digital camera spectroscope

This simple mod allows the slide to be mounted to most cell phones, turning them into digital spectrometers.

How-To

Step 1: Salvage springs from an old floppy disk

You need four floppy disk ( 3"1/2).

Comments

[wgrube]

You can get a diffraction grating by peeling off the aluminium layer from a cd... I've tested this and it works very well.

[alltootechnical]

Yeah, I did that too

[coonass]

1) Is there a quick and dirty procedure for capturing spectra using this slide and a cell phone? 2) is there a library of spectra to which any spectrum you captue can be compared? A program for doing this automatically? If so, this would be no end of a useful tool. Parents could run quick checks for contraband substances in their kids' rooms; anyone could determine right then and there exactly what they are looking at. Narcotics cops could dispense with cobalt thiocyanate and just look at the stuff they capture to get immediate identification of the aliquot. No kidding, this would rock. Last question - how well would Motorola's i860 camera phone work with this system, if you just know off hand? I don't have an immediate need for an IR absorption spectrometer, but you never know, do you?

[blakeredfield]

Get your hands on a grating from rainbowsymphony.com and test it by placing it in front of your camera. I have several of these, that I use for educational purposes in an astronomy lab and the students love it. Never thought ot taking pictures of the spectra though! They're gonna get a kick out of this!

[LancePenney]

I've used laboratory spectroscopes to identify pure elements (eg. Hydrogen gas. Each type of atom absorbs tiny pieces of the light that hits it and reflects the rest. The particular parts of the light spectrum it absorbs is unique to that atom. The combined colors that the atom reflects appear as white light so we can't see this with just our naked eyes. If we pass the light through the diffraction grating, the light gets spread out so we can see all of it's component colors and the pieces that were taken away by the atom (dark areas). This type of picture shown from the phone is called an "Absorbance Spectrum" and is just like a "bar code" that is unique to the atom we are looking at. If we use the spectrometer get an absorbance spectrum of an unknown atom and compare it to a database of knowns, we can identify it. We also use "mass spectrometers" to identify unknown atoms. They're a lot more complex because they rely on electric and magnetic fields to manipulate the atom. In the picture is a few examples of the "emmision spectra" of a few elements. This is the same basic thing as the "Absorbance Spectra" except we only include the colors that the atom took away. This type is more commonly used in labs. I got the pictures from webpub.allegheny.edu/dept/chem/resource.php and meri.njmeadowlands.gov/lab/instruments.html

[Kent]

The most common use for a simple spectrograph is identifying types of light sources. Mercury lights are very different from sodium vapor, and they both have prominent lines (which will appear as colored images of the source) in contrast to incandescent bulbs, which produce light by heat. Light made by heat has a smooth spectrum, and the image will be a smear, like a rainbow. LEDs have more pronounced lines than thermal sources, especially red LEDs. As you get to blue LEDs the lines are smeared more. White LEDs use a UV LED and a phosphor, so it spectrum looks a lot like thermal. Some have a yellow LED too to make the light less blue. I bet you could see that with this spectroscope. I would love to see a picture of a UFO taken with this. It would be interesting if the UFO was using mercury vapor bulbs. That would suggest a mirage of distant street lights. Fireworks could be interesting. You could probably identify the elements used to produce the colors.

[mju.cat]

This is really awesome! The only thing I'm having difficulty with is trying to analyze the resulting image - does anyone know of any way the spectrum can be analyzed for specific absorption bands? Or is a comparison with a full spectrum the only way to go about it? If there's a way to do this with even some vague accuracy, I can use it for protein estimation without having to wait in that long queue!

[das300]

Have u given any examples of what it can be used for ?

[cvxdes]

Spectroscopes can be used for thousands of things. They use them in drug tests some times, but mostly in chemistry.

[rusty0101]

At question may be what can this spectroscope (a cell phone camera with a diffusion grating in front of it) be used for. The quick (and actually wrong) answer is 'not much'. A longer answer is that if you have the right software in the camera, or where you send images to, you can do field data collection of spectra to see what various substances in the world contain. One thing to consider as part of that is that you will need more equipment to make good use of this tool, as you will want to isolate the light being reflected (or emitted) by the material in question, from the ambient light in the area. It is also best if the sampled source is narrow enough that clearly defined lines or dots can be sampled, to compare against the appropriate spectra samples. Besides chemistry, spectra photography is used in the analysis of the light emitted by stars to identify both the contents of the star in question, and the redshift of the star to determine it's speed of movement towards or (usually) away from us. (the camera phone slide probably won't give you enough information to do redshift analysis of anything you can photograph with that camera. Sorry.)

