Make Your Own Spectrometer for Under $1

Sure, and you need to be sure the grating is the actual lines/mm advertised as well.

If it's something plastic you buy for $1 off ebay i would expect there to be a fairly wide tolerance on that. Those things are intended for funky optical effects and such, not precision instruments.

You could even get better results using something like a DVD as the grating as those are usually produced within reasonbly tight tolerances to assert reliable playback. With CD's this is not really the case as there are some discs that 'cheat' the standard a bit to fit on a few more minutes of audio and part of that can be decreasing the track spacing to a value out of standard but still tolerable to most playback devices.

Meh. Just measure a good 532nm DPSSL or a 405nm. If it comes out within a 0.5nm at 3 or 4 meters, it should be fine. Don't measure a cheap red as I have measured some that claimed to 650nm and turned out to be 660nm.

If you have a hene, then the spectral line width is narrower than a DPSS. That would be a clear test of accuracy as it should read 632.8nm exactly.

paul1598419 said:

Meh. Just measure a good 532nm DPSSL or a 405nm. If it comes out within a .05nm at 3 or 4 meters, it should be fine. Don't measure a cheap red as I have measured some that claimed to 650nm and turned out to be 660nm.

Click to expand...

405nm is no good for this. They vary by as much as 10nm.
It has to be DPSS or gas.

paul1598419 said:

Meh. Just measure a good 532nm DPSSL or a 405nm. If it comes out within a .05nm at 3 or 4 meters, it should be fine. Don't measure a cheap red as I have measured some that claimed to 650nm and turned out to be 660nm.

Click to expand...

Will do, the only laser(s) I have that operates DPSS is 532 but I do have a bunch that when put on a grating only shows one line. I should be giving it another go a little later this morning.

Edit: I don't have any gas lasers unfortunately!

Yeah so using a known wavelength laser is like what industries do when standardizing/calibrating instruments. If its a nm high or low then you know to add or subtract that to your measurements (without going into statistics and what not)

It's not that simple because the relationship of the measured distances to the wavelength isn't linear. It being off by 1nm near 532nm doesn't mean it will be off by 1nm near 650nm.

Little late to the party ... saw my name mentioned earlier in the post.

Yes, that picture shows all wavelengths as the same colour but cameras aren't good at picking apart wavelengths that are fairly close together, at least, most cameras anyway. Not to mention having your monitor actually reproduce those colours with any sort of accuracy.

Single line Argon or HeNe are your best bets for testing accuracy here, preferably both, as that gives you two sources of calibration at either end of the scale - the wavelengths of those won't drift noticeably and they are narrow line compared to other sources most of us have access to.

I'll need to pick up some gratings and test this out myself with the 301 I have... I need a bigger laser budget...

I have the gas lasers, but because a lot of people here recently don't, that's why I didn't mention them. If you had a 6328 Angstrom He-Ne and a 488nm single line argon, which I do, they would be the preferable ones to use. I was just trying to make it as easy for him as possible.

So I give up! Well not really, I ordered the linked gratings to give it another go. I have a sheet from the same seller that says 1000/mm but idk. I know for a fact my grating was perpendicular to the wall, and nothing moved while testing. There was alot of time spent making sure my little rig had all perpendicular cuts.




I'm still getting 527nm as my 1st order, and on my single line 532s the same. So something is definitely wrong. I also noticed that while sliding my grating in the track I made, just over a hair and the gaps between each order deviated by about 1/4". Could it be that the baseball card holder is causing the grating to be bowed the tiniest amount thus throwing off the final numbers? It has too, I'd be willing to blame user error on myself but I spent alot of time to make precise 90° angles.

Cali, is the first order wavelength brighter or dimmer than the next one? This may indicate whether or not it is actually 527.

To be honest, while it operates with 6 lines the first order is definitely the brightest, but once it changes to 8 lines the brightest spot goes down the line to 4-5 order. Although that doesnt seem to make any sense.



Edit: heres a picture of it, (on right, the one I sold is on the left) while operating in 6 lines and the 1st order is definitely the brightest.

With a flexible grating you'll never get a perfect measurement. It's why I gave up spectro'ing my lasers with this sort of rig.

Would you please link to where you got those freak multi line 532nm lasers?

Does a known (single line) 532nm pointer register as 527nm as well? Is the distance to the first order on either side - left and right of the main beam - the same distance? If so, The line count might be off slightly. I would use the equation with 532nm held constant, and calculate the line count from from that instead of the other way around.

CaliKirk, if you PM me with your address, I'll send you a large piece of the one I still have in an envelope. No charge. It is definitely 1000 lines/mm.

Atomicrox said:

Would you please link to where you got those freak multi line 532nm lasers?

Click to expand...

I only have the link to the sellers store - LEDgroup - I don't really remember which exact listing it was as e bay deleted my previous purchases from earlier this year. The thing I do know are the lasers were labeled "laser 301". I went through many different sellers and had worse luck with ones labeled YL laser, SL laser, Laser 303, 303 laser, 301 laser and the ones I kept rebuying did not have a star cap and were labeled laser 301. Sorry for not really helping on that, ill try to find out! It definitely was a gamble through every seller but this one i purchased the most from.

Cyparagon said:

Does a known (single line) 532nm pointer register as 527nm as well? Is the distance to the first order on either side - left and right of the main beam - the same distance? If so, The line count might be off slightly. I would use the equation with 532nm held constant, and calculate the line count from from that instead of the other way around.

Click to expand...

The distance between 0 and 1st order on each side are the same. I think either the count is off, or possibly that the rig I have is causing the grating to bow a tiny amount, cause I am using a cut sheet. I did however slide the grating a small amount after I finished marking on the wall and with the 8 line laser on, I noticed the lines separation distances changed. Each separate test of known (single line) 532, the 1st order I came up with 527.75, 526.05, 526.08, and 526.9. For the 8 line on 1st order got 526.95 & 527.81 on separate tests. :wtf:

I know i know, im doing something wrong.. :undecided: