Belt noise - Help on belt alignment and idler mounting

I built a self-designed core-xy printer inspired by the Hypercube. The prints come out fine but when I move the extruder I get a rattling sound from the pulleys, mostly from the idlers in the back.

For the pulley mounting I used m3 washers which I sanded down to the diameter of the inner bearing ring of the pulleys. I put these on the above and below each pulley on the screw providing equal distance to the printed part. I hoped to get a clean setup this way by fixating the inner bearing ring without touching anything else. However the movement of the pulleys is not perfectly smooth, tilting slightly when rotating. I suppose the relatively large inner diameter of the m3 washer and the not perfectly round shape after sanding still leave some room for misalignment here.
Is there any better solution for that? Just don't tighten down the screw and abuse it as a bolt?

Apart from that I think there's still an issue with the general alignment (z-height, idler rotation axis being tilted by belt tension or so). I tried to calibrate it using calipers measuring the belt distance from the top of the frame but it's pretty cumbersome. Are there any tips or strategies on getting the alignment right?

Thanks in advance for your feedback.

ps: On the motor side I could fix the problem by fixating the motor vertical axis against being tilted by the belt tension.
pps: This thread seems to describe a similar issue. However it is unclear to me in what way the pulleys were "tapered" to fix the problem.

The pulleys have two bearings each so should not "tilt" as they turn. If they are tilting, it is likely that the shaft is too small in diameter. You should probably use either a 3mm steel rod, cut to length, or get shoulder screws that have 3mm diameter shoulder. You will find they fit tightly in the pulleys and do not allow them to tilt.

As far as alignment is concerned, that is best addressed by designing the parts to put the pulleys at equal heights and spacing everywhere. If the vertical position is a function of where you position the parts along vertical t-slot members, you can make a spacer from wood or metal and use it to set the positions of the pulley mounts using the bottom or top of the printer's frame as a reference.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
The pulleys have two bearings each so should not "tilt" as they turn. If they are tilting, it is likely that the shaft is too small in diameter. You should probably use either a 3mm steel rod, cut to length, or get shoulder screws that have 3mm diameter shoulder. You will find they fit tightly in the pulleys and do not allow them to tilt.

Thank you for the feedback. I ordered some shoulder screws, unfortunately the real fitting screws (ISO 7379) are apparently only available starting from 4mm. Got some 3mm steel rods which should have much tighter tolerances to compare too.
How do you confine the idlers in the axial direction? I find very hard to find something that touches only the small inner bearing ring.

Quote

As far as alignment is concerned, that is best addressed by designing the parts to put the pulleys at equal heights and spacing everywhere. If the vertical position is a function of where you position the parts along vertical t-slot members, you can make a spacer from wood or metal and use it to set the positions of the pulley mounts using the bottom or top of the printer's frame as a reference.

I will try using temporary spacer for positioning in the T-slot - thanks for the idea!

Those little pulleys can be difficult to use properly, and I've always had doubts about their operating life in a 3D printer. I prefer to use stacked bearings to make pulleys from larger ball bearings that are likely to last a lot longer and are easier to mount, both because of the larger size. They are heavier, so moving mass increases. In my 3D printer I used stacked F608 bearings for the pulleys. In my sand tables I used F625 bearings.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Interesting idea. The belts just run on the flat outer surface then? Do you see any negative effects compared to using idlers with teeth (where applicable)?

I received shoulder screws and 3mm steel rods. Unfortunately my Powge idlers have something around 2.95mm inner diameter. Using standard bearings would be nice since they are available with documented tolerance classes to get a real proper fit.

This is how i fixed my problem. These pulleys suck. The bearings are horrible. Might try and redesign my machines to use 5mm bearings somehow.

Thanks a lot for sharing! Looks like a good additional idea after aligning everything as good as possible.

I let the belt teeth run on the smooth surfaces of the stacked F608 bearing pulleys without any visible artifacts in the print surfaces. There's a Gates technical paper somewhere that says if you're going to use belt teeth against smooth pulleys there should be at least 9 teeth in contact with the pulley at all times. In a corexy mechanism, belts bend 90 degrees around the pulleys (except the drive pulleys where they bend 180 degrees). If there are 9 teeth in a 1/4 of a turn, there would be 36 teeth in a full turn. With 2 mm pitch belt, that means circumference would be approximately 36 x 2 mm = 72 mm. 72/pi = 22.9. F608 bearings are 22 mm in diameter. The stacked F608 bearings are a bit heavy, and they need an 8mm shaft that's also heavy. Stacked F608s will accommodate 10 mm wide belt.

F625 bearings have a 5 mm bore and smaller diameter. They are much lighter weight. I put a twist in the belts in my sand table so that the smooth back side of the belt rides against the pulley. The sand table doesn't require it for "print quality" but to keep noise down when I run the thing at 1500 mm/sec.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]