For my final project, I would like to make a ruling engine. This is a precision machine that rules (or engraves) fine lines into a surface and is often used to create diffraction gratings. I am starting by reading through some papers and looking at similar projects others have worked on.
I met with Quentin to talk through my plans for the project, and it appears that it the ruling machine would have the best chance at working if I were to decouple the diamond scribe's motion from the sample's. I can use a diamond scribe on a wheel plus a way of moving the sample perpendicularly to the grating lines, perhaps using a single-axis flexure (maybe with a piezo stage). Zach's machine building class project is a good source for inspiration here. What is important here is to have a closed feedback loop for positioning so that there is high repeatability, even if small adjustments need to be made after the initial motion, and using a flexible mechanism will allow for much greater precision than a leadscrew. I aim to make very small but functional grating. In case you're curious, I will add my messier pre-Quentin-meeting documentation in Ruling Machine Planning. Here is some concept art I made using whisk.fx.
Achieving precision motion is one of the most interesting parts of the ruling engine to me. Therefore, I will aim to complete a precision xy motion stage first, and if I can do this quickly, will then try to figure out how to add the ruling capability as a stretch goal. Achieving the precision needed to make a diffraction grating with lines consistently spaced just a few micron apart is very challenging, and I've done a lot of brainstorming on that problem.
I have begun ordering parts for this and planning it out more concretely, keeping a BOM internally to keep track (this will be posted once its complete). Today (11/26) I am working on designing the motor control board. I am referencing Anderson Zelarayan's Output Devices Week to get an idea of how to use the DRV8428PWPR motor driver to drive a NEMA17 stepper motor, which alerted me to an example in the datasheet that enables 1/8 microstepping. I'll start by trying to replicate this. It seems to be possible using the Carvera but I may need to use the laser.