Finding wavelength and intensity of a specific light

[johnny1985]

Hi people

not sure if this is feasible but this is the idea and like to hear your comments

i wan 2 shine a specific color of light (could be colored LED, normal light,)

to determine the wavelength n intensity of it,

so i am thinking of using a photodiode and an opamp connected with a arduino ic to interface with my PC so that i can determine its wavelength and intensity.

 

[Simon Bridge]

You also need the frequecy responce of the photodiode.
LEDs are not monochromatic. Few light sources are.

 

[johnny1985]

You also need the frequecy responce of the photodiode.
LEDs are not monochromatic. Few light sources are.

but from this i found, is it not doable?

https://www.chem.wisc.edu/deptfiles/mcclain/Lab 4 LED Spectrometer 2010.pdfnot sure if this is the correct way to replace the NI-USB6008 interface,
but my idea is instead of using NI-USB 6008 i will replace it with arduino UNO to my PC and use ENERGIA to code the arduino and finally use matlab/python to plot the graph,

 

[Simon Bridge]

Of course it is possible to build a digital spectrometer.
http://www.arborsci.com/rspec-explorer
Does not change comment.

You need to make sure the PD can detect the light you want to measure, and you should allow that the light is probably not monochromatic.
Note: the detector in the link only detects the presence of light. The author suggests using different color sources for their known characteristics. You'll probably do better using a CCD.

http://sci-toys.com/scitoys/scitoys/light/spectrograph/spectrograph.html
http://nznano.blogspot.co.nz/2011/12/homemade-spectrometerspectrophotometer.html

 

[johnny1985]

Of course it is possible to build a digital spectrometer.
http://www.arborsci.com/rspec-explorer
Does not change comment.

You need to make sure the PD can detect the light you want to measure, and you should allow that the light is probably not monochromatic.
Note: the detector in the link only detects the presence of light. The author suggests using different color sources for their known characteristics. You'll probably do better using a CCD.

http://sci-toys.com/scitoys/scitoys/light/spectrograph/spectrograph.html
http://nznano.blogspot.co.nz/2011/12/homemade-spectrometerspectrophotometer.html

so replacing the PD with a CCD array in my link will do better?

but for the llink, if i were to shine a monocolor light it will give me the wavelength right? or wrong?

 

[meBigGuy]

Wrong

The pd tells you only about the intensity of the source. Nothing in the pd output signal has direct information about the wavelength of the source.

A photo diode is a (relatively) low speed light intensity detector.

The link is telling you how to accurately determine how much red light is absorbed by a solution. It is only about intensity.

 

[DrClaude]

You need a monochromator.

 

[Simon Bridge]

The simplest way to build a spectrometer is to pu the light through a diffraction grating and photograph the results.
Note... early spectrographs used black and white photography. Read the links I gave you and make sure you understand how the devices work.

 

[johnny1985]

ok, but how come in the link i provided, they were able to use LABVIEW to plot "intensity vs wavelength" graph??

i will continue to understand the links you guys have provided and hopefully think of something,

 

[meBigGuy]

ok, but how come in the link i provided, they were able to use LABVIEW to plot "intensity vs wavelength" graph??

The intensity vs wavelength graph was not plotted by labview. (I'm assuming you are talking about fig 1). That is a reference plot they are supplying to confuse you. Actually, it is just a reference of the absorbance of their sample. It was not plotted by MATLAB. You can multiply the reference LED spectrum (fig 2) and sample absorbance, scale, then integrate, and that will give you the single value result your PD would measure.

Is there somewhere else in the article where they infer or plot intensity vs. wavelength?

The diffraction grating/photograph suggestion is a great way to achieve what you want.
Or, a diffraction grating, then move the PD through the image, use several PD's, use a CCD, or pick a spot you are interested in. Depends on what it is *exactly* that you are trying to accomplish.

 

[johnny1985]

in figure1 they show the Absorbance Spectrum of Grape Kool-Aid

and in figure 2 they showed Output Spectrum of Red LED

or missed out something?

 

[meBigGuy]

in figure1 they show the Absorbance Spectrum of Grape Kool-Aid
and in figure 2 they showed Output Spectrum of Red LED or missed out something?

As I said above, those 2 figures are reference plots to give you some idea of what is going on. They were not computed by MATLAB. Matlab will compute 1 number which is the intensity of all modulated light, regardless of frequency.

A photodiode can only give you a measurement that is the sum of all light frequencies to which it is sensitive. If you want to know the frequency of a light source, you need some sort of optical filter (prism, diffraction grating, colored filter, etc) that breaks up the light before the photodiode.

 

[johnny1985]

As I said above, those 2 figures are reference plots to give you some idea of what is going on. They were not computed by MATLAB. Matlab will compute 1 number which is the intensity of all modulated light, regardless of frequency.

