No Maslow yet but following the sled/ring modification threads with interest.
I woke up this morning wondering if instead of a ring with bearings on sleds rolling around it, we could just use two large internal diameter bearings e.g. these have 100 mm internal diameter and would allow the router to be mounted in the middle (as with the ring) and one chain would be attached (slightly offset to keep both chain attachment points the same Z distance) to each bearing.
I suspect it might be more expensive (large good bearings are) but perhaps mechanically simpler, and easier to get accurate / less parts than a ring and two sleds with 2 wheels/bearings each. Welding the chain attachment points straight to the bearings may not work (and less easy to swap bearings if ever needed) but for a test I assumed that pipe clips (screw tightened) would work.
I would be interested to hear what people think, and happy to draw a pic if helpful,
please see the top of the âthrowing my hat into the ringâ topic.
you are describing what we always thought would be needed to get triangular
kinematics, and finding bearings that would work, not clog up with sawdust
handle the angular load (as opposed to axial loads that most are designed for)
and not cost a fortune was a major problem.
Thatâs why the ring approach is so attractive. And the linkage kits are a futher
refinement that takes less material and space.
Thanks dlang and larry, I have finally caught up with the bearing constraints on that monster thread:
Axial loads - use angular contact bearing
Sawdust - use sealed bearing
Price - small bore bearing
20mm ID, sealed, angular bearings are fairly cheap - so what would happen if we ran just the router bit, but not the whole router (using a collet extension if necessary) through the middle, with the router mounted 4cm off the base with standoffs, which could be at 12 oâclock, 4 oâclock and 8 oâclock, to avoid the chain angles?
Does the effect on the sled balance rule this out (chain attachment point too far from sled/router centre of gravity)? i.e. the chains would be connected much closer to the base and the compensating options are not sufficient i.e. chain attachment height is not adjustable without large L shaped brackets (which I was ruling out due to the force they would put on the bearings), changing frame angle = too much more friction, using a spindle for less weight than a router means more complex Z-axis devicesâŚ
you need the chains to attach to the sled fairly near the balance point.
you have radial loads, not axial loads on the bearings.
so if you were to go with small bearings (and 20mm inner diameter is pretty
small in this context, you would need very large L brackets to clear the router
(4" min, 6-8" on some routers) to get out from under the router, then several
inches high to get to the balance point (which is further out since you are
moving the router away from the workpiece)
also remember, you have to have these clear of the sled so that wide, fancy
shaped bits can clear the workpiece.
at this point, you are going to be putting torque on the bearings, not just
radial loads
I think the result would end up every bit as complex as the linkage/ring
approaches
@dlang thanks for more input - adding (rolling) supports under the L brackets to mitigate the torque on the bearings would solve that but make it even more complicatedâŚand the extension collet bumps up the costs
Just make your sled so that is has more mass than the router !
(Iâm mostly kidding, however increasing sled mass will âlowerâ your CG (toward the work). Depending on your router a 10-15mm steel plate could do the trick⌠but that adds even more cost and complexity.)
!