laser induced breakdown spectroscopy

[jonsmirl]

Could you use this rig to do a demo of LIBS like the instruments on the Mars rovers? Are there low cost lasers capable of forming the plasma? I'm not sure if you can do any useful identification without the ultraviolet spectrum. This demo would be done in the context of high school physics.

It does appear to be possible to do UV imaging using the PI Cam.
https://www.mdpi.com/1424-8220/16/10/1649/pdf

[leswright1977]

The software probably could (I have implemented peak hold for pulsed applications, just not pushed it to GitHub yet, still playing with it)

The problem is the UV. With the spectroscope used, there is a lot of glass (and probably plastic) between the input aperture, and the output. It is doubtful that measurable UV would make it through. You would have to build a spectroscope front-end with a reflective diffraction grating and Fused silica optics.

Then of course there is the camera, but the paper you attached has a solution for that! (thanks for that by the way, it looks really interesting. Might even be possible to get thrice the resolution out of it with a B&W image and some programming!)

Worth looking into if you are sol inclined!

[jonsmirl]

I do more work with the robots and the cameras. We are working with this camera currently:
https://www.cnx-software.com/2021/01/25/firefly-dual-lens-ai-camera-module-comes-with-rockchip-rv1109-or-rv1126-processor/
https://www.firefly.store/goods.php?id=133
Dual sensors -- one is 1080P visible light, the other is 1080P IR. Also a good AI unit that can run tensorflow. This unit can be flashed with software that turns it into a USB camera. I also have several 4K camera boards, but they are more difficult to work with. One has binocular 4K and can do real-time depth mapping.

Are there cost effective solid state lasers available that can create the plasma? Power should not be an issue, the Mars rover does three 10W 5ns pulses per second.

Here's someone who built a desktop LIBS rig.
https://www.youtube.com/watch?v=jHG8zm7kfOo&ab_channel=toc1955

[leswright1977]

V3 is out, for pulsed applications. Also supports CSV Export.

Re: Solid state lasers, not really. The cheapest solution would be to tear down and re-purpose a Tattoo laser head. I did this some years ago, and rebuilt it as the engineering was particularly poor. To break down a dielectric with a laser requires 10's of Megawatts of peak power, so no bare diode will come close. Maybe a fibre pumped Q-switched microlaser would do, see this article here: https://www.sciencedirect.com/science/article/pii/S0079672718300107

[leswright1977]

I have nuked 2 picams so far, however, I have learned some things about the stripping process (what not to do), and the results mid strip have been very promising.
I have got some more cameras on order...

[jonsmirl]

I have not tried this on the Rasp Pi, but most sensors have an enable pin. That pin allows you hook multiple sensors (like IR, visible, UV) to a single SOC and then cycle through the sensors capturing images. A common use of this is for facial recognition. You use two senors a couple of inches apart - one visible, one IR. Then combine the signals using binocular math to build a 3D map of the face. This can't be defeated with a photograph.

Another thought, for a spectrum you just need a one dimensional reading. Maybe you could take a single sensor and create three different masking bands in the sensor -- IR, visible, UV. Now a single lens setup could expose the sensor containing the three bands.