sadam wrote:
The “belt fighting,” waviness, oscillation still occurs when performing a test cut dry with the router off and the dust collector hose disconnected.
Would replacing the control board with one from the store potentially correct the issue?
probably not, can you reduce your tension when doing the calibration and still
get a good report.
David Lang
I can try finding the minimum, but if I don’t have the retraction (2100) and calibration (2100) forces set pretty high I don’t get reliable retraction, especially of br. I’ve found that without reliable retraction I can’t get convergence in finding the corners.
2100 is very high, It might be time to try and reduce friction in the arms. Do all arms require 2100 to retract or only some? I have found sanding the bearing surfaces allowed me to reduce force considerably
sadam wrote:
I can try finding the minimum, but if I don’t have the retraction (2100) and
calibration (2100) forces set pretty high I don’t get reliable retraction,
especially of br. I’ve found that without reliable retraction I can’t get
convergence in finding the corners.
yes, you can’t reduce the limits below what’s reliable for your arms. If you
have one arm that takes more current to be reliable than the others, it may be
worth taking that one arm apart to improve it.
David Lang
sadam wrote:
I can try finding the minimum, but if I don’t have the retraction (2100) and
calibration (2100) forces set pretty high I don’t get reliable retraction,
especially of br. I’ve found that without reliable retraction I can’t get
convergence in finding the corners.
how do the measured anchor locations compare to the automatically detected
locations?
David Lang
Before performing the manual anchor measurement, I found advice elsewhere that talked about sanding the bearing surfaces of the arms and reels. I disassembled each of the arms and sanded the three surfaces with 220 grit sandpaper. After cleaning with a damp rag I followed other advice to lubricate the surfaces with machine/paste wax. I replaced my broken encoder rollers during reassembly and shortened the belt tails that likely broke them.
My retract and calibration forces are now much lower, 1100/950. The “gentle tug” to start extension is actually gentle. Calibration over 1000x500 7x7 ran cleanly and quickly. Dry cuts using the code where the oscillating motion occurred were fine and smooth. I’ve been able to successfully run the program while cutting. It feels good to be making things instead of fixing things.
Moral of the story: Spools should turn freely kids. Sand the contacting surfaces if it feels like a press fit.