Public Lab Research note

Oil testing kit: Assembly + Initial Calibration

by ethanbass | October 23, 2015 21:34 23 Oct 21:34 | #12331 | #12331

Hi, I'm Ethan, a member of the Oil Testing Kit beta program, located in Chicago. Here I've collected some notes and pictures from the assembly and calibration of my oil testing kit.


1.jpg 2.jpg Unloading the oil testing kit

3.jpg Oil testing kit contents: oil samples, cuvettes, eyedroppers, pipettes, etc.

4.jpg Contents of the spectrometer v3.0 kit (my card-stock was a little warped coming out of the box but it wasn't too problematic)


Preparation of the DVD diffraction grating:

I cut the DVD that came included with the kit into quarters, but unfortunately none of my fragments split apart cleanly.

9.jpg Instead, almost all of the aluminum stuck to the bottom of the disk, so these pieces were useless as a diffraction grating.

lye_bath.jpg I am experimenting with MrBumper's method of loosening the aluminum layer in a lye bath.

10.jpg In the meantime, I found an old, scratched-up DVD that split apart cleanly

Spectrometer housing:

  • The assembly of the spectrometer's housing went fairly smoothly after I consulted the materials on the website, but the reproductions in the booklet were very hard to make out.
  • In spite of the warning included in the instructional booklet, I initially folded the top of the box with the inside facing out, so that the cable holes were misaligned.
  • This hazard could be avoided if the top of the box were printed with a pattern to differentiate between the inner and outer surfaces.

spec3.jpg Desktop Spectrometer v3.0

Velcro bench, block and camera:

5.jpg My assembled camera block

7.jpg 8.jpg Assembly of the diffraction grating angle (I opted for extra tape)

  • The velcro bench system is clever because it's sturdy enough and you can adjust it easily, or add new parts to the spectrometer, if necessary.
  • I had a lot of trouble getting the camera to stay mounted. Even after doubling up the tape on the back of the camera, I found that torque from the camera's USB cable is strong enough to easily dislodge the PCB, if the spectrometer is not handled extremely carefully.
  • Has anyone used the mounting holes in the PCB to screw down the camera? I think this would help a lot to stabilize the camera and improve replicability of the results.
  • Why isn't the face of the camera block cut so that it supports the entire length of the PCB? Are the mounting blocks cut at Public Lab or are they a pre-fabricated part? I couldn't help thinking that having full contact over the entire length of the camera would provide a lot more stability, even using just the sticky tape. It would also open up the possibility of using the PCB's 3rd mounting hole, next to the micro-USB port.
  • Another thought I had was that it might be possible to reduce torque on the camera by fastening the USB cable to the back of the mounting block.

Cuvette frame assembly:

11.jpg When I laid out all of the pieces, I didn't know what the function of each part was meant to be or how they should fit together

  • It would have been helpful to have a diagram with a short explanation of each piece, so that it would be obvious which piece is the cuvette holder, which piece is the laser attenuator and so on. Or the parts could have labels printed on them.
  • It was difficult to interpret the documentation because the edges of the black card-stock are very difficult to make out in the pictures (even more so in the reproductions included with the kit).
  • It would be useful to have diagrams with arrows illustrating the directionality of the folds, and/or a video.

12.jpg 13.jpg This was probably the easiest fold, but it took me a few minutes to figure it out, even though it's obvious once you see how it's meant to go!

14.jpg The proper method of folding the cuvette platform took me an extremely long time to figure out. My paper-folding skills are definitely not up to snuff, but better documentation would help.

  • I had to notch the smaller triangular holes with a razor blade to get the laser holder to push through the middle of the “M-folded sheet.” I spent a while trying to force the "V-folded sheet" through the holes before resorting to the razor blade, resulting in some slight crumpling along the edges of the laser holder. I see that some other beta testers had the same idea.
  • To my mind the smaller triangles also do not allow enough clearance for the laser to slide in and out of position easily enough.
  • I also found it useful to notch the edges of the laser attenuator slit so that the laser attenuator could slide in and out more easily.

16.jpg Modification of the laser attenuator slit with a razor blade

  • One small note on the laser attenuator: it is difficult to determine which opacity you have it set to, since you cannot easily see into the aperture when the cuvette holder is assembled. This could be a problem if you are trying to compare methods with other scientists, or even just to achieve consistency within a single testing session.

15.jpg Laser attenuator

done.jpg My fully assembled kit


  • I found that pointing the spectrometer directly at a CFL bulb, even from a distance of several feet away, always resulted in overexposure of my spectra.
  • I had success using light reflected from a piece of plain printer paper, as suggested in the calibration notes.
  • I noticed that, for some reason, reflected light did not activate the “CFL detection” icon in the spectral workbench interface, even though the peaks appear to be much the same.
  • In my first attempts to calibrate the spectrometer, the instability of the camera was a major stumbling block. After noticing that the calibration was shifting (sometimes dramatically) between consecutive spectra of the same CFL bulb, I opened up the spectrometer's housing and found that the camera was no longer firmly mounted to its foot. After re-mounting the camera, I was able to get much better replicability over several spectra!

Screen_Shot_2015-10-22_at_7.59.48_PM.png Replication of CFL spectra when camera is securely fastened

Screen_Shot_2015-10-22_at_8.22.58_PM.png Comparison of overexposed CFL spectra with proper exposure from reflected light

  • I think the calibration procedure on Spectral Workbench should be modified to specify which green peak to use for calibration, since there are two peaks roughly corresponding to the “mercury 3 line,” falling 4-5 nm apart from each other, as discussed in a prior research note.

Thanks! I look forward to working with all of you on this exciting project.


Hi Ethan,

Your CFL spectra look really good. It's not that common to resolve both of those green lines with a PL spectrometer, so you must have done a good job assembling yours. That's a good point that everyone should be aware that the big green peak is really two peaks. I think the new calibration procedure at Spectral Workbench 2.0 mostly solves this problem by displaying an image of a spectrum with both peaks as you calibrate. I think version 2.0 actually works and saves your results, so you should be able to use it now. Jeff should confirm that though.


Reply to this comment...

Cool. I can't figure out how to use version 2 to capture spectra. Maybe i am not looking in the right place, or it is not up yet. So far, I've only figured out how to use version 2 to look at and compare existing spectra.

Reply to this comment...

thanks for all the assembly notes. One note on the attenuator-- there are registration marks on the attenuator strip that correspond to the different fade levels-- they aren't marked because keeping them in registration required putting them on the cut layer.

Reply to this comment...

Ethan - version 2 doesn't include a revamp of the capture interface -- just analysis. We've gone a bit late on our schedule for rolling out analysis, so probably won't do too much to the capture interface for v2.0, but hopefully in an upcoming version.

Reply to this comment...

Great, thanks for the feedback -- I added some of your notes to @tonyc's note on a redesign of the camera mounting:

Also, SWB 2.0 doesn't include a rewrite of the capture interface (maybe 2.1? hopefully!) so it's only in the analysis pages.

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

Reply to this comment...

Login to comment.