Some one tell me how do I make a digital spectrometer.

[Inspectorcritic]

I'm on a budget and so things have got to be cheap and is what I want to do is remote testing of bio tissue to find the true ID. Okay there is always known to irradiate the subject with IR radiation or UV or RGB. Me I'm thinking remote and safe distance like using a picture irradiating that with IR now what type of a device would I make to read it? I found one light meter, but I doubt if that is going to be what I need. You can take a look at it tell me if that will work.

 

So I am think I just get a IR diode shine it on a picture and use a light meter in the other hand to see how much was absorb and reflected, right?

 

Also I have it in mind to use to view inter net pictures and determen if the tissue is real or fake or somthing else. Maybe I am thinking a color sensor with an indicator light, that would work for me too if anyone knows how to make one.

 

Notice in the light sensor they just used a digital volt meter I thought that was pretty clever and saved money. Can you tell me will that digitally inform me of the wavelength or is it going to be the intensity of lumens and not wavelength?

Edited May 5, 2013 by Inspectorcritic

[swansont]

You need a frequency discriminator of some sort. I just went to a good talk about spectrometers — I liked the line: "To build a spectrometer you need to make a rainbow." Usually done with a diffraction grating, and then calibrated with some known emission lines.

[Inspectorcritic]

Yes you can make a cheap one, but is what I needed is one for very low light so I guess I have to irradiate it with UV or IR I need to know what gizmo is needed for detecting the absorbsion and reflection. Heres a cool home made web page of one.

http://sci-toys.com/scitoys/scitoys/light/spectrograph/spectrograph.html

 

[John Cuthber]

When you say this"Also I have it in mind to use to view inter net pictures and determen if the tissue is real or fake or somthing else" it makes me think that you are hoping to do something impossible.

 

Can you explain exactly what you wan to do?

[pwagen]

Maybe this can be of interest:

 

http://www.raspberrypi.org/phpBB3/viewtopic.php?t=40614&p=336870

[dumbbloke]

Save yourself a lot of time and effort and buy one--it will work immediately and accurately. you don't have time to study digital electronics enough to become sufficiently competent.

[Inspectorcritic]

I would like to but some of those guys well screw ya real bad. I was just taking a look at diffraction grating sheet 8"x 165" roll, one had it price for $310.00 another had the same product price at $8.65 and many other abnormalities in pricing.

 

Take a look at this home made light job. It states it has a 5 gain channel one UV, RGB, and IR readings. Could I put my efforts into this unit then get either a IR or UV light diod, radiate the subject and measure the absorbtion and reflectance to give me the ID of the matter. Well it work?

http://bigbro.biophys.cornell.edu/~toombes/Science_Education/Light_Meter/Build_Your_Own_Light_Meter.pdf

 

Or well this just tell me the lumens ambient light reading?

[EdEarl]

Internet pictures such as my thumbnail on this page do not contain true colors, only red, green, and blue (RGB), because our eyes only see those colors. That we see all the colors of the rainbow is an illusion. Spectroscopy will not give the same results from internet pictures as actual things.

[Inspectorcritic]

Could I take that into account deduct it and recalibrate the spectrum from matching it to a known source then giving me an index of RGB limitations comparing it to the true source? Theres always a way, don't you think?

[EdEarl]

Could I take that into account deduct it and recalibrate the spectrum from matching it to a known source then giving me an index of RGB limitations comparing it to the true source? Theres always a way, don't you think?

I doubt it, but I am surprised by science all the time.

[Klaynos]

You can go from an RGB value to a single frequency in most cases. Colour is complicated (google colour spaces) as the human interpretation is not trivial.

 

If you had a monitor with a narrow bandwidth on the pixel RGB frequencies you could turn each single channel on in turn and use them to calibrate. Arc lamps are a better option I suspect though.

 

I have used simple ocean optics spectrometers, they're pretty good in most cases. I've also built thz time domain spectrometers, you need the right bit of kit for your purpose. If you use a grating you will need to combine this with a filter if you want to cover the whole visible and near uv and ir for the best results, depending on your grating.

[John Cuthber]

Could I take that into account deduct it and recalibrate the spectrum from matching it to a known source then giving me an index of RGB limitations comparing it to the true source? Theres always a way, don't you think?

No

Like the eye, the camera can not distinguish yellow light at 590 nm from some mixture of red light at 650 nm and green light at 530 nm.

They will both give the same output.

Since they give the same output, there is no way to look at that signal and say what the illumination was.

 

You need to have the wavelength discrimination near the object you are looking at.

[bEarthly]

2 potatoes and a length of bell wire.