Core XY belt paths

I'm building a core xy printer based on the fabotatum design (no crossing belts) as in the image below. It's coming along fine but I was reading some threads here about how to adjust the tightness of the belts. I think digital_dentist showed some images that got me thinking about something.

I've always been of the assumption that for this design all the belt paths must be either perpendicular or parallel to the x/y carriage. But after seeing some of the images of how people were suggesting doing the belt tightening, I'm not sure that's necessarily true.

So my question is which belt paths (if any) are required to be perpendicular/parallel to the carriage? If possible I would like to use the drive pulleys that carry both belts (upper ones in the image) and move them outword to tighten the belts, but would that cause any issues?

Checkout the excellent writeup that the_digital_dentist made about belt paths.

CoreXY Mechanism Layout and Belt Tensioning

-os3dp

In your case, you won't be able to use the dual idlers as tensioners. You'd have to use two stationary idlers and two moveable idlers.
A simple rule is: all belt segments that change length have to be 'true'.

Quote
os3dp
Checkout the excellent writeup that the_digital_dentist made about belt paths.

CoreXY Mechanism Layout and Belt Tensioning

-os3dp

Thanks for that link. That's a good writeup. @o_lampe, thanks for your response also. I think what I can do is make the motor mounts moveable in a forward/backward direction. That would allow me to adjust the tension without affecting parallelism.

The sad truth is, a lot of people lay out the belt paths incorrectly and then wonder why print dimensions are off sometimes but not others, or never realize that there's a problem and just accept that varying belt tension is a "feature" or the mechanism. It isn't. In a properly laid out system, the tension will not vary with extruder carriage position. In this diagram, segments labeled A-H must be parallel to the guide rails for proper operation of the mechanism.



The diagram above isn't the only possible layout for the mechanism. I saw one design in which the motors were placed in the M segments. That should work fine, too, but will require extra pulleys to ensure sufficient wrap around the drive pulleys. The diagram above is probably the optimal in terms of minimizing the number of pulleys. Every pulley you add increases moving mass and friction.

This diagram shows all the places you can adjust tension without screwing up the motion of the mechanism:



I recently built a large (about 1.8 x 1 m) corexy mechanism for a sand table that put the belts on the same level and twisted/crossed them in the M segments. I didn't leave it that way for long- the belts rubbing against each other at the twists/crossover were grinding rubber off the backs of the belts, and the belts kept climbing and riding on the pulley flanges. I converted it to stacked belts and all problems were solved.





Edited 1 time(s). Last edit at 10/13/2018 06:50AM by the_digital_dentist.

---

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

Hello guys,

I am currently thinking of a new CoreXY build (about 400x400mm base size) and therefore, I was thinking about the motion system and came to a new, slightly different belt path. What if two of the belt connections to the gantries are swapped (see figure)?

I already tried to figure out disadvantages vs. original CoreXY configuration and found some introduced deviatoric torque in the gantry when moving the whole x axis carrier, which scales with the belt distance on one one side of the gantry. But since the lever is really small (compared to the racking behavior in H-bots), it should not be that much of an issue?

What do you think of this idea?

Best
Andy

Hi Andy,

You can also move the idlers off the gantry closer to each other, so the belts to the carriage are one above the other and are centered in the middle off the gantry

Hi gforce1,

yes, this would be another special case, which I have already seen in the "Ratrig" 3D printer. These designs allow for closed belts which I would be interested in (less mass on the gantry since one instead of two belt clamping mechanisms are needed).

It looks like you've taken the twisting torque from the extruder carriage, converted it to two unequal torques, and moved it to the bearing/pulley blocks at the ends of the X axis. I've seen different belt arrangements that claim to eliminate this or that torque, but in the end does the torque they're trying to eliminate somehow affect the print? Unless the X axis is twisty and/or there is play in the extruder carriage bearings, neither of which is present is a reasonably built machine... meh. OTOH, autoleveling in printers can allow unflat, flexible beds to be used, so maybe playing with the belts attachments to the extruder carriage can allow twisty X axis and sloppy extruder carriage bearings to deliver quality prints. Try it and let us know!

Edited 1 time(s). Last edit at 01/06/2021 09:14AM by the_digital_dentist.

---

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