Ringing Part 8: Make it worse before making it better
I approach certainty.
Throught my illustrious scientific career, I have found that the most direct way to figure out what is wrong is sometimes to make a hypothesis about the cause, and then amplify that cause to see whether it’s actually involved in the system.
In the previous log entry I said that I would install PLA bushings and see what effect that had on ringing, either increasing it (if the bushings are too tight) or removing it (if the bushings run without friction). And that’s exactly what happened.
Here’s what happened when I installed PLA bushings with the bearing blocks fully tightened (i.e. probably misaligned and producing more friction than necessary):
And here’s what happened when I loosened those bearing blocks so that the bushings could self-align:
I did my best to angle that print into the light to show off any surface artefacts, but there’s really not much to show! There were some vertical lines but they’re not ringing, I think they’re caused by the bed being so loose on the bearing blocks that it was sliding around when the Y axis braked.
So I mean, this is good news! Getting to the heart of the matter.
So okay: the problem with using LM8UU bearings or bushings is that they’re really quite long and they don’t have much play inside them; like if I printed a 8.2 mm ID bearing and put it onto a 8 mm rod, I’d have to get that rod aligned to within 100 microns for it to still be okay.
The other ‘problem’ with these bearings/bushings is that so much of it is in contact with the rod at a time, and we all know that if you want to reduce friction with a surface, then one of the best ways to do that is to limit the contact area.
So the idea is to try a bushing that’s shaped like this:
The contact surface is short (10 mm), so it presents less friction to the rods. And it’s only in the center of the bushing, so there’s a lot more practical space in there (10.8 mm, so that’s 1.4 mm of play). The amount of play the rods will actually have is, of course, limited by the length of the contact area. One could even see a case where you just have a contact area of 5 mm in length to give the rods the smoothest ride possible.
The reality of a rod like this is that it would wear down pretty quickly, it might even become part of monthly maintenance to swap out the bushings to keep things nice. But if they fix my ringing problem and tide me over until those ‘Drylin bushings’ arrive, or if they prove more useful than those bushings entirely, then this is a pretty cool experiment.