From the video, it looks like the warping (curvature) of the web might be one of the biggest sources of variation. I assume it's due to a combination of uneven heat application and latent stress in the wood being relieved.
Possibly consider etching lines into the non-reflective side of glass mirror for linear stability? Which will still only be as good as your laser units/step. Secondarily the finer the lines, the more important a vernier graticule or mask in front of the receiving quadrature sensors as to block stray detection from the side.
With multiple sensors and being able to laser cut arbitrary slots it would be interesting to see if you could make this into an absolute position encoder. Not sure what to use it for, but it would be cool.
Think expansion and contraction of the wood will matter? I've wondered if a (stationary) threaded rod and an inductive sensor could be used for positioning. Should be very repeatable, so the errors in the rod could be mapped for better accuracy, if desired.
@maximthemagnificent I definitely think that's an interesting idea! If I had an inductive sensor handy I'd absolutely set that one up and give it a go. I'll have to add that to the list. If you give it a try before I do, I'd love to hear how it goes.
@adamsiwek7995 I definitely think some 3d printed version comparisons are in order. Although may need to swap out my 0.8mm nozzles if I want to make it fair :) I'd love to try some resin ones, but I haven't found any resins yet that are opaque to the IR light used by the optos.
@GermanMythbuster haha, very true! Definitely an option, but one thing I don't love about the film encoders is that they aren't rigid. Plenty of options for tackling that, but hoping for a non-compliant middle ground :)
@@bubsbuilds Maybe just glue the encoder film to a acrylic strip to make them non-compliant. I use relatively cheap uv epoxy for binding acrylic sheets optically clear, I bet that works for encoder strips too. But I don't know if you limit your resolution due to different diffraction indices of the Encoder Film (most likely PET) and carrier. Possibly a better choice is to also use a PET as a carrier, although acrylic is stiffer.
Definition of insanity, not spending cca 50$ on aliexpress for 5 micron glass linear encoder + rpi pico and call it a day (60$ in total). But rather spend > triple on wood, electric components, time, electricity and laser machine wear. 🤯🤯🤯🤯
From the video, it looks like the warping (curvature) of the web might be one of the biggest sources of variation. I assume it's due to a combination of uneven heat application and latent stress in the wood being relieved.
Possibly consider etching lines into the non-reflective side of glass mirror for linear stability? Which will still only be as good as your laser units/step. Secondarily the finer the lines, the more important a vernier graticule or mask in front of the receiving quadrature sensors as to block stray detection from the side.
With multiple sensors and being able to laser cut arbitrary slots it would be interesting to see if you could make this into an absolute position encoder. Not sure what to use it for, but it would be cool.
are you going to put two sensors 90 degrees out of phase and interpolate to multiply and get higher resolution in the end?
Think expansion and contraction of the wood will matter? I've wondered if a (stationary) threaded rod and an inductive sensor could be used for positioning. Should be very repeatable, so the errors in the rod could be mapped for better accuracy, if desired.
@maximthemagnificent I definitely think that's an interesting idea! If I had an inductive sensor handy I'd absolutely set that one up and give it a go. I'll have to add that to the list. If you give it a try before I do, I'd love to hear how it goes.
I'm 30 years and counting on that idea, so I suspect you'll beat me to it.
@maximthemagnificent haha, I have some of those on the mental shelf myself! The snail's pace race is on :)
How about comparing that to the 3D printed ver?
@adamsiwek7995 I definitely think some 3d printed version comparisons are in order. Although may need to swap out my 0.8mm nozzles if I want to make it fair :) I'd love to try some resin ones, but I haven't found any resins yet that are opaque to the IR light used by the optos.
@@bubsbuilds MAybe spray paint them?
or just buy 180LPI or 360LPI Encoder Film for like 10$ for 5m.
Solves the problem cheaper than you can cut them :)
Isn't the cost in the readers?
@GermanMythbuster haha, very true! Definitely an option, but one thing I don't love about the film encoders is that they aren't rigid. Plenty of options for tackling that, but hoping for a non-compliant middle ground :)
@@bubsbuilds Maybe just glue the encoder film to a acrylic strip to make them non-compliant. I use relatively cheap uv epoxy for binding acrylic sheets optically clear, I bet that works for encoder strips too. But I don't know if you limit your resolution due to different diffraction indices of the Encoder Film (most likely PET) and carrier. Possibly a better choice is to also use a PET as a carrier, although acrylic is stiffer.
Definition of insanity, not spending cca 50$ on aliexpress for 5 micron glass linear encoder + rpi pico and call it a day (60$ in total). But rather spend > triple on wood, electric components, time, electricity and laser machine wear. 🤯🤯🤯🤯