After years of great service, the last few days my delta printer seems to have had some X (or Y, it’s hard to tell) drift while it’s printing. I thought it might be a bug in Marlin 1.1.0RC (right cube) so I switched back to the 1.0 that has worked for years (left cube) and see the same thing. The other axis (Y, or maybe X) shows no drift - its sides are straight up and down.
Isn’t this sort of odd for a delta printer? Right now I’m considering:
My printer’s frame is our of whack (not likely - it’s built like a rock and wouldn’t explain the back and forth in just one axis)
Cooling issue? I don’t have a heated bed and my room’s about 65F/18.3C. But this also hasn’t been a problem in the past and for drift like that in just one direction…
I had similar issues with my delta and tracked it down to the rod end bearings. Check them at the effector and on the other side for any play / wobble.
I would expect the most likely cause of that type of issue on a Delta to be tension on the bowden tube and/or wiring harness. If your ball joints have loosened up, that could explain why you didn’t see the problem before.
@Whosa_whatsis no way, this is tipical drivetrain issue on one of the axis. Belt picked up some hard debris and did it’s thing with pulley and got damaged, either belt or pulley. Usually replacing belt fixes this.
Thanks guys. My belts are a little long-in-the-tooth - will replace them (and inspect gears) in a few days when I get back to this. Arms and bearings all seem to be in good shape.
I always see this when belt paths are not perfectly aligned and tension changes as the carriages go up/down. It’s always good to do a “6 tower” test – this same thing, taller, but one at each tower, and one at each intermediate point between the towers. The one with the most effect will give you a hint. If it’s an intermediate point between two towers, then it’s the tower opposite.
Yep, belt issues. I had a printed belt tensioner ziptied to the carriage for stability that had twisted and was (quietly enough that I didn’t notice) rubbing against the frame. Obvious in retrospect but I missed it initially. Put three new belts on and the 6 Tower Test printed perfectly.
