Public Lab Research note


Calibration Pictures

by ckmurt | | 3,353 views | 14 comments |

Read more: publiclab.org/n/5244


My calibration did not produce sharp lines and I was told the slit may be to wide. These are the pictures of the unit. Any help would be much appreciated. Link to the calibration: https://spectralworkbench.org/analyze/spectrum/1997


14 Comments

Hi! I didn't realize it was a kit -- i thought you'd built it yourself from scratch. In this case the slit is probably correct since it's been laser cut as part of the kit. Your assembly looks great! Was the spectrum of a compact fluorescent lamp? Can you upload just a photo you take through it while pointed at such a lamp?

Awesome to see one of these in the wild!

Is this a question? Click here to post it to the Questions page.


Did you do the infrared conversion?

In any case, I bet your camera is out of focus. You do need to focus it closer, whether or not you do the conversion -- it needs to be focused on the slit itself. A good way to focus it is to point it at a barcode about 4-5" away and turn the lens. Heating it with a hair dryer can loosen up the threads.

Check out step #10 on this page: http://publiclaboratory.org/wiki/dsk

Is this a question? Click here to post it to the Questions page.


I did the IR conversion, I tried to focus it on a barcode 3" away before I assembled the kit. I will try to redo it. Will upload the pic of the lamp image in a moment. and yes, it was a CFL


If you have a chance to upload a photo of the barcode while it's out of the case, that can help us narrow down what's going on.

Once we solve this I'd like to post it as a "common troubleshooting issue" on the wiki... would that be OK? Your photo illustrating the problem is excellent and we can likely show a before/after pair.

Is this a question? Click here to post it to the Questions page.


I posted a photo of the barcode at 4". I have other distances at well, its focused from approx 3-5 in. You're welcome to use this on the wiki.


Well, it's definitely in focus now, if that's 4". How are your spectra looking now?

Is this a question? Click here to post it to the Questions page.


Hi, ckmurt - i double checked by going through the build process from scratch again, and you're right -- 3-5" is too close. For whatever reason it is much more successful when focused approx 9" away from the lens. I'm not sure why this is, but try refocusing one more time (sorry for the trouble!) at 9" and tell me how it goes! See these images below:

Default focus: default

Focused at 4" away (resulted in blurry spectrum as above in ckmurt's data): 4"

Focused at 9" away: Focus at 9"

Spectrum at 9": Focus at 9"


Nice detective work, I'll give that a try today and see how it goes.


So i refocused it, this was the most recent calibration, the red and green channels are producing very clear and crisp peaks and valleys, however, the blue channel is still very fuzzy. Is this just unavoidable because of the camera? most recent calibration: https://spectralworkbench.org/analyze/spectrum/2136

Thanks, definitely an improvement

Is this a question? Click here to post it to the Questions page.


The blue is not so bad... i think that blue peak may just be broader, see these examples:

Is your bulb a "warm" fluorescent? There are more peaks and more detail since they coat those bulbs with europium and terbium.

Is this a question? Click here to post it to the Questions page.


I've been tweaking the setup today, its much crisper. I think some of the noise on the blue spectrum may have been from the blue light on the webcam, I taped it over and re positioned it a little bit, the blue end of the spectrum is clearer now. most recent: https://spectralworkbench.org/analyze/spectrum/2148


also it seems that there might be some problems with the calibration program. It often gives odd numbers for the wavelengths, at least for me.


one final note, my camera is facing inward in the box which seems to give it better geometry with the slit on mine.


How's the band separation on this spectrum for a CFL? (https://spectralworkbench.org/analyze/spectrum/20875)

I've adjusted the focus of the webcam to 9". The calibration example in the workbench looks fabulous - how can one come closer to achieving that with the desktop spectrometer kit?

Is this a question? Click here to post it to the Questions page.


You must be logged in to comment.

Public Lab is open for anyone and will always be free. By signing up you'll joined a diverse group of community researchers and tapped into a lot of grassroots expertise.

Sign up