homemade laser spectrometer.

Why not use a prism?

?

Using a prism would be harder to do, because the relationship between index of refraction and wavelength isn't linear.  If you have a chart/table/GOOD curve-fit showing the exact index at every wavelength for your exact prism, then using a prism should be very doable, and it would even be based on first-principles instead of having to calibrate your set-up like with a diffraction grating.


ETA: Misread somethings.  Really, using the diffraction grating can very easily be based on first-principles as well.  The allowed diffraction will always satisfy the diffraction equation, so if your spacing is known, then you can find the wavelength based on the angle of the diffraction (and the angle comes in with measuring the distances from the different order diffractions and the distance from grating to screen where you measure the dot).  Still first-principles, and can still be very accurate.  

The difference in using a prism vs. a diffraction grating is, well, everything.  The diffraction gratings grating uses diffraction (duh), and the prism uses refraction.  All the calculations will be different, and either one is really doable  As long as you know the specific measurements of your diffraction grating then the first-principles calculations come out much easier and less convoluted (the non-linear dependence I mentioned above).

So, to sum up what pullbangdead has said. You can use a prism, however it is incredibly more difficult.

Kenom said:

The only thing you need to do is make sure your diffraction grating has got a small enough number of lines. I went with 500 lines per mm.
http://cgi.ebay.com/Laser-Diffracti...1QQcmdZViewItemQQ_trksidZp1638Q2em118Q2el1247

The smaller the number the more spacing there will be between the lines or spectrum. So obviously, if you want to determine the small shift in wavelength in a bluray for example, you want to move it further away from the grating than I have, and a small number of lines per mm on the grating. I think 500 is the smallest you can go.

Click to expand...

Thanks for the link. The seller has 500 line/1000 line pairs for around $10 with free shipping, so ordered a pair.

I would like to get a better idea of the wavelength variations in some different bluray diodes and even a few reds. I can actually see the effect of Red diodes wavelength increasing, because the dot gets Dimmer as the current is increased to a high level while looking through laser glasses. So, yea, good sensitivity will be needed to see a few nM difference.

Can't wait to try this!

Interesting build, true classical instrument!

It might be feasible to use a piece of a CD or DVD as the grating - if you cut a piece close to the outer edge the lines should be fairly parralel, and also in the right order of lines/cm to get usable results.

in order to get a more accurate viewing of the change in wavelength move your measuring stick further from the diffraction grating. it will increase the spaces between wavelengths the further out you go like the illustration below.

You can use a dvd or a cd. You can cut the layer of the recordable material off the plastic and shine the laser through that. it is also a diffraction grating. There are also diffraction gratings inside dvd or cd readers inside the sleds.

I had a similar idea using a LaserCheck. A bit simpler, but havent tried it yet.

1) Measure the output of a laser with a thermopile based meter

2)Adjust the wavelength on the Coherent LaserCheck until the values match.

I will have to try it sometime and see what happens, id imagine if both meters are properly calibratred, it would be a fair assumption the wavelength could be estimated quite easily

or maybe not.... :-?

Hows that 532 scratch build coming along btw?, cant seem to find the thread?

does anyone know if a multimode diode will put out an innacurate reading on something like this? Supposedly there was a batch of 635nm, 660nm, 700+nm and 808nm among this assortment of 9mm diodes. However, there appears to be only one wavelength. 635nm. I have one that literally maxes when pushed super hard at 5mw. One that peaks at 10mw, on that is 50mw, and one that I've yet to determine but is bright as hell. All of them show in the same place on my spectrometer. It's damn odd that a listing of 50 diodes of ranging wavelength would all be the same wavelength.

wannaburn said:

I had a similar idea using a LaserCheck. A bit simpler, but havent tried it yet.

1) Measure the output of a laser with a thermopile based meter

2)Adjust the wavelength on the Coherent LaserCheck until the values match.

Click to expand...

I doubt that would work well - the wavelength you enter into a meter like the lasercheck is only used to compensate for the sensor. This compensation is not linear with wavelength, and one compensation factor could be valid for a number of wavelengths.

It's damn odd that a listing of 50 diodes of ranging wavelength would all be the same wavelength.

Click to expand...

If they are all 635, you are damned lucky ;p

Ordinarily I'd agree with you. However most of these diodes peak at 5mw. This is pathetic for these 9mm diodes with huge heatsinks inside the cans. The highest I saw one of these diodes capable of was 35mw of 635nm. I'll have a few of these listed for sale possibly. I'll tell ya, a 30mw 635nm laser is briiiight.

Benm said:

[quote author=wannaburn link=1228109929/20#22 date=1228277431]I had a similar idea using a LaserCheck. A bit simpler, but havent tried it yet.

1) Measure the output of a laser with a thermopile based meter

2)Adjust the wavelength on the Coherent LaserCheck until the values match.

Click to expand...