A photodiode can only give you a measurement that is the sum of all light frequencies to which it is sensitive. If you want to know the frequency of a light source, you need some sort of optical filter (prism, diffraction grating, colored filter, etc) that breaks up the light before the photodiode.

so by inserting an optical filter will help the PD detect the wavelength?

 

[meBigGuy]

not exactly. It will only tell you that the frequency of the intensity measured is within the passband of the filter. You need a different filter for every band you are interested in.

That is why the diffraction grating is a preferred solution. It spreads the light based on frequency, so you can look at the intensity at every position to determine the intensity vs frequency.

 

[johnny1985]

not exactly. It will only tell you that the frequency of the intensity measured is within the passband of the filter. You need a different filter for every band you are interested in.

That is why the diffraction grating is a preferred solution. It spreads the light based on frequency, so you can look at the intensity at every position to determine the intensity vs frequency.

even though i am interested in only 1 color i still have to use the diffraction grating? cause i thought by shining a 1color LED will be a more straight forward case

 

[DrClaude]

even though i am interested in only 1 color i still have to use the diffraction grating? cause i thought by shining a 1color LED will be a more straight forward case

First, as Simon said, LEDs are not monochromatic. Second, if you want to know at what wavelength the intensity of light is maximum, you need to be able to scan the wavelength going into the dectector. That's why a said you need a monochromator. If you want anything remotely accurate, you will also need to do some serious calibration.

By the way, the document you have been citing describes a lab on colorimetry, where the goal is to measure the concentration of a solution, and not characterize in any way the light source.

 

[meBigGuy]

If you want to determine the frequency of a light source, you need something that responds differently to different frequencies within your band of interest, and some way to calibrate it. Also it must respond such that you can get the precision, sensitivity and interference rejection that you want.

You have not really described exactly what you are trying to do. What possible range of sources, what required accuracy, what intensity sensitivity, what frequency selectivity, how fast, what budget, etc etc.

 

[Miles Whitmore]

I TA'd a lab once where we had the students measure the emission spectra of different light sources using a spectrometer. It output a nice plot of intensity vs. wavelength onto a computer. I think it's used more for educational purposes than state of the art research, and is more expensive than building your own, but since I don't know what you plan to use it for the link is below, as well as a plot of Helium emission it was used to measure.

http://www.vernier.com/products/sensors/spectrometers/visible-range/svis-pl/

The students were able to measure the spectra of a variety of sources: atomic emission spectra, a white LED lamp, the fluorescent lighting in the room, and even the light reflected by brighter objects in the room. To measure emission I think you have to also buy an optical cable sold separately, since it's primarily for measuring absorption spectra of samples in cuvettes.

 

[sophiecentaur]

The simplest way to build a spectrometer is to pu the light through a diffraction grating and photograph the results.
Note... early spectrographs used black and white photography. Read the links I gave you and make sure you understand how the devices work.

That's OK if you just want an 'idea' of wavelength and intensity (I'm assuming you mean a digital photo) but, unless you use a monochromatic sensor with a known response, then just looking at the RGB values (or the computed Y value) on a computer file of the camera output will not give an accurate value of the light intensity over the range of wavelengths. Having said that, it could be an instructive exercise as long as the reality of the limitations of using a digital camera is appreciated. It goes without saying that the exposure and colour balance must be set to Manual, with fixed shutter and aperture settings if there are a number of pictures involved. If you don't do that, the camera will be busy trying to make all its pictures look 'right' according to the rules it's been programmed with and it will give misleading (quantitative) comparative readings.

 

[Simon Bridge]

This correct... details are in the links above.

 

[sophiecentaur]

This correct... details are in the links above.

Yes but the contents of those links may be heavier than necessary for someone just considering how to get 'a result'. Links tend to generate more links and it can end up being unreasonably divergent so I was trying to condense the problem into a paragraph. It's not hard to obtain a spectrum with a slot and a prism (no box required even, if you do it in a dark room) then you can take a snap with a digital camera and look at the values on the file. Quick and dirty and a lot cheaper than what the advert is showing. Refinements can follow.

 

[Simon Bridge]

2nd link post #4
Its a plastic pipe with a diffraction grating, includes how to use with a camera and an program to generate a spectrum graph.

 

[sophiecentaur]

2nd link post #4
Its a plastic pipe with a diffraction grating, includes how to use with a camera and an program to generate a spectrum graph.

I think we're reading the OP differently and I see where you're coming from. Imo, the wording suggests either a problem with the language or a limited knowledge of the problem. That link certainly tells you how to build a spectrometer but I wonder how user friendly it actually is for a beginner. We shall see when the OP tells us how he has got on with the project.