There's been a lot of talk about simplicity lately...  I tend to think this is a slippery slope...  It's the old Mac OS vs. Windows argument..  An article Strive for Elegance, Not Simplicity covers much of my opinion on the matter..

Design Justification:

Obviously this isn't a simple device.. There's a lot going on.. You could probably reduce the system to a few filters, lenses, a laser and a cheap spectrometer...but then that is all you end up with..a semi-organized collection of components fashioned into what could be better. My goal with this is to create a piece of equipment that I can use for years to come.. Something that can withstand moving from one side of the room to the other without having to realign the optics and re-calibrate it..something that looks good on my bench and produces an output that I can use. All of this in addition to the stories I hear of how many people have tried to build a raman system and have achieved only mediocre results..which is why I'm guessing there hasn't been an open source raman system out until now. It's not easy and oversimplifying comes at a cost, but thankfully the reward is great when you find a balance.

I explain in fair detail how this system works below, what I don't think I covered was how a normal raman system works.. Without spending a lot of time, they're usually not based on two edge filters like I use..they typically use a notch filter which combines the function of the two edge filters into one..eliminating the need for some of the mechanics, etc. in my system... I decided to go this route because the cost of using a notch filter is far too high for the normal hacker..this adds some complexity, but has to be acceptable otherwise this project would not be worth making open source because too few would be able to build one.

I originally started with a different design (as you can see from the design logs).. That design was my first concept, and seemed logical in its approach as it was modular and was set up much like an optical table.. After much consideration, I decided to go with the more enclosed and rigid structure that you see in the newer pictures and renderings. The older design would work, and probably be ok.. But ambient light, vibration, alignment, adaptability to different optics, mounting, and a few other challenges started stacking against it.. The new design keeps ambient light and reflected laser light out of the picture... It makes alignment easier since it is a rigid tubular structure, only requiring minute adjustments.. It is much easier to adapt to different optics. Mounting the whole structure becomes easier since there are fewer points to have to attach.. Although it may look complex, the whole approach is much simpler to deal with since the parts are simply printed and screwed together. I wanted to find a balance between simplicity, durability, ease of construction, easy to source parts, and a finished looking device that won't fall apart, and require tons of maintenance..

I'd also like to cover something else....

Challenges:

As I mentioned, this isn't a simple device. It is a cross between several disciplines.. There's electronics, programming, optics, mechanical design, chemistry, and maybe a little physics.. A proper raman spectrometer is a very expensive investment.. Trying to bring this down to my goal where a middle school teen can build one and enter it as a science fair exhibit is a moderate challenge.. Designing one that fulfills that desire and will be a useful addition to my bench is another. Obviously, this is the first iteration, and I will probably follow it up with an even better design which is more elegant..but one has to start somewhere...and this is somewhere for me..

So, a major challenge has obviously been the optics.. Deciding on the Crossed Czerny-Turner Configuration was a process..I originally started out with an aberration corrected monochromator grating and was going to use the raspberry pi camera... after some suggestions, investigation, and testing, it looks like the Crossed Czerny-Turner Configuration and a Linear CCD Array are the better option..

Another challenge has been the mechanical layout.. Fitting this system into a mini ITX case was interesting, thank goodness for my linear calipers and openSCAD..  

Where it's going from here:

I'm currently working to get prototypes of the PCBs constructed and tested out..  I'm switching between that and getting the filter selection assembly designed..  

The PCB design has been a challenge in itself due to the licensing issue with Eagle, and the corruption of the files.. I posted a couple times with designs for the control board, and now I am in the process of redesigning it because I'm apparently overworked or dumb...I didn't stop to think the steppers I used in the design for the cuvette tray and filter assembly are unipolar motors and I designed the circuit using the easyDriver stepper controllers which doesn't like unipolar motors..so I'm switching over to something like the ULN2803..  They use twice the IO pins on the MCU, but shouldn't pose a problem..  I am hand wiring a prototype board before going forward with the Eagle design...this time I'll make sure the bugs are out before proceeding..funny how the simple stuff can get you sometimes..