Loading project Things
A large Hbot 3D printer
Loading README and project media…
NOTE This is still a work in progess, it is currently printing quite well. UPDATE Changed the Y rails to Makerslide and Makeslide carriages and VWheels.
A large build volume 3d printer based on the CoreXY concept.
It uses 2020 Misumi extrusions for most of the frame.
For historical purposes...
Recently re wired to be Corexy rather than HBot. This seems to reduce racking of the X Gantry. See
I see no difference between using Spectra line drive and GT2 belts, in fact if anything it was slightly smoother using the line drive.
For Spectra drive the four idlers on the XGantry are grooved 5mm bearings that bolt directly to the extrusion slots. One of the X carriage bearings is raised 10mm to keep the lines parallel.
The four end idlers are grooved 5mm bearings that are screwed directly into the tapped center of the vertical extrusion. This provides very solid idlers than can take the tension of the spectra when it is tight. The grooves are necessary otherwise the spectra just runs off the top or bottom of the bearing. These bearings are stacked on top of each other with a washer in between, this allows the cables to run on top of each other and no cross over is needed.
The two end idlers are 2x flanged bearings back to back, and another pair separated with a washer above it, Then two of the idlers on the X Gantry are raised to keep the belt riding level. I avoid the cross over of the belts by running the two belts one above the other in parallel.
The scissor lift idea was nice but the prototype showed that a Nema17 did not have enough torque to lift the bed. I think the bottom of the scissor lift needs to lift 4 times the weight so any pressure on the top caused the motor to skip.
The new Z actuator is a length of 500mm of 2040 VSlot, with a printed carriage and v wheels. Driven by a McMaster ACME 1/2" leadscrew with 10TPI, Bearing blocks using 608 at top and bottom and a Nema17 with 5mm/0.25" flexible coupling. The bottom bearing block isolates the flexible coupling from taking any weight and compressing. As the 1/2" leadscrew is pretty big it has no wobble as far as I can tell and having bearings top and bottom seems to be very stable. I also used a printed Leadscrew nut printed in PLA and it seems pretty snug and smooth. I also used a half height leadscrew nut and spring to implement antibacklash which appears to be necessary.
There is a length of 2040 that spans from the Z actuator to the other side of the frame where another 500mm length of 2040 VSlot has a vslot mini carriage which is bolted to the end of the 2040 beam. Halfway along this beam is a short piece of 2020 perpendicular to it, this provides the third attachment point for the bed.
The most important part of this design is the moving knot, which is attached to the top and bottom and runs over 4 v groove bearings on the cross beam, This keeps the beam level while allowing a single actuator on one end to move it up and down. This is the far end, and here is a view of the whole assembly.
The bed is attached to the cross beam at three places to provide three point bed levelling. I use a scheme where a 4mm stud is screwed into the tslots, and a 25mm length of aluminum rod, tapped to 4mm is screwed onto that, this can be turned to move up and down the stud. At the top of the aluminum is a cut out that the glass bed fits. Next to this, and also bolted into the tslot, is an M4 bolt with a conical spring bolted to the top, this pushes the glass at the three attachment points up into the slots on the aluminum rounds. This keeps the glass flat without deforming it, and allows it to be leveled at three points by turning the aluminum rounds. Some jam nuts are used to screw up to the aluminum rounds to stop them unscrewing once levelled.
So this adds several printed parts... The Carriage, a bracket to hold the leadscrew nut, and bearing blocks. Also some lathe work to cut a thread into the center of some 1/2" aluminum round, and lathe a groove at the top that the glass can slot into. (If the bed is not heated this could be a printed part).
Aluminum parts are from Misumi with Misumi part numbers
OR
Getting the top of the frame square is essential, measure each end of the Y rails to make sure they are parallel. Next the belts need to have exactly the same tension to keep the X gantry straight, this was introducing non-square prints for me. Loosen the X gantry on both Y carriages so it sits loosely on the Y carriages, then measure the distance of each end of the X Gantry to the front corner of the frame. Now tighten one or other belt until the X Gantry is square to the Y rails, then tighten it down to the Y carriages.
If you find cylinders print out oblong or squares are not square this is probably the problem.
CC BY SA
WolfHBot by Jim Morris is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.Based on a work at https://github.com/wolfmanjm/wolfhbot.
