Hi guys, I want to post this to get some feedback on a X carriage I have previously designed to address the inconvenient ears that #Smartrapcore had and have re-designed now to allow symmetrical belt tension adjustment in addition to the previous iteration’s “ear-less” design.
I had to embed the tensioning mechanism in the X Carriage itself because I had no other way to do it with a simple method in my current #ScoutCoreXY #3dprinter design and you can’t get any simpler than a bolt tensioning for both belts and in a symmetrical fashion on top of that.
I haven’t seen yet a design that would address the belt tension in a CoreXY so that you end up with a balanced belt tension.
Below you can see the result, still in WIP but very promising unless I find a fatal flaw or someone else points me to one, for which reason I am showing it here.
Feel free to head over at thingiverse:
where the X Carriage design is hosted.
Also you can view it in 3D on Sketchfab @
CoreXY X-Carriage w/ Tensioner (exploded view) - 3D model by thorfab [f00b1df] - Sketchfab .
Wouldn’t a belt balancing mechanism need to tension each independently? I’m thinking a teetering mechanism that would apply tension to the opposing belt. It would need to lock in place after balancing.
An easier solution would be to use a gauge that can measure belt tension, then used to balance it manually.
If you install the belts (which should be the same length) with exactly the same teeth in the locking gripper on the carriage design that I propose, you have a balanced system, it just isn’t tensioned.
By having the slide gripper pulled towards the center of the carriage you are now applying the same amount of tension on both belts. It will never get out of balance because you only have one adjustment system.
Also if you have a locknut it will never loose tension due to vibrations. I don’t think it can get simpler than this. It’s so simple I am afraid it really has something terrible that I have omitted 
so why do you need it on both sides of the carriage?
I’m afraid I don’t understand the question…
One side of the carriage has the belts’ grippers fixed (embedded in the plastic part). The other side has the slide. There is one Bolt that goes through the whole carriage to pull the Slide. Only one side moves and pulls both belts.
I hope this makes it a bit clearer … :-/
I think this is the opposite of what you want. The belts in CoreXY don’t need to be exactly the same tension. You can use differential belt tensioning to square up the gantry and compensate for small build errors or inaccurate printed parts. That’s part of the beauty of CoreXY.
Hmm, after reading your response I incline to agree. Never thought of imbalanced system as an advantage but squaring the top plane is a difficult one because you can’t put there a cross bracing (it has to remain open).
Back to the drawing board …
After giving it some thought I realize that the squareness of the frame can be pretty much “roughed-in” at installation time. So one will insert the belt with the same teeth in the gripper as I have suggested before, then work it out by advancing by one tooth into the gripper until one can measure the frame straight. (But maybe that needs much finer adjustment than what the GT2 pitch can provide, although with such a long belt as for CoreXY it might be enough to also compensate for the belt stretch)
Only then one would apply proper tension with the mechanism I propose.
Would that work? Having two slide mechanisms in the design is a bit of a stretch for such a small 3d printed part. And, in the lights of the previous explanation, seems unnecessary.
Another thing: once you’ve dialed in the squareness, every time you want to adjust/play with the tension in belts you would have to do it the exact same way on both adjusting mechanisms if you had one for each belt, which is difficult to achieve with repeatable results.
You should only need to adjust the belt tension twice ever. Once at initial install and then again maybe a week later to take up a little bit of creep. Then they stop stretching. If you need more than that, it probably means you have PLA parts overstressed and they are slowly failing under load.
It’s ok to have a single joint tensioning point, but I recommend having independent tensioners too. (Slotted motor brackets are really easy.) 2mm of belt tension adjustment can be a lot of gantry angle error.