of very sensitive and well focused/slotted phototransistors to "see" only directly upward.

    Let me know if this is way off track or confusing...