So I’m working on designing a 3D printer (long story short, T-slot rails are expensive around here, so I’m designing something that uses square aluminium tube instead), but I’m having a hard time with deciding on the kinematics. Originally I was going to use coreXY, since everyone else seems to use it. But the way the belts cross each other is off-putting for me to account for (it makes me uncomfortable). So I started looking at H-bot. My question is, which would be better? Why are there more coreXY printer than H-bot now in days?
I should also mention, I do know about stacking the idlers in a coreXY to avoid having the belts cross each other, but if I can use H-bot and not have to do that, I will.
The belt path H-Bot uses generates toque, causing racking. By crossing the belts coreXY is able to achieve X and Y motion without generating any torque.
Maybe I can run this idea past you then: I was planning on having the Y axis be on rods with SCS8UU’s, and then the X axis on an mgn rail. Would a 3D printed part attaching the rail to the bearing still give that degree of playin this configuration, or do I have to use metal?
I don’t think your reasoning for choosing H-bot makes much sense. You can hide the crossing belts if you need to, unless you use plate construction like the Markforged printers you’ll generate torque.
I see what you mean. I guess I’ll have to go with coreXY then. My other reason for going with H-bot was that it it looks easy to add adjustable belt tension by making the outer idlers adjustable. With coreXY, it looks like the anchor points on the carriage are the only place to do it.
You don’t need to cross the belts if you place one above the other. For the belt tension the simplest option is to make the motors adjustable. Or let them swivel on one bolt, like you would do on a Prusa I3MK2.
@Rien_Stouten I knew about stacking the idlers, I just couldn’t find anything showing how it worked like that (until after I posted this). The belt tensioning idea I was not aware of, that’s a good idea. @Baldur_Norddahl thank you, that’s exactly what I needed (though, Alex’s video showed the issue much more clear).
Slotted motors are definitely the best way to tension CoreXY. It’s hard to do in a tidy way on the X carriage.
HBot is never better than CoreXY, under any circumstances. It’s just a bad gantry design. The mechanics are inherently unstable and your linear hardware has to resist belt forces in a way that is difficult for them to do. Basically the only way to make it really rigid is to make the Y carriage bearing spacing (along the Y axis) at least, say, 1/4th the gantry bridge span. That burns up a lot of useful Y axis travel.