I've recently started 3D printing my own optomechanics as well. It's such a good idea that I was surprised I hadn't heard of others doing it until I ran into this project. If you want some other inspiration I know Thorlabs let's you download Solidworks files of all their optomechanics. I have access to Solidworks thru my university so it's easy for me to convert those to .stls, but I don't know how else you can do it. One of the most useful things I've printed is a breadboard. Here is one on thingiverse: https://www.thingiverse.com/thing:3207833

Thank you for the spectrometer details. I'll keep an eye out for your builds. 

Yes, I found that in the Michelson. It was actually easy to get a quite precisely steerable mirror mount using machine screws and springs. Also I am keen to build a VIS spectrometer and to try to get good resolution, say 1 nm (if one was happy with less, the Hamamatsu C12666MA would be the way to go). Was going to build  a crossed Czeny-Turner but seems like the BOM ends up being significant with the mirrors and grating. Probably you are going beyond the DVD grating approach too, do you have a design in mind? 

I intend to make very cheap one with a DVD-transmission grating, and another with proper reflective grating. Both will have old 50mm camera optics  on either side of the grating and as few degrees of freedom as possible with regards to the optical pathway. It will probably be variations of this design (that I don't know the name of):

I have some spherical mirrors lying around so I might play with those too, but unless the above proves very impractical, that's the way I'm going.

(and in the far future I would like to look into spatial heterodyne spectrometers)

This is interesting Esben - I look forward to seeing your optical builds. I guess you are working towards your Raman spectrometer. Cheerio, Andrew

Thanks. Yes that is the plan. In the assembly and alignment of the spectrometer I've come to realize how convenient it is to have movable parts. After the Raman spectrometer I intend to make at least two VIS-spectrometers - one of them suitable for integrating in the classroom. Hopefully spectrometers will be less of a black box for my students then.