Detect light bulb on / off

Discussion in 'The Projects Forum' started by fender7802, Jan 8, 2013.

  1. fender7802

    Thread Starter Member

    Apr 7, 2012
    43
    0
    Hi folks,

    I have a question about designing a circuit which detects whether or not a light bulb is turned off or on. This circuit will be electrically isolated from the light bulb (a standard 60W incandescent). I have to use a dry contact sensor, which reads "open" when it sees a resistance greater than 2600 ohms, and "closed" at anything less than that. My first thought was a photoresistor in series with the sensor, but I can't seem to find one whose resistance will decrease to 2600 ohms in near full light reliably. My second thought is a photodiode, but I am not exactly sure how I would use this. If it's set up for reverse-biased conditions, and the light shines on it, will its resistance decrease? I also am using a voltage source of 3.3V.

    Thanks for the help.
     
  2. WBahn

    Moderator

    Mar 31, 2012
    17,787
    4,807
    Use on that has a higher on resistance to control a transistor acting as a switch to bring in a physical resistor having suitably low value into the sensing circuit.
     
    fender7802 likes this.
  3. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The sensitivity of photo-resistors varies a lot even if they have the same part number. Some are sensitive and others are not.
    Also the sensitivity changes when the temperature changes and when the photo-resistor ages or rusts away (caused by humidity).

    A photo-transistor or a photo-diode with your own transistor is more reliable.
    If you don't know how to use them then look in Google.

    A reverse-biased diode does not conduct. It "leaks" a small current when exposed to light or heat. A diode with no bias generates a tiny voltage when exposed to light because it is a solar cell.
     
    fender7802 likes this.
  4. tracecom

    AAC Fanatic!

    Apr 16, 2010
    3,871
    1,394
    That's odd. Radio Shack part number 276-1657 contains 5 LDR's in three different sizes. I tried a random sample of 10 pieces and they all went under 1kΩ in a matter of a few seconds when held about 2 inches from the CFL in my desk lamp. Some went as low as 300Ω.
     
    fender7802 likes this.
  5. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    2 inches is tiny.
    I think he wants the circuit to detect that the light is turned on when it is 20 feet or more away (any light in the room).

    At a distance of 20 feet the cheap Chinese RadioShack photo-resistors will be hundreds of thousands of ohms.
     
    fender7802 likes this.
  6. fender7802

    Thread Starter Member

    Apr 7, 2012
    43
    0
    That's correct, it most likely will be in an assembly which I suppose could be anywhere from 5 to 20 feet away from the light bulb. I think the best idea is to use the transistor method. So I want to set up a transistor as a switch (base bias), but my question is how do I "bring in" the resistor which has the value to close the contact switch? Is it in the collector circuit?
     
  7. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    "Transistor method"?
    Which circuit are you talking about?

    Maybe you have a reverse-biased light-detector diode that is amplified with a transistor, or you have a photo-transistor?
     
  8. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    You could use a phototransistor as your sensor, and have it drive the LED in a Vactrol, which internally illuminates a photoresistor. Sorta like a Rube Goldberg machine.:D
    Or you could tell us more about this "dry contact sensor". What is it? Does it have a part number? How about a datasheet?
     
  9. fender7802

    Thread Starter Member

    Apr 7, 2012
    43
    0
    I don't think that photodiode method would work because of the fact that the light just increases the leakage current and doesn't decrease the resistance like you said. So if I want to use a phototransistor, should that be open-base and then are you saying using another (regular) transistor as well? I'm just not seeing the part where the transistor tells the sensor to see a resistance of 2.6k ohms or less. Thank you so much.
     
  10. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Read post #8, if you missed it.
     
    fender7802 likes this.
  11. fender7802

    Thread Starter Member

    Apr 7, 2012
    43
    0
    This dry contact sensor is something I am required to use for this design (I'm an intern at an electronics company) and they use this dry contact sensor in conjunction with their cloud monitoring which tells the user whether or not something 'is' or 'isn't' (I think they use this circuit as flood sensor). I don't have access to a schematic unfortunately.
     
  12. tracecom

    AAC Fanatic!

    Apr 16, 2010
    3,871
    1,394
    Sorry, I didn't realize that the distance was 20 feet. I assumed it was closer because a 60 watt bulb doesn't give what I would call "near full light" 20 feet away. Have you asked for a lumens spec?
     
  13. fender7802

    Thread Starter Member

    Apr 7, 2012
    43
    0
    That's ok, I haven't asked about that. I probably should but I'm just focused on the design of the circuit at the moment.
     
  14. tracecom

    AAC Fanatic!

    Apr 16, 2010
    3,871
    1,394
    I don't see how you can design a circuit to detect light if you don't know how much or how little light you have to detect.
     
  15. fender7802

    Thread Starter Member

    Apr 7, 2012
    43
    0
    It's supposed to detect light or no light, so it's safe to assume a room that is pitch black compared to one that has all of it's lights turned on.
     
  16. WBahn

    Moderator

    Mar 31, 2012
    17,787
    4,807
    But what IS a "dry contact sensor"?!

    How does it determine that it has less that 2600Ω across it? Is it looking for current or voltage or something else?
     
  17. WBahn

    Moderator

    Mar 31, 2012
    17,787
    4,807
    What qualifies as having "all of it's lights on"? I have rooms in my house that are not very bright at all when all of its lights are on and other rooms that are much brighter with only a fraction of the light on. At some point you have got to deal with determining where the boundary is between the being light and there not being light. You can put this point off for awhile by building in a wide range of adjustability into your circuit.
     
  18. fender7802

    Thread Starter Member

    Apr 7, 2012
    43
    0
    I don't really know what it is, I just know that the on the computer, it says "open" if you connect a resistance larger than 2.6k ohms across it.

    Can you tell me if this circuit would work?

    http://i337.photobucket.com/albums/n386/fender7802/phototransistor_zps8eaf704e.png

    The idea is that with no light, the base current is zero, making the diode in the collector circuit open, making the contact sensor see a high resistance. With light, the diode is forward biased in the collector circuit and is given a resistance of about zero. Then, 2.5k ohms is connected directly across the sensor and it would read "closed". My question is, is this the resistance the sensor would see?
     
  19. WBahn

    Moderator

    Mar 31, 2012
    17,787
    4,807
    I don't think so. But I don't know how the sensor is detecting when it is connected to less than 2600 ohms, so I have no way of knowing. If the sensor is applying a voltage to the load and looking for a certain amount of current to be flowing, then you could be messing up, even damaging, the sensor with this circuit.

    You might be best off using a MOSFET across the sensor terminals and using the LDR in a simple voltage divider circuit to control the gate voltage to the MOSFET. But even this gets tricky if you don't know whether the sensor can have one of its terminals ground (or suppy) referenced in an external circuit.

    Without knowing more, the safest way is probably to use an optoisolator (or use a phototransistor in a fashion similar to an optoisolator).
     
  20. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
Loading...