[chooseausername]

I'm afraid you can't take pictures of a full spectrometer using a digital camera ... Most digital camera use a RGB filter on their censor, and most RGB filters don't cover the full spectrum. Most of the time, you'll get a partial spectrum with a lot of holes because of the RGB filter ... I think it would work better with a gray-scale censor (surveillance camera for instance) : you'll get a gray scale picture of the spectrum (color intensity instead of color+intensity), but it should be more complete than with a color censor. I can't find diffracting grating here ... where did you buy yours ?

[BlueFusion]

It's "Sensor" not "Censor" and they don't have a RGB 'filter' but they MEASURE red green and blue individually. This does not mean you cannot get a full spectrum. The human eye measures red,green and blue separately also. So anything you can see with your eye, you can capture with a digital camera and reproduce with a screen. The human eye as well as camera sensors work by interpreting halfways. Yellow light affects red and green, hence is interpreted as halfway i.e. yellow.

[chooseausername]

Thanks for the correction about spelling "sensor".

About "RGB filter", I wanted to mean "Color filter mosaic" or "color filter array" of course : Each pixel of the sensor having a different color filter (R, G or B).

I know that the human eye uses a similar method.

But like I explained in my next comment (the one with bold writings), if I said that, it's because I've personally noticed that none of my cheap DC or Camcorder were able to capture spectra correctly (ie, captured spectra were less complete than when I watched them with my nude eyes).

For instance, a captured spectrum of the direct sun-light (which is supposed to display some absorption lines) had a giant hole between one or several narrow B strips, and an incomplete G band, etc ... It looked like the spectrum of a fluorescent lamp.

My interpretation of this phenomenon :
- R, G & B filters used by the sensors of my DC and Camcorder are of a too bad quality (narrow B filter, composite G filter, and large R filter), and I've wrongfully generalized by saying that color sensors were unable to capture complete spectra ...

Then, DoreJC sent me a private message and gave me a link to a French site showing some very complete and perfect spectra captured with a simple webcam. And that's how I went to the conclusion that my cheap DC and camcorder must be really terrible ...

[Meispantyshot]

You get those from rainbowsymphony.com

[chooseausername]

Hey thanks for the link ! They're definitively less expensive than what I thought ! They cost only $US 0.4, and someone was trying to sell me one for 5 Euros ........... I can't believe that ...

[dorejc]

$US 0,4 each if you get 50 glade ...... and the internationnal shipping cost to France is $US 42,50. TOTAL $62...... be carefull

[chooseausername]

Oh, thanks for pointing out that !
I did not read the bottom of the page .......... :o/

So yes, in that case, 5 euros is a better price ! =o]

[chooseausername]

I got your message.
J'ai eu votre message.

About my previous comment, I wrote that because I failed to film or photograph a full spectrum using my digital camera (Kodak C310) or my camcorder (Samsung VP-D371).
Au sujet de mon prEcEdent commentaire, j'ai Ecris cela parce que je ne suis pas parvenu A photographier un spectre complet avec mon APN (Kodak C310) ou mon camescope (Samsung VP-D371).

So I guess that some cheap DC and Camcorder will not make it ... unless we want inaccurate spectrum ...
Donc je suppose que certain APN et Camescopes bon marchE sont A dEconseiller ... A moins de vouloir des spectres incomplets ...

Just for reference, could you, please, tell us what device you tested it on ?
Juste pour rEfErence, auriez-vous, s'il vous plait, l'amabilitE de bien vouloir nous indiquer les appareils sur lesquels vous avez rEalisE vos tests ?

=o)

[dorejc]

This device was tested with three cell phones . Nokia 7373 ( 1.3Mp ) not so good. Nokia N80 (3.2mp ) pictures ok Nokia N95 ( 5 Mp ) picture OK

[andrew_29]

dude just cut and bend some safety pins it would work the same

[GorillazMiko]

Looks cool, especially those last two pictures (same, right?). Great Instructable!

[thebboy]

what actually is this and what does it do?

[alvincredible]

it separates the light so that you can see individual colors! btw, that mount is ingenious! it can be applied to so many things! thanks for the info!

[thebboy]

coolio