LED/Photodiode advice

  1. I am currently doing a university project looking into potentially energy saving ideas. The project has a 'novel' element to it, in that ideas don't have to be "new", but for example you can take an idea and apply it to a new environment.

    I am looking specifically into lighting (as most do) and am trying to build on the idea of "intelligent street lighting" (see http://www.tvilight.com).
    I was wondering if it would be possible to set up a photodiode/LED combination in the same unit, so that in one state it acts as photodiode, in the other as a light source.

    Major problem I soon encountered (if it to be possible) was getting it from light source back to photodiode. The converse I believe could be triggered when the light received by photodiode fell below a critical level. This would not be possible for light source back to photodiode without another additional application (perhaps IR sensor?).

    Any help is greatly appreciated.
  2. jcsd
  3. berkeman

    Staff: Mentor

    Welcome to the PF.

    Are you wanting to use the "photodiode" portion to detect movement of people and vehicles at night, to turn on the on-demand streetlight? I think the sensors are a bit more sophisticated than just photodiodes, but I don't know what is used for that function. Do you know how the sensor portion is typically implemented?

  4. Thanks for your reply and I'm sort Im finding it difficult to be clear.

    I want the photodiode to act in the same way as a solar cell, and utilise the stored energy to (at least partially) operate an LED light. What I want is for the photodiode to BE the LED, so all in one 'bulb'.
    On the tvilight models I am unsure specifically what is used, that would probably of much help though.
    I don't think IR will do the job, but perhaps UV?
  5. berkeman

    Staff: Mentor

    But if there's enough light for the solar cell to draw energy, then it's light outside and you don't need the streetlights on... :smile:

    Are you thinking of charging a battery to store the day's energy to operate the LED light at night? You're not going to get enough energy to make the streetlight very bright at night. Even though LED lights are pretty efficient, they still draw a lot of power to put out the kind of light you need for a streetlight.
  6. An LED cannot be a photodiode. It doesn't have a "photodiode-like" mode of operation. (If that is what you were asking)

    BTW, In the tvilight they show a sensor unit on the pole controlling the bulb.

    You need to bullet list the features that you want to support. (make it up) This is a very important part of starting and analyzing any project. A basic marketing brochure, so to speak.

    In order to store enough energy in the daytime to operate a light at night, you will need to have a substantial solar panel and battery. That's also true even if you have line power and just want to significantly augment the power from the power line. You should do some basic calculations based on the illumination you want and the worst case operational scenario.

    I have solar powered pedestrian crossing flashers near my house that only operate when someone wants to use the crosswalk (and presses a button). They have a substantial (>2 sq foot) solar panel, and it is just an intermittently used warning light, not night-time illumination.
  7. OK --- I stand corrected. learn something every day! It's leakage current changes with light like any diode.

    The photo current is very small in an LED because the geometry of the junction is not designed to maximize it like it is in a photo diode. But it does exist.

    What are you wanting to do exactly, with the LED? Use it to sense the ambient light? Motion? I see functional conflicts as you try to use one element both as a light source and as the sensor that decides whether the light source should be on or off.

    From a practical point of view, the circuitry required to switch modes, and produce the required sensitivity, may be far more expensive than a simple photodiode.

    But first you need to clearly explain the functionality you imagine you want. How it behaves to the observer.
  8. sophiecentaur

    sophiecentaur 13,259
    Science Advisor
    Gold Member

    Absolutely. Do it 'because you can' but not because it's a good engineering solution (it can't be).
    Randomly, the idea made me think of the amazing Gramdeck, from the Fifties and early Sixties. See See here which was a poor tape recorder and would mess up a perfectly good record player.
  9. Thanks for your replies to all of you it's been very helpful.
    The intention was to detect light present, so when the daylight falls below a certain value, this would be registered by the photodiode and a switch would reverse the operation to an LED light. Main issue (like I said originally) is getting something from LED back to photodiode, there would have to be some other device to perform this operation, and so everything blows up in my face from there.

    An example they gave for the project was taking the piezoelectric plates installed in Japan in 2006 and again in 2008 that could run the ticket gates al peak hours when crowds were at their biggest. One group in previous years went on to apply the idea to street lights, offices and traffic lights.

    Im totally stuck at this point, and thinking I should totally begin again at the idea table with time currently against me.
  10. sophiecentaur

    sophiecentaur 13,259
    Science Advisor
    Gold Member

    If you have ever ridden a bicycle with a dynamo, you will be aware of the detectable extra power you need to expend when it is actually lighting a 2W filament lamp. To produce 100W of power on a continuous basis would be exhausting. The idea of Energy Harvesting does have a very few possibilities, where there is a vast amount of wasted power and where the system is already working as efficiently as possible. Pinching (significant) energy from the crowds going home from work will just serve to knacker them and they will rebel against it, once they learn about it - claiming to be much more knackered than they actually are. They are 'human'.
    If you want to think of methods to improve the Energy Situation then saving it is far more fruitful than harvesting it. Don't use power tools for small jobs, wrap up warm. Turn off street lamps when there's no one there. Keep your car for ten years and use a bike. Fit pedals to trains and buses (haha)
    Your idea is, basically, an Engineering Proposal. As such, you have to start and finish with actual numbers. How much energy do you want and where, in detail, will it come from?
Know someone interested in this topic? Share a link to this question via email, Google+, Twitter, or Facebook