diffraction gratings
Re: diffraction gratings
Thanks Dinesh! We found the first attempt had ~1400 lines/mm. Lots to explore and improve with this setup, so we'll continue with it for the rest of the semester.
Re: diffraction gratings
I just discovered this thread today, and I wish I had sooner, to help you on get on the right track, but it looks like you were lead in the right direction. Check out the expanded Seven Single Beam Projects page on my web site, featuring the new Seven Single Beam Projects for the Web, http://edweslystudio.com/Pedagogy/7SBP/ ... e7SBP.html, with the Bonus #1: Gratings being particularly pertinent to this topic. http://edweslystudio.com/Pedagogy/7SBP/ ... ectB1.html
This is a Beta version, adding to it irregularly, maybe I'll complete it before I kick the bucket!
This is a Beta version, adding to it irregularly, maybe I'll complete it before I kick the bucket!
"We're the flowers in the dustbin" Sex Pistols
Re: diffraction gratings
Ed, I'm pretty sure the spectral lamp is krypton. Like 99% sure. The only thing preventing me from being 100% sure is that no sane person would use krypton for a discrete line demonstration.
After making a few, we gave up on the large divergence gratings. Setting the plate so far from the divergence point, there was way too much scattered light from all kinds of directions. So the gratings tended to have very Interesting looking but useless dark bands swirling through them.
Another difficulty with diverging beam transmission to produce the grating is that the mirror can be placed at multiple positions in two dimensional space to produce the same (central) angle incident onto the plate. Really a job trying to figure out the best balance of intensity and plate coverage. So this problem was sort of over-subscribed, with two spatial range variables to determine one final domain result.
So we went back to the collimated beam, a nice one inch circle where we could divert all the unwanted reflections away. And we didn't have to black the edge of the two inch wide plate.
We added another mirror to make gratings via transmission hologram (with our traditional "pulling off the protective coating" ceremony that accompanies the unveiling of a new first surface mirror)... first mirror behind the plate reflects the beam 90 degrees, second mirror off to the side reflects the beam another 90 degrees, so that it is parallel to the incident beam and traveling in the opposite direction (not through the plate). Then a third mirror positioned along the line of the parallel beam reflects it back on the plate from the incident side to produce the grating. With our 1 degree precisions in measuring third mirror angle and subsequent measure of first order light dispersion, we could make gratings from 300 - 1600 lines/mm with an uncertainty of +/- 25 lines/mm.
I totally expected the lines per mm to be proportional to the angle of incidence of the second beam, but it was not. A plot of six test plates' lines/mm vs 3rd mirror angle (which translates to Incident angle) was best fit by a shallow quadratic (parabola) curve. We set up to test the widest discrepancy between quadratic and linear fits, an angle that the quadratic curve predicted 300 lines/mm, while the linear fit predicted 350 lines/mm (not much beyond our uncertainty!) and proceeded to create three 300 lines/mm gratings in a row. They make nice holiday gifts!
After making a few, we gave up on the large divergence gratings. Setting the plate so far from the divergence point, there was way too much scattered light from all kinds of directions. So the gratings tended to have very Interesting looking but useless dark bands swirling through them.
Another difficulty with diverging beam transmission to produce the grating is that the mirror can be placed at multiple positions in two dimensional space to produce the same (central) angle incident onto the plate. Really a job trying to figure out the best balance of intensity and plate coverage. So this problem was sort of over-subscribed, with two spatial range variables to determine one final domain result.
So we went back to the collimated beam, a nice one inch circle where we could divert all the unwanted reflections away. And we didn't have to black the edge of the two inch wide plate.
We added another mirror to make gratings via transmission hologram (with our traditional "pulling off the protective coating" ceremony that accompanies the unveiling of a new first surface mirror)... first mirror behind the plate reflects the beam 90 degrees, second mirror off to the side reflects the beam another 90 degrees, so that it is parallel to the incident beam and traveling in the opposite direction (not through the plate). Then a third mirror positioned along the line of the parallel beam reflects it back on the plate from the incident side to produce the grating. With our 1 degree precisions in measuring third mirror angle and subsequent measure of first order light dispersion, we could make gratings from 300 - 1600 lines/mm with an uncertainty of +/- 25 lines/mm.
I totally expected the lines per mm to be proportional to the angle of incidence of the second beam, but it was not. A plot of six test plates' lines/mm vs 3rd mirror angle (which translates to Incident angle) was best fit by a shallow quadratic (parabola) curve. We set up to test the widest discrepancy between quadratic and linear fits, an angle that the quadratic curve predicted 300 lines/mm, while the linear fit predicted 350 lines/mm (not much beyond our uncertainty!) and proceeded to create three 300 lines/mm gratings in a row. They make nice holiday gifts!