Make Your Own Spectrometer for Under $1

[HydroSean]

Spectrometers can be used to measure the wavelength of your laser. You can easily spend thousands of dollars on precision spectrometers, but for the hobbyist this one is incredibly cheap and fun to build. Granted the resolution is much lower but it still works just fine.

Disclaimer: You must wear the appropriate safety glasses for your laser, I am not responsible if you damage your eyes or anything else.

Here is what you will need (click for link):

• Diffraction grating - any diffraction grating will do, but you get a pack of 25 here for $13 which is about 50 cents a piece These will work with a 7W NUBM44 so long as you don't shine it on the paper part.
• Safety glasses
• Laser
• Large open space
• Measuring device

This is a picture of the ones I purchased. NOTE Mine are 1000 lines/mm, we will need this number later in the calculation of your wavelength.

Set up your diffraction grating parallel to a flat wall and measure the distance from the wall to the diffraction grating (I know my garage isn't flat but you should find a flat wall to do this with).
NOTE: The further your grating is to the wall, the more accurate and precise your measurement will be

Shine your laser through the diffraction grating making sure that the beam is extremely perpendicular to the diffraction grating. You should see something like this on the wall, one spot in the middle and two spots on each side. Measure the distance between the middle spot and one of the side spots.

Now you're done! Just calculate your wavelength with the Fraunhofer diffraction formula shown below

where "n" is the grating order (here n = 1), the desired wavelength is lambda (in meters), theta is the angle your diffraction grating (in degrees) makes with the diffracted spot on the wall, and the grating spacing is d. For a 1000 line/mm grating (d = 1.0 x 10^-6 m).

Here is an example calculation (make sure your calculator is in degrees, not radians):
The distance from the diffraction grating to the wall was measured to be 180.0 inches and the distance between the middle spot and the side spot was 90.7 inches.
To calculate theta; theta = arctan(90.7/180) = 26.743
To calculate lambda; lambda = (1.0 x 10^-6)*sin(26.74) = 4.499 x 10^-7 m
To go from meters to nanometers, multiply by 10^9, so the result is 449.9 nm!

If you are confused, please refer to Cyparagon's thread created much like this thread for an alternate explanation below.
http://laserpointerforums.com/f44/make-your-own-laser-spectrometer-62605.html#post890672

 

[paul1598419]

This is how I used to measure mine before I got a spectrometer. I bought a single sheet diffraction grating about 20 cm X 60 cm and cut it to the size I needed. I always used metric measurements, though. Same pitch too. 1000 line/mm.

 

[HydroSean]

This is how I used to measure mine before I got a spectrometer. I bought a single sheet diffraction grating about 20 cm X 60 cm and cut it to the size I needed. I always used metric measurements, though. Same pitch too. 1000 line/mm.

The beauty of the whole thing is it doesn't matter what measurements you take because they get converted into degrees, then multiplied by the diffraction grating order. Heck you could use lengths of string if you wanted to, the point is that its extremely cheap and fun, which is great because it gets people excited for science!

 

[diachi]

The beauty of the whole thing is it doesn't matter what measurements you take because they get converted into degrees, then multiplied by the diffraction grating order. Heck you could use lengths of string if you wanted to, the point is that its extremely cheap and fun, which is great because it gets people excited for science!

Still best to use the same units if you can, but it does work fine either way. Switching between imperial and metric is how you accidentally crash probes into Mars after all .

Fun experiment - I remember doing it in high school - never had to use it, but it's handy to know about if you want to figure out the wavelength of a laser in a pinch.

Edit: Also - if you aren't good at math then there are online calculators that you can use. Just plug in your measurements and they'll spit out a wavelength.

 

[Mattronium]

Also, if the answer doesn't look right, check the measurements to make sure that they are in meters. The most common error I make is to enter it in mm etc. So make sure to use the correct units.

It also is worth emphasizing a little more that the variable "d" is the distance between slits. So d is the reciprocal of the lines/mm so a 1000 lines/mm is 0.001mm (note that it is in mm, not m), and 500 lines/mm is 0.002mm between slits.

Just some thoughts.


What I do is have an excel document that I enter in the formula and keep track of my previously calculated wavelengths.
If I want to set up the equation to give the answer directly as nm (10^-9 m) I use:
λ = (1000000/d)*SIN(ATAN(s/l))
d = diffraction grating lines/mm
s = distance between orders
l = distance to wall
λ = wavelength

Otherwise, the standard equation is λ = (1/(d*1000))*SIN(ATAN(s/l))


And just a pet peeve, but if you want to be scientific, keep track of your sig. figs.

 

[paul1598419]

Just noticed there's a tilt to your grating as one of your first order principal maxima spots is higher than the other. Might also be a good idea to mention that there may be more than the 0th and 1st order principal maxima.

