I got my c12880MA and Arduino breakout board this week and I am starting to refine and comment the Processing code for a better display of the spectra. So far I have added the calibration data for the chip and stretched  the display for the pixel output rounded to nm. I will post the code and any other files like the pixel calibration conversion  with instructions to a new project. Thank you guys for making this chip accessible to a broad hardware platform.

I did some experimenting with this. you  get a response, but the signal is extremely weak, and the resolution is not good enough to do traditional raman matching to the published databases of chemicals.   

Just wondering if you were using the 405nm laser diode for your tests? I am hoping to use a 532nm laser, I am thinking that it may cut down on fluorescence. Also do you think it would be worth while to cool the board to cut down on any dark noise?

Hi,

It is possible to sharing or sending gerber files of breakout board?
Thanks

What about using a fully blown spectroscopy software to run your spectrometer setup? I'ves just started working on this: https://hackaday.io/project/11264-universal-software-for-diy-spectrometers

There is a new Hamamatsu micro spectrometer out: c12880MA. Is this project compatible with it? Or doea it need to be updated?

Thanks

there is an updated board being sold here: https://groupgets.com/campaigns/278-c12880ma-mems-u-spectrometer-round-9

Did the boards come?

boards did arrive, you can buy them now here: http://www.pureengineering.com/store

I did have some sensors as well, but they sold out nearly instantly.  I have more sensors arriving this month sometime. 

Did the boards come?

HPK is a niche manufacturer. Don't expect from them mainstream prices. Do expect great niche products. Most higher end spectrometers run HPK linear sensors. 

I need the IR version for my project, so I'm forced to pay ~$500 to buy a single unit (C11708MA). Can I purchase your PCB separately without the spectrometer mounted?

does your hardware or software correct for the uneven spectral response?  

http://www.hamamatsu.com/resources/pdf/ssd/c12666ma_kacc1216e.pdf

The Hardware just buffer's the I/O and provides additional hardware. The correction would have to happen in software. Right now there is no correction in the example code I have now. 

once that's in the code you could calibrate it against a dedicated spectrometer in a lab

May I suggest looking into megapixel smartphone camera modules in conjunction with a small piece of 3d printed mechanics? diffraction grating foil [1] can be used to obtain "sparse" tomographic data from which hyperspectral images can be recovered. [2]

Usually a simple slit would do but there is also the possibility to swap the slit for a coded aperture [3], improving sensitivity while preserving a well-defined dataset for spectrum reconstruction.

How about sub-nm for sub-$10?

-----------------------------

[1] e.g. "Diffraction Grating sheet 1,000 lines/mm" (ebay, amazon etc.)

[2] http://www.cg.tuwien.ac.at/research/publications/2012/Habel_2012_PSP/Habel_2012_PSP-Draft.pdf

[3] http://en.wikipedia.org/wiki/Coded_aperture

I tired this already, with ok results.

 http://www.pureengineering.com/blog/foldablemini-spectrometer

I used a smartphone camera.  and a diffraction grating from a DVD. 

I think its great for understanding how it works and teaching people about it, but it doesn't do the same as the hamamatsu part.  

I want to buy a different variation of this spectrometer. Can you point me to a source for buying these? In specific I'm after the IR version, Hamamatsu C11708MA

It appears that you can assemble your experiments, more or less, with COTS equipment. We had to design and build everything from scratch. The physicists did the theoretical part and developed requirements and we engineers & technicians designed and constructed the equipment. Of course, my experience was 20 years ago, so availability of equipment has improved. 

I am not a physicist, but they are in very different frequency regimes. The NQR is about quadrupole resonance with the extra-nuclear charges (electrical polarization of the molecules), and NMR excites the bipolar magnetic characteristics of the nuclei. Either one provides adequate signals only for certain molecules and or nuclei. These phenomena occur for certain compounds of nitrogen (among other elements) which, by fortunate coincidence are present in both high explosives and hard drugs. The NMR requires a strong magnetic field, so in my mechanical engineering capacity, I was engaged in building large, strong, magnets (none was superconducting). This was low field NMR. Both require RF excitation, and an RF engineer developed the excitation and reception means. 

I do not know how these phenomena mesh with Raman spectroscopy. It would be a very interesting exercise to take your equipment to a NMR/NQR lab to see if there is any sensible correlation.  You would be looking to correlate some characteristic of surface absorption via light vs. intrinsic magnetic and electronic properties. All are electro-magnetic properties, but lie in vastly different regimes. 

Pure Engineering, 

I contributed to the design of a NMR/NQR spectrometer about 20 years ago. An A/D ISA board and a Pulse Programming ISA board (two of three required) and the packaging for a FedEx-sized scanner for a portable Contraband Detector (of hard drugs and high explosives). Two US patents. We won an R&D 100 Award in 1995 for it. Glad to help if needed. 

Charles Moeller

cmoel888@aol.com

That is fairly sweet, It's amazing how small things are getting. I'm not sure if this sensor will help with nuclear magnetic/quadruple resonance, but let me know if there is a way. I wonder what wavelengths of light are best for drug and explosive detection. 

Where can you get these little modules? Do I even want to know how much Hamamatsu wants for one?

https://groupgets.com/campaigns/61-hamamatsu-c12666ma-mems-u-spectrometer-batch-2

via #Open Source Science Tricorder

(Oh look, there's a GroupGets logo on the PCB…)

For that price there must be magic inside that little metal case ;-)

I think the price is pretty decent considering what you get. It does not have enough resolution for what I want to do but it is still pretty neat.

The spectral resolution is about 10-15nm in practice, but it's slit limited, and if you back out the point spread function of the slit it might be possible to get it closer to the ~2nm/256 channels that the sensor has. The output with 5V on VIN tends to only be around 0-3v, and with a 10-bit Arduino ADC (-2 bits from noise), that leaves only around 7-bits (128 values) of dynamic range, so likely not really enough to do this well. It really needs to be coupled with an external ADC for useful work, and with the AD7940 14-bit ADC used on the Arducorder (or any similar part), you get a much greater dynamic range, and can oversample to 15 or 16-bits without too much trouble. Absolutely needs an external opamp off the output line, both for SNR (as per the datasheet), and to prevent damage in an Arduino environment.

I'm excited to see what folks get up to with these! They're beautiful little instruments.

I need sub-nm resolution for what I want to do (make optical band-pass filters)

The magic is basically a slit, grating and a line sensor.

http://www.hamamatsu.com/jp/en/C12666MA.html

My guess why the costs are high is because the volume is so low. Once a company like canon sticks it in all their printers and cameras this would drive the price down to under $10. but until something like that happens we are stuck with the high prices.

They are still arguably difficult to manufacture, even on a slightly larger scale. Look at fiber optic amplifiers and multiplexers, essentially the same thing, but range into the several thousands.