Re: B&Wtech 473nm unit / Spectrometer Mods & Info
Hello, I am new here on the forum mainly because of these spectrometers and the 473 lasers delivered with them (interested in raman-spectroscopy). I received 5 of these spectrometers from the ebay seller mentioned above. One had a damaged focusing mirror and the pins of the ventilator swapped, rest seem to be OK. Also 4 out of 5 lasers work OK, the 5th losing power after a short while (thermal issue ?). I use these spectrometers with the Spectrum Studio software mentioned above without problems. Perfect for calibration (it has the coefficient fitting included). Did not manage yet to have the coefficients stored in the spectrometer internal memory though.
There are a few peculiarities with the spectrometer however, which can be expected considering their original use. Important and related to the last posts is the fact that there is a cut-off filter at the entrance (and of course the baffles blocking most of the detector, which have to be removed). It blocks most of the light below approx 480 nm (few nm variation in the ones I received). This is understandable considering they were used as a raman/fluorescence sensor, for which you want to block as much as possible the original laser beam. In addition, I have the impression (but might be wrong) that this filter is fluorescing when having 473 nm laser light at the input. Luckily, the filter can easily be removed: open the spectrometer, unscrew the tiny set screw at the spectrometer input (inside the spectrometer) and remove the small black cylinder from the input (again at the inside of the spectrometer). This opens up the spectrum below 480nm (I get typically a range 380nm to 640 nm with some variation from one spectrometer to the other). I have not yet seen higher order spots appearing, but this can be due to my limited testing with this configuration so far.
Spectrometer noise is for all the ones I received between 1300 and 2200 counts, but when acquiring first a baseline dark signal spectrum and substracting it from the other acquisitions (Spectrum Studio can do this on the fly), the remaining noise level is only a few 100 counts and can be further reduced by avering several spectra (also function in Spectrum Studio). On an total signal of more than 60000, this is perfectly fine for me. All of them have however a few "hot" pixels: pixels with either high dark signal (easily removed by background subtraction) or unstable gain (more difficult to remove).
Straylight internally in the spectrometer is an issue with a high intensity input signals. It causes all kinds of strange features in the spectrum.
The beam splitter delivered in the unit is interesting. It has an dichroic mirror inside, reflecting most light close to the 473nm laser wavelength but much less other (higher?) wavelengths. Again: designed for the original application. I could acquire some raman spectra (although still limited quality) of aceton, alcohol etc, just with the delivered laser, the beam splitter and the spectrometer.
Did not try yet to improve the alignment. Not sure if I will do either. Planning however to change the wavelength range of at least one by rotating the grating in order to have a wider spectral range by combining minimum 2 spectrometers.
In summary: I had quite some "fun" with them so far.