 

[Cel]

Great info on this thread! :beer:

How accurate is this method? (try it on 532 or 473)


Oh, and gratings sometimes melt

 

[Pman]

So just how accurate are we talking here as I want to try this. I have only one difraction grating and I'll buy more but if it won't differentiate between say a 473nm and 470 or 467nm then it's not going to be worth it for me to mess with. Just how many nm will it differentiate between?
It's mentioned that they might burn which is definately something I'm concerned with as we are talking 4-6W with a G2 or less with a 3 element. I just have a hard time believing it's going to be accurate enough. I'm really hoping I'm completely wrong about that though.
Ugh, sounds like I'm not appreciating this and that's absolutely not true. I just need something more accurate than its a 445 and not a 462nm if you follow me without having to send the lasers away. It's not like it's unbelievably important either but I have some units that are quite a bit different than the norm. Thanks guys.

 

[AaronT]

I just want to find some glass diffraction grating.

 

[HydroSean]

Also, if the answer doesn't look right, check the measurements to make sure that they are in meters. The most common error I make is to enter it in mm etc. So make sure to use the correct units.

It also is worth emphasizing a little more that the variable "d" is the distance between slits. So d is the reciprocal of the lines/mm so a 1000 lines/mm is 0.001mm (note that it is in mm, not m), and 500 lines/mm is 0.002mm between slits.

Just some thoughts.


What I do is have an excel document that I enter in the formula and keep track of my previously calculated wavelengths.
If I want to set up the equation to give the answer directly as nm (10^-9 m) I use:
λ = (1000000/d)*SIN(ATAN(s/l))
d = diffraction grating lines/mm
s = distance between orders
l = distance to wall
λ = wavelength

Otherwise, the standard equation is λ = (1/(d*1000))*SIN(ATAN(s/l))


And just a pet peeve, but if you want to be scientific, keep track of your sig. figs.

Yes, significant figures is a very crucial part to making measurements in industry and research, but I tried explaining this as simply as I could without deterring people with more in depth calculations. Perhaps I, you, or someone else could create another thread going into detail and I will link it at the end of this thread. Or post it in a comment and I'll paste it in at the end of the thread (giving credit as well)

Just noticed there's a tilt to your grating as one of your first order principal maxima spots is higher than the other. Might also be a good idea to mention that there may be more than the 0th and 1st order principal maxima.

Yeah the tilt is definitely something I should correct, I will take a much better picture when I get home and repost that. Good catch!

Great info on this thread! :beer:

How accurate is this method? (try it on 532 or 473)


Oh, and gratings sometimes melt

Yeah gratings do melt so watch the power! I haven't calculated the exact precision but I will do that in my off time on the weekend and post it at the end of the thread

So just how accurate are we talking here as I want to try this. I have only one difraction grating and I'll buy more but if it won't differentiate between say a 473nm and 470 or 467nm then it's not going to be worth it for me to mess with. Just how many nm will it differentiate between?
It's mentioned that they might burn which is definately something I'm concerned with as we are talking 4-6W with a G2 or less with a 3 element. I just have a hard time believing it's going to be accurate enough. I'm really hoping I'm completely wrong about that though.
Ugh, sounds like I'm not appreciating this and that's absolutely not true. I just need something more accurate than its a 445 and not a 462nm if you follow me without having to send the lasers away. It's not like it's unbelievably important either but I have some units that are quite a bit different than the norm. Thanks guys.

That depends on how far away you put your grating to the wall. If its less than a meter or so you most likely will not be able to have enough resolution. If you put it further away from the wall you will be able to see the different wavelengths

 

[CurtisOliver]

Someone should try this out with a HeNe as we know that the emission spectrum is very narrow. It should be a good test for determining the accuracy of the setup.

 

[Pman]

Are there not glass ones then or are you saying they are too expensive?
My curiosity is high enough to buy a stack of those to try it out. I can certainly try some known wavelengths such as 532nm. I would assume my 473 jet and dragon 589nm lasers to be pretty much right at those wavelengths. They are both just over 80mW peak so I don't see them having a burning issue with the grating.
Paul, did you do comparisons between the grating and your spectrometer?
I guarantee you that I'm not the only one intrigued by this inexpensive method.

 

[CurtisOliver]

They do offer glass diffraction gratings, but you could always use ND filters to limit the power. DPSS lasers should be pretty close to their stated wavelengths.

Thorlabs sells a 12.7x12.7mm 1200 lines/mm B270 glass diffraction grating for £55.10 ($74.97)
Thorlabs link

 

[Pman]

Well I just ordered that pack of gratings from Amazon. What cheap filter would I need that they sell at Amazon to knock the power down enough for lasers in the 3-7W range? I see a 3 pack of 3 different ones but it's meaningless to me.

 

[Cel]

Well I just ordered that pack of gratings from Amazon. What cheap filter would I need that they sell at Amazon to knock the power down enough for lasers in the 3-7W range? I see a 3 pack of 3 different ones but it's meaningless to me.

Maybe a reflection off glass would suffice. Just be careful where the other photons land.

 

[CurtisOliver]

If you use a 2 OD Filter, you will reduce your output to just 1%.
I have given a link to a absorptive filter, but with lasers like the NUBM44 and M07E maybe best to reflect off at 45 degrees to avoid thermal damage.

Absorbtive Option

Reflective Option