Light Data Transmission

I am looking for ideas and suggestions on using light for data transmissions. Range will be up to 500 feet, during daylight hours, and outdoors. Want to use relatively common LEDs and photo diode detectors, due to pricing. The sending end will be moving and sending out light 360 degrees. (Looking at automobile LED running lights. Some of the bulbs radiate 360 degrees) The receiving end will have will have a very basic and crude tracking system, so possibly a very simple and broad focus lens can be used. What I am looking for is thoughts on what color of light to use. Since there will be high ambient light can I use a certain color/frequency and some type of filter to get a better signal to noise ratio than if I just use white and a sensor that will look at all the background light? --- Also has anyone been able to get long range, using simple systems without the use of lenses?

I think the project will be difficult, with light diminishing in a inverse square rule 500 ft there will be so little light you would probably need a telescope at the receiving end.

Colours will depend on the background as the sun has full spectrum light anything that is behind it will reflect all colours in the white light spectrum. A black background with high absorption flocking would be needed, but at 360 degrees will be impossible.

As far as colours go and RGB LED would be a good start with.

Thanks, Malcolm --- I agree with you about being difficult, that is why I am looking for ideas from others. I need to go ahead and run some tests and maybe learn something. I want to start with new high brilliance LED, and probably sensors that peak in the blue region. I can block or shadow overhead or open sky. Probably start by pulsing at 100 KC and see how far apart I can get a signal through. After a few tests, I will have a better idea of what is reasonable.

I think being a low cost experiment maybe indoor testing can be done to increase receiver to maximum possible then take it outside for the combined test. If you are not going to get the range indoors it is highly likely you won't get it outdoors either.

I have done quite a bit of work in optical imaging using active sources in the past. What you are trying to achieve is not trivial, Do the maths, optical units and measurement can be challenging, so get that sorted first. I have found that to get good range you need to modulate the light, and use phase detection. Additionally, a narrow spectral source, like a laser is preferable (you can get them out of old CD drives). To get wide coverage, the laser can be scanned, there are a number of ways to do that. Lasers can be modulated at high frequencies, unlike LEDs. If a laser is used, a narrow band (interference) optical filter may be used at the receiving end. The combination of modulation and optical filtering minimises the impact of DC sources, like the sun. To avoid saturating the detector with sunlight, even with a filter, put the detector down the end of a tube. This is a good low power solution. The colour probably is of little concern, although both solid state optical sources and detectors have better efficiency towards the red end, see data sheets. Check out info on light scattering in air if the air is not clear where you are, and consider that when looking at the colour to use as well.
What sort of data bandwidth are you talking about. Large area photo detectors have very high capacitance, which limits the bandwidth of the trans-impedance amplifier used to amplify the photo current.
It is possible to just AC couple at the receiving end, but you have to choose a modulation scheme that has a DC average signal, or data slicing is difficult.
Just some random thoughts, there is a lot more that could be said,
cheers,
Richard

Hi Richard I appreciate your random thoughts and will throw out some more from here as I sit inside next the stove. (Snowing, temp. at 22 and dropping)
You mention the need to modulate the light. I plan on pulse modulating (CW). I think that is the most probable method of getting the signal though.
You suggest using a laser as it can be modulated at higher frequncies. I haven't tried it yet, but expect to be able to run LEDs at 100 KC. I can also slow that pulse rate down quite a bit if would help. Would phase detection give me a better sig to noise ratio?
Your suggestion of a laser out of a CD drive is interesting, however I don't think they are very high powered.
I plan on picking the signal off the photo diode using a differentiator so the DC and low frequency noise should not bother. I can set the DC level so it does not go into saturation. Should be easy to have feedback to hold it at a given level, just need to watch maximum current at max. light levels.
As to filters, I was wondering if it benefit me to use a blue filter. In “white” LEDs quite often blue is the dominate frequency. I can block off open sky, so maybe increase my signal to noise ratio? ? ? ? ? ? ?
Thanks guys----

RED
BLACK
AND
GREEN.....

these colors are the best for LED, as far as the colors go...

You do seem to have chosen a very challenging method to communicate data in the chosen environment.

It might help to get suggestions if you could reveal what your intended application is.

What is the reason that you want to avoid (nice cheap, simple, off-the-shelf) radio systems ?

Just a thought: the infrared system in TV remote controls is chosen for its lack of sensitivity to white ambient light, so could an IR system work ?
(The receiving photodiode has an extremely dark red filter)

IR might work only if the photodetector has a really high sensitivity in that region. Richard and Vikasbly44 recommend Red and Green (and apparently black) which is on purpose. If you google air absorption spectrum (or something like that) you would see that red is the least absorbed, then green and so on.

If Bern was using a CMOS digital camera, he might have even easier time receiving the signal as CMOS cameras have high sensitivity in the red region and can have digital filters applied on incoming source.