Ten months ago when I first read about the ramanpi, I immediately bought a small collection of some of the things on fl@tc@t’s bill of materials. In hindsight I should probably have done a little more research before shopping.¹
For example I bought a 10µm optical slit. Theoretically this is a very good choice of slit width, because the pixel width of the TCD1304 is 8µm.²
This is of course in a perfectly well aligned spectrograph. Mine won’t be perfect, so I’ll go for more light throughput. However I’m not about to pay another 100€ or so for a new slit. The price pr m² for that kind of thing is terrible.
Instead I set out to make one myself, I hope it will be good enough. Anyway, enough talk. Here is my El cheapo DIY SM1-compatible slit:
The El cheapo slit consists of a 3D printed structure fitted in a Thorlabs SM1RR retaining ring. Four small neodymium magnets (Ø=2mm, h=1mm) are positioned under the two razor edges comprising the actual slit.
In this last picture the El cheapo slit is screwed into the SM1 lens tube holding the collimator lens (an achromatic doublet):
Oh and of course: here is the openscad file for the DIY 3D printed optical slit.
Update: I’ve measured the attainable slit widths to be as high as 0.13mm and as low as 25µm. These figures are very approximative.
The important thing is that the razors are easily quite accurately aligned. The human eye is surprisingly good at evaluating parallellity.
¹ I’ve already substituted the cheap ebay laser with a proper JDSU µgreen laser. I’ve also moved away from the Czerny Turner spectrograph, and so I’m left with a few spherical mirrors that I haven’t found a new use for.
² Sometime ago, somewhere I don’t remember, I saw a calculation showing the optimal slit width vs CCD pixel width.