Perhaps totally unnecessary I am trying a different route after being inspired by the Maslow mark II gantry type. I too noticed that if the sled wobbles a bit, the measurements of the part being cut are off. So goal number one is to let the routerbit always be perpendicular to the bed. Goal number two is getting some z-information, just like with a gantry-style router.
As you can see, I made the arm quite stiff, so goal 1 will be achieved. Goal 2 on the other hand…
My first impression while trying are that the bed and supportstructure are not stiff enough, especially around the place wher the arm pivots.
This is my solution for the chain attachment, two ballbearings (65x50x7mm) around a collar, under the router. When the router is in the lowest position the routercollet and the underside of the collar are flush, so no depth is lost.
I needed to do something like this, because the ringkit or the linkage kit do not have enough reach.
Here is a video:
I can’t test yet, because unfortunately the sliding bearings of the AEG router are not tight enough. They catch when I try to move the z motor (this was no problem with my former Festool router, with the same z method) so I will have to make something different.
Also, because the “sled” is lighter than the former sled, the chains hang much more. Might have to put some load on it.
So, still a lot to do, but thought I’d just share.