Thinking about different types of extruder designs, and especially remote direct drive, I wondered if anyone's ever made a cartesian bed slinger printer (Ender 3 style) where a direct drive extruder on top of the X carriage is driven by a stationary E motor at the end of the gantry via a belt. This would of course turn the extruder hob whenever there's motion along the X axis, so the X and E motors would have to be synchronized, with effective extrusion being the difference of the X and E motor motions (thus my allusion to CoreXY in the name). Are there reasons not to do this, aside from needing an over-engineered belt assembly with closed loop belt? It seems like you could get whatever reduction you need just with an appropriate belt pulley, and no need for a gearbox on top of the hotend to get the necessary torque to the extruder hob.

Of course as described this only makes sense for bed slinger printers, since if you have to move the gantry in the Y direction, you have the motor mass moving just like a normal direct drive. Of course it might be extensible to "CoreXYE" with an appropriate belt layout.

  • 2
    $\begingroup$ Prior to our makerspace shutdown, a student mechanical engineer and I had a discussion about this exact topic. He had seen a paper espousing the concept. Unfortunately, it's been more than a year and for my brain, it's more like dog years. I found the concept interesting, but neglected to pursue it. Suffice it to say that it's likely someone has given the idea some consideration. $\endgroup$ – fred_dot_u May 7 at 0:24

Wow, Kevin O'Connor (of Klipper) just pointed me to a real CoreXYE design by Vlad Strateniuk called REXY, with video available:

There's an open Klipper issue about supporting it and apparently Marlin was already adapted to run on it.

  • $\begingroup$ Very interesting! Doesn't such a system have a limit on extrusion speed? I guess the extruder motor compensates speed of the carriage, so, sometimes it has to run faster than a normal extruder would do. Maybe you could dwell upon such issues in the answer? $\endgroup$ – 0scar May 7 at 6:08
  • $\begingroup$ @0scar: Possibly. I think you need reduction gearing of some sort because the belts can't handle the torque needed at the extruder hob, and reduction puts a limit on extruder motion, but that mostly affects retraction speed, not extrusion speed, and shouldn't matter; even with 10:1 reduction you're still fine, and 3:1 or 5:1 is probably enough. $\endgroup$ – R.. GitHub STOP HELPING ICE May 7 at 13:35
  • $\begingroup$ There is the matter of needing the sum of speeds, but I think if you have 2 E motors like in the REXY, the one going the opposite direction of the XY belt motion (needing added speed) can run at very low current and the one running in the same direction (needing difference of speeds) can run at full current. But that would require dynamically controlling the current, and I'm not sure how fast drivers can switch it. The total speeds probably aren't enough to need this, anyway, though, unless you're trying to do a 3-minute benchy... $\endgroup$ – R.. GitHub STOP HELPING ICE May 7 at 13:36
  • $\begingroup$ I take it that everyone is forgetting that you need a seamless E belt for this to work at all. The person who posted that video said that they are using normal glue to hold the belt together, but that they are looking for one that is seemless. $\endgroup$ – user77232 May 7 at 14:03
  • $\begingroup$ there are tons of deign issues, but the result is a super light direct drive, so those can be seen as tradeoffs. $\endgroup$ – Trish May 7 at 14:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.