"I doubt that would work well - the wavelength you enter into a meter like the lasercheck is only used to compensate for the sensor. This compensation is not linear with wavelength, and one compensation factor could be valid for a number of wavelengths."




Example, if you measure a 803t at 407nm, the numbers consisantly match between the lasercheck and my laserbee. If I drop the power of the 803t, the laserbee  will read at say 60mW, I wont get the same reading on the lasercheck. This seems consistant with the findings of others who say the wavelength goes up on the 803t's as the current is increased..

I dont have the time yet... but I will go deeper into that one, and test with more lasers of different wavelengths. If you have used a lasercheck before, you will notice that they are pretty accurate, and the mW always changes when you enter a different wavelength value. Going from 405nm entered down to 400nm will show an increase on the meters mW rating up to 80mW higher.

It may work in favor to have another member with an updated lasercheck try this as well, and see if there is any consistantcy there.

The only diodes I would really find the wavelength testing practical for is the 803's as they tend to keep everyone guessing

I'll tell ya, a 30mw 635nm laser is briiiight.

Click to expand...

I'm sure it'll give any dvd-burner based pointer a good run for its money in terms of visibility.

If you have used a lasercheck before, you will notice that they are pretty accurate, and the mW always changes when you enter a different wavelength value. Going from 405nm entered down to 400nm will show an increase on the meters mW rating up to 80mW higher.

Click to expand...

Well, the lasercheck uses a photodiode sensor with a certain wavelength response, which is compensated for based on the number you enter. But lets for example say that this sensor has a sort of bell-shaped curve for sensitivity, with 1 being the most sensitive. You'll have a 2x correction factor at 50% sensitivity on either side of the peak, but no way to tell which one your are looking at.

In practice i guess you can still use this approach to tell 420 or 400 nm apart, as long as the sensitivity curve is steep enough in that range (which it seems to be).

This is in line with typical photodiodes: many have their peak at the near-ir edge of the spectrum (800 ish nm) and taper off to either side. This is usually not a need bell curve, and actually very steep in the 400 to 500 nm range. Telling 655 from 635 would be an enirely different matter though, since its near flat there.

great to see others experiment with diy-spectroscopy as well!
i got me a few handheld-spectrometers back then, cardboard-style, and amazingly accurate/useable! yes, you can measure lasers by observing the dot on a wall quite well.
i got mine from here:
http://astromedia.de/shop/csc_fullview.php?Artikelnummer=406.hsp
http://astromedia.de/shop/csc_fullview.php?Artikelnummer=407.NDI

first one is the spectrometer, second one the grating only. german page. to sum up: 1000 lines per cm, in the spectrometer its perhaps 1cm² (surely large enough), the 'meter is 7.90€, the grating only is 4.50€ for 152 x 38 mm (!). they ship intl, no costs stated. drop them a mail?

i built a siy-spectrometer very similar to kenom's back then. i made a round "projectionscreen", to have a constant distance to the grating and therefore a linear scaling. so far the theory. it didnt work, somehow the angle didnt change linear. i only used "secure" ~405nm, 532, 808, 1064nm and the multitudes of them (like the second angled violet dot would be at 2x405nm = 810nm). it didnt turn out linear at all, after fiddling, calculating and researching for hours i abandoned it.
but besides that, it worked really well! the limiting factor was (here) that the dot was so fat, it covered several nm scale. but then i didnt use a slit like you, kenom.

i made a thread somewhere, lets see if i find it again.

i can only encourage everyone to experiment with that stuff too! its fun!
no grating? yes, cds/dvds work well! only downside is its projecting the dot in a circle. other than that they are fine!
know those funny "i see everything colored"-googles like here?
http://astromedia.de/shop/csc_fullview.php?Artikelnummer=275.GFB
http://astromedia.de/shop/images/MSF+Kerze1.jpg
try it, lets see if its already enough?
and yes, finally, the cd/dvd grating could work too, for first experiments at least..

manuel

found my old thread with many pics:
http://www.laserpointerforums.com/forums/YaBB.pl?num=1196518505

hey, we used the same wooden stuff as a base! ..but i have more laser-action-pics

/threadjack

manuel

Incredibly more difficult (with prisms) depends on whether you're going to eyeball it, or use a computer. With a computer, you can use the table easily. If you don't have the table for the prism in question, you could probably obtain it from a blackbody source, referenced to a known source (e.g. a HeNe laser). Additional precision can be had with a motorized lens and a galvo, so that you can "zoom in" on the interesting parts of the spectrum. That's probably cheaper than buying a proper detector, all in all.

Straight razors make good entry slits, by the way. Not too hard to set them for a very tiny opening.

If you want higher or lower period gratings, there are transmissive types down into the <110/mm range and up to 3600/mm on Edmund. Also, some people have made them by taking pictures of printouts from the computer on black and white film and using the negatives, but unless you have a good grasp of diffraction optics, that doesn't net you anything (it's rather cheap, though).

Just some thoughts.

For my own, I'm still trying to settle on a suitable imaging chip.