–replace all the non-vertical chain with heat-set pre-stretched spectra
(dyneema). I’d move the motors to the bottom corners, and run the chain purely
vertical, eliminating most chain sag, as well as “plucked string” movement
caused by sudden tension changes.
you don’t have enough vertical space to have enough chain, and your spectra will
still sag, so you haven’t eliminated the problem, just changed it.
we have chain sag pretty much taken care of in software, so there’s not much to
gain from this.
If you could eliminate the chains entirely and go to some other line on a
take-up reel (especially if you can measure the movement of the line instead of
the motor) you would have a win.
–grow the frame to improve line geometry to the whole working surface. I’m planning a 14’ top beam, 3’ above the ply top.
this is good
min force 7.64 pounds, max force 24.88 pounds (stock 3.26, 33.74)
you probably don’t need to go nearly as high 26" above would be 8.38, 33.09
–to make light 3d work possible without worrying about sled support, i’m
considering mounting my non-sled to a pair of roller-coaster track kind of
pipes, that in turn are mounted to a pivot at the center of the top beam. The
pipes/sled tracks would then ride at their bottom/outer edges on 1x3’s mounted
on edge around the work surface. In an ideal world, I’d learn the firmware
language and add polar or radial kinematics support for the firmware, and use
a single point attachment since the pivoting rails prevent any localized sled
rotation.
I’ll say what I’ve said many times before, if you want a rail system, go with a
corexy setup and you will be much better off. you would eliminate all the ugly
math and dependence on gravity as a force and have a fairly conventional CNC
machine.
sled rotation doesn’t matter, so don’t go to a lot of effort to eliminate it.
these rails will add a lot of extra weight to move, a lot of extra friction, and
a place for sawdust to accumulate and affect things.
