Get full raman signal

[neutrino]

Hello,

first of all i want to thanks Luc for his great work and sharing all the information and experience.

Let me short introduce my self. I am from Germany and togehter with good friend we builded the open raman spectrometer.

We are using all the parts of the performance edtion but the following parts are different(at the moment):
- Slit 20um
- BFLY-PGE-12A2M-CS camera
- 10 mW laser diode from eBay
- selfmade base plate
- modified cuvette

Now we finished the assembly and get some signal. For us it seems like the raman signal but we are not sure.
Also we are not sure if the BFLY-PGE-12A2M-CS camera is sensitive enough to catch all of the raman signal.
We decided modify the cuvette. Here the AC127-019-A is external on a cage mount so we can use the lense easily for alignment.
We also tried to put the lenses on lens tube and SM05/SM1 cage mounts, because we had trouble to fix the AC127-019-A on CP14/M without pan/tilt.

Attached a photo of the setup(yes, it looks horrible because it is still a prototype), and two images of the signals of 2-propanol/IPA and 2-propanone/acetone.
It seems we can only see the three peaks around 3000 cm-1 of 2-propanone and one peak around 3000 cm-1 of 2-propanol
We played with exposure time and gain but this is all we get. If we switch to Mono 16 we see a black/dark image only

Thanks in advance for you help


Regards
Neutrino

[Luc]

Hi!

Thanks for sharing these modifications

From what I see in the propanol image, it does indeed look that you are getting a good Raman signal!

It would be best to have an intensity plot to confirm and analyze the results, in particular the resolution that your laser is offering (a good tip is to look at the narrow peak at 1000cm-1 or if you can resolve the 2 nearby peaks at 3000 in propanol). Did you try using the SpectrumAnalyzer program? I believe I validated the software with the 12A2M but my memory might be wrong. In any case, you can import tiff/bitmaps in matlab or octave for the analysis.

Generally speaking, I worked a lot in the past with 12A2M cameras. They're unfortunately very noisy compared to Sony Pregius sensors by a factor >2 in low light conditions (faint peaks). If you're on budget, you can have a look at Daheng cameras.

Apart from that congratulations for your setup because, from what I see up to now, it looks fully operational

[neutrino]

Hello,

thanks Luc for your reply and sorry for late response.
We have only limited time at the weekend to work on the spectrometer.

We followed your advise to use SpectrumAnalyzer and we get a good spectrum of i-propanol.
We could see much more peaks as on live camera image. It showed almost all of the peaks.
Also the shapes of the peaks where nearly the same as on reference spectra. But unfortunately we did not take a image of the same.
We tried to optimize the signal which went wrong and led to much more worse signal than before
Attached a image of the latest quality(of course, very bad) We did not calibrate software thoroughly which is cause of wrong scaling
About the camera we think we will buy the Flir 31S4M.

[Luc]

Although it's a bit noisy we clearly identify isopropanol on your spectrum

What integration time did you use? With long exposures you should get a very clean curve like on the website. If not, it means alignment can be improved.

I recommend the 31S4M indeed

[neutrino]

Hello,

again sorry for late response.
We ordered the 31S4M and waited until it arrived before we continued working on the spectrometer.

After setup the 31S4M we get much more better signals than 12A2M
Attached a image of i-propanol.
It seems we are on a good way now.
Thank you very much Luc for your support

[Luc]

That's good to know!

On the occasion, I will make a post about cameras performances & how to quantify them but I don't have a lot of cameras to test (only 3). I made a setup for cameras evaluation in the past so I already know exactly how to do it