right now kinematics.cpp is unreadable in terms of what it actually does. There are a lot of greek variable names, but no diagram or explination of what is what.
We have done a LOT of work to figure out the various sources of possible error and add calculations to the kinematics to allow us to account for them (even though we are still struggling to get a calibration routine that gets the right values). All the other v-plotter/bipod/calwhatever drawing robots just do the simple triangular calculations and either live with the errors, or don’t realize they are there.
a cleaned up and documented kinematics.cpp would be valuable to a lot of projects
Are you referring to both the triangular and quadrilateral kinematics? I find the triangular kinematics to be somewhat readable…(yes, doesn’t have diagrams and it’s not really spelled out, but it’s also fairly simple). The quadrilateral is definitely an issue.
I’m referring to the triangular (as far as I’m concerned, the quadrilateral is
legacy that would go away if we ever did incompatible hardware)
the general idea of the triangular is easy to understand, it’s the subtle things
like the compensation for the sprocket diameter and where the chain leaves the
sprocket to become straight, chain sag, chain stretch that are not clear.
I’m still working the hardware end of a Maslow Mark II, but since that one doesn’t have the sled rotation to deal with, I would think the quadrilateral might be the right way to go. Haven’t thought about it very carefully, though.
The original quadrilateral kinematics attempted to figure out how much the sled would rotate when the sled endpoints were fixed (didn’t swivel like in triangular kinematics). If your sled design doesn’t rotate, then its real simple trigonometry to figure out where the sled is.
