So..  When I first started this project.. I had some ideas on how I was going to accomplish this.  I was thinking of using the raspberryPi Camera Module, and a diffraction grating I bought on eBay..along with the two edge filters, etc..  I had a lot of parts designed that were sort of an optical table style..with sliding lens holders for the edge filters.. All in open air for the most part..

Lessons learned..  

• I probably bought the wrong diffraction grating for what I was thinking of doing.. I did some more research and decided I was going to go with the crossed Czerny-Turner setup..  I am still thinking that is the best option for this device..  
• The raspberryPi Camera Module has a number of issues stacked against it.  Noise, bayer filter, sensitivity, etc.. This system would probably benefit more from a Linear CCD Array..  That doesn't mean the raspberryPi Camera Module is completely out, I may still design a lower cost option for that...using a monochromator diffraction grating (the original grating I bought actually) and rotate the grating, etc..  We'll see where that goes..
• The original design I had almost completed when I entered thehackadayPrize has been retired in favor of the new tubular design which eliminates a lot of issues...structural rigidity, vibration, stray light, mounting, and so on..plus the new design looks a lot cooler.

So, it's come time to put my brain to work and get the imaging side of this done..  I've started a research campaign to compliment all the great help people here have given me..  (I appreciate it!)..  

This document is giving me a lot of useful information..  I've never worked with spectrometers, let alone raman systems before I started this project a couple months ago...and my knowledge in that area is lacking..but I'm trying to make up for it..

Starting here, I began my planning..

I need to begin by determining the entrance slit size...From B&W Teks website, they have this blurb about the entrance slit... 

The function of the entrance slit is to define a clear-cut object for the optical bench. The size (width (Ws) and height (Hs)) of the entrance slit is one of the main factors that affect the throughput of the spectrograph. The image width of the entrance slit is a key factor in determining the spectral resolution of the spectrometer when it is greater than the pixel width of the detector array. Both the throughput and resolution of the system should be balanced by selecting a proper entrance slit width.

The image width of the entrance slit (Wi) can be estimated as:

Wi = (M2×Ws2+Wo2)1/2

Where M is the magnification of the optical bench set by the ratio of the focal length of the focusing mirror (lens) to the collimating mirror (lens), Ws is the width of the entrance slit, and Wo is the image broadening caused by the optical bench. For a CZ optical bench, Wo is on the order of a few tens of microns. So reducing the width of the entrance slit below this value won’t help much on improving the resolution of the system. The axial transmissive optical bench provides much smaller Wo. Thus it can achieve a much higher spectral resolution. Another limit on spectral resolution is set by the pixel width (Wp) of the array detector. Reducing Wi below Wp won’t help to increase resolution of the spectrometer.

Under the condition that the resolution requirement is satisfied, the slit width should be as wide as possible to improve the throughput of the spectrograph.

So... They also state above this that...

The most common slits used in spectrometers are 10, 25, 50, 100, 200 μm, etc. 

And.....200um is .2mm ...  Which I think is on the sketchy boundary of at least my 3D printer....  So that puts to rest any notion of printing the slit.  Which I had tried previously with some sad results.. They weren't terrible, but they weren't great.

So, I will start with a 25um slit.. I plan to acheive this using the 2 razor blade approach..  I would like later on to make them adjustable using tiny stepper motors..I found a bunch on ebay that should work fine for this... 

Next would be the collimating mirror.. I am currently working on calculating the focal lengths and angles required..  Not an terrifically easy task..   Ultimately, I believe at this point... since my first mirror's focal length is 80mm, that's the distance the mirror goes from the slit..  

The above images are used to calculate angles..  My collimating mirror looks like it will be at about a -15deg angle from the entrance slit..  This website explains it pretty well...

I am using a flat holographic diffraction grating for this attempt..  So, the rules are fairly simple..

The next mirror is the same concave mirror so the rules above apply there too..  This mirror will be at a -128deg angle from the grating surface..  This will probably change when I look at this tomorrow with a fresh brain...  

On a side note.. I just placed an order for these optics...  I bought a couple combinations of mirrors with different focal lengths to experiment with..  These do not have to go into the final design, and suitable eBay replacements probably can be found much cheaper.. But I am gearing this up to be done before the August 20th deadline!

ProductPriceQtyTotal

• 1200 Grooves/mm, 25mm Square, VIS Holographic Grating
  
  • Stock No. #43-216 $135.00
• 25mm Dia. x 50mm FL Protected Aluminum, Concave Mirror
  
  • Stock No. #43-466 $37.50
• 20mm Dia x 80mm Focal Length, Spherical Mirror
  
  • Stock No. #46-239 $37.50
• 50mm Dia. x 100mm FL Protected Aluminum, Concave Mirror
  
  • Stock No. #43-471 $42.50
• 25mm Dia. x 50mm FL Uncoated, Double-Convex Lens
  
  • Stock No. #32-624 $28.50