During some sanding on the arms to get rid of the friction I tested to pinch the arm half together and just with a little force the belt friction increased significant. The rubber in the belts act like break pads on a bike.
When all anchors are installed at the same height as the spoil board, the bending force on the arms creates this type of pinching between each other. This is particular true in the corners. First due to the angle of the belt, and second the amount of the belt rolled in on the spool.
If you lift the anchor up so that the belt is extended out in a perpendicular angle, Would that remove most of the bending force in the arms and thereby minimize the pinching?
I designed an anchor to print out but hasn’t got time to print them yet.
If you list the the anchor up so that the belt is extended out in a
perpendicular angle, Would that remove most of the bending force in the arms
and thereby minimize the pinchng
yes, it is possible, I think it¢s a good idea (the closer the belts are to flat,
the less any Z error makes)
you do have to tell the system that you have done this. the Z coordinates for
each anchor (aka z offsets) defined in maslow.yaml need to be updated
I have designed and used these anchors
there is a lot of force on these, so they need to be pretty beefy to hold up. I
run a 3/8 drill through them after they are printed and then get a 3/8" bolt
that is long enough that the anchor rides on the smooth part of the belt
(sometimes I bother cutting off the threaded part, sometimes I don¢t)
note that these have a configuration value for how far the bottom belt is from
the surface you are attaching the anchor to. you don¢t want the belts pulling
up, but it take into account the length of the bit and thickness of the material
you typically cut for the ideal height (I would consider doing calibration on
top of a 2nd spoilboard to let the anchors be a bit higher, but still not pull
up when doing the calibration, you can then use a negative spoilboard value to
tell the system about this after calibration
This is a good topic to have in public so others can learn, and as you post
pictures and work through qustions, it can clarify the process for others.
Good observation. I would imagine the sag in the belt (particularly for large frames) would affect this as well. Unfortunately that would be hard to correct, because the sag angle will be a function of the extension of the belt. This line of reasoning makes me wonder how a Maslow that used a rack and pinion system instead of belts might work. From each anchor would extend a rack that could reach almost to the opposite corner. The Maslow would then have a pinion gear on the sled for each rack. Wouldn’t be awkward at all…
Fun. It would probably work and without much software changes. I think you would need to have an L shape at the anchor end so that the effective movement was directly towards and away from the anchor. encoders and everything else would still work the same…Just really slow. There would be some pretty big forces as one tried to accelerate long levers sideways. I suppose it might work with a really big frame and a small working area. : ) I don’ think the tradeoffs would be worth it.
I use 1 metre aluminium square tubing to extend the anchor points from the wall mounted anchor point. Each is at the same relative height to the arms. They are nearly flat due to the connection at the wall. Has the advantages of being rigid, so belt expansion is reduced, less belt needs to be rolled out, so it’s quicker to connect, I am tempted to go to 1.5 metres, which would put the extension up against the corner of the work surface. I originally was using a vertical setup but prefer the horizontal. On a vertical setup it would reduce the need to climb a ladder to connect belts.
Raising the anchors has made a difference, together with sanding the spool bearing surface. I can now run a retraction force of 600 reliably.
