Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Light Frequency Detection

Status
Not open for further replies.

bibbinator

New Member
Hello,
I have been googling to try and figure out how to go about designing and building the circuit described below, but alas I can't seem to find the correct terminology to zero in on the types of components I need, let alone the design.

Overview:
I would like to construct a device to measure for the presence of certain light frequency/frequencies. It could be done one at a time (choose which frequency to test for) or sample a range and display the results. A similar example to what I want to build can be found here:

**broken link removed**

Use:
Plants have certain frequencies of light they require for vegetation and fruit/flowering. When growing indoors, the critical frequencies of sunlight must be reproduced. Traditionally this is achieved through broad spectrum lighting such known as High Intensity Discharge and achieved using metal halide type or high pressure sodium. These are VERY hot and power hungry. The new type of lighting is with LEDs. The problem with LEDs is that they are not an ideal match in terms of the frequency of light they emit.

For example, red LEDs are typically just above the range that plants like most, so the best lights are ones where you can evaluate a large sample of red LEDs and choose the ones closest to the target frequency for the light.

I would like to build a sensor that can evaluate a spectrum of light, or take individual frequency samples, to assess the effectiveness of the light and suitability for use as a growing light. You can see the attachment for a diagram of the light spectrum.

LED manufactures of course provide diagram showing their LED frequency. I simply want my own independent way to quickly and simply evaluate the frequency for experimentation, and also to evaluate other lighting systems.

Question:
So, my simple question is: What type of detectors/components are suitable for detecting specific frequencies of light? I'm mostly concerned with the detection at this point, not the actual building of the whole circuit.

If you've read this post this far, thanks! Any help, advice or guidance much appreciated!

Brett
 

Attachments

  • PlantLightSpectrum.gif
    PlantLightSpectrum.gif
    21.9 KB · Views: 330
What you want is a colour sensor =) Photodiodes sound like they would work. You can pick their wavelength sensativity in a pretty wide range. Check out Digi-Key Corporation - USA Home Page and search for photodiode They sell 657 different kinds from 254nm's to 1300nm's
Note this is just the 'peak' wavelength, they'll pick up others so you'll probably want to use a filter over the sensor, filters for those specific frequencies may be hard to find. Take a look at Wikipedia's entry on photo diodes and look at the spec sheets for the diodes on Digikeys website and you should have a better clue of where to proceed. You may be able to google for photo diode circuits to get an idea of how people typically use them. With a good filter a photo diode should be pretty accurate.
 
Last edited:
Sounds like a great project, keep in touch with whatever you come up with. It's always nice when people post novel projects.
 
Not accurate enough?

Okay, I looked at some of the photodiodes and they provide the correct function I'm looking for (detecting frequency), but isn't the accuracy a problem?

The manufacturer publishes a datasheet showing the expected frequency of emitted light, and the photodiode manufacture also shows a datasheet of the expected response curves, but doesn't one of them have to be known to be accurate in order to make a valid test?

In practical terms it means that I see a red LED and it looks red, and my photodiode detector says it's red, but given the variance in each part there's no way to know if it's the actual 625 nm frequency I'm looking for.

Since the LEDs or other light sources will vary, at a minimum then I at least need to calibrate the detector with a known light source, or I need a detector that can be used to determine the frequency within some narrow range of error.

Preferably I would be able to sample the incoming light at a variety of fixed frequencies. Photodiodes could work if there were more available for different frequencies, then I could build a circuit with dozens spread out over the spectrum give more sample points.

The only other obvious solution is to use a physical interference filter with know specs to ensure only light of a certain frequency is passing through.

Anybody have any ideas? comments?
 
You pretty much have it nailed there bibb, it's not easy, not if you want that precises a measure of specific wavelengths. Diode's + filters would be required and filters would be expensive and difficult if not impossible to find or construct for such a narrow bandwidth. You could look into commercial optical spectrometers but the price would probably kill you.
 
I just thought, for calibration you can use LED's they are a narrow if not mono chromatic light source (excluding white LED's) And measuring a 'white' light source (the sun) you can get a feel for the responce curve of the various photo diode's you're using.
 
I have also had a interest in this topic for use in a greenhouse environmental monitoring (and one day control) project that I'm working on in my spare time.

I thought I'd add the little insight that I have acquired. The light spectrum between 400-700nm is what is referred to as Photosynthetic Active Radiation (PAR) or Photosynthetic Photon Flux (PPF). It is, as you might suspect, the wavelength of light that plants require for photosynthesis.

The key with measuring this narrow band of wavelength is having the proper photodiode and even more importantly the proper filter on the sensor head. There are a handful of vendors of PAR/PPF sensors out there, **broken link removed**is one **broken link removed** is another. Depending on your point of view, the price really isn't all that bad (at least for Apogee) considering the accuracy of the measurement that you get in return. Of course, if you wanted a dozen of these, then you'd really start to feel the pain.

Several years back the B.C. Ministry of Forest required a quantity of these sensors for their research. They decided to roll their own to keep their budget reasonable and were generous enough to publish the results. They experimented with various photodiodes and were able to obtain decent results in the process.

I hope to use this document as a guide in building my own sensors at some point down the road when I have the time, meanwhile I purchased one from Apogee to get me up and running and to use as a "standard" for comparison purposes when I do get around to taking a shot at building these from scratch.

At any rate, best of luck in your endeavors.
 
White LED's cover between about 400 and 700nm's but it's very peaky. You'd have to buy a white LED from a manufacture that published it's output curve in order to use it to offset the perceived brightness.
 
A piece of glass, some white paper and a tape measure. :)

Just shoot your LED through a prism. being LEDs run at very specific wavelengths they will show up as very fine lines when they are run though a prism. ;)

Yea, I know its too simple and obvious. :eek:

Use a microcontroler to run a robot around that does the measuring for you! Or use a video camera and a converter system so you can look at the colored line on your computer monitor. :D

Now do me proud and over complicate this! I know someone will! :p
 
Niceone tcmtech, I'd never have thought of that and now I'll never forget it =) Home brew optical spectrum analyzer.. A piece of glass <smirk> Add a digital camera and put it in a dark room taking short exposures and you could probably come up with a curve of the relative strength of the various output frequencies of a white LED. Old copiers laser printer or CD/DVD rom drives tend to have really high quality optical prisms in them.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top