Large printer X&Y prototype done after many design revisions. Now I understand why @Deezmaker gave away prototype parts as part of his kickstarter (there are so many!).
I’m using a self-designed carriage + $5 wheels mounted horizontally on 20mm extruded aluminum (aka experiment #1). The axles are 6mm aluminum rods (experiment #2). Currently using wood dowels on the cross bars just for sizing purposes. Eventually they will be replaced with carbon fiber rods (experiment #3), since I don’t want to cut the carbon fiber until the very end. The next step is to design and print up some motor mounts to see if the motors have enough torque to turn the shaft. If not, the design will have to change so I can fit in some gearing. The red knobs are just to keep people from poking themselves on the ends of the rod.
Nice work @Chris_Horton Hope you don’t mind doing some friendly feedback
You’ve got 2 more shafts than you need to support your carriage. You will just be adding friction and binding problems with 4. Unless I’m missing something in the design of course.
I agree with @Tim_Rastall as this design is kinematically the same as the Ultimaker, tantillus, and ingentis, all of which are already over constrained. The addition if a fourth prismatic joint on each axis will lead to more binding if not perfectly aligned. The trolleys on the outer bars and a single middle rod should give you enough structure.
“Overconstraint” isn’t inherently a bad thing. It’s only a bad thing when your alignment is off. The reason why a lot of reprap designers moved to 3 linear bearings instead of 4 was because the people building the machines would regularly have absolutely terrible alignment. 3 bearings allow a little wiggle room for things to still work, where 4 bearings might cause binding. I see no “problem” with the dual-rod design, due to the drive system. It’s a bit wasteful, imho, but not “wrong” per-se, so long as the carriage is aligned properly to begin with and pulleys don’t slip, etc.
Over a span that long, the extra rods may need to be there for rigidity. I can’t be sure without either trying it or doing a bunch of math. I’m curious to hear how the carbon fiber rods work out there.
Part of the reason for terrible alignment in earlier RepRap designs was the difficulty of adjustment. I had a rough time with the X axis rods on my Prusa i1 until I popped the 4th bearing out of the carriage. In that case, there was no structural advantage to the 4th bearing, so it had to go.
Thanks for the questions! This was a completely independent design, but there are only so many options. I was going for simple/cheap, but large. Thus the regular extruded aluminum vs vslot. Industrial vslot is much more expensive, and the linear bearings for 80/20 are like $70 a piece.
The wheels I have are $5 each from http://framingtech.com. They’re probably not 100% perfectly round, but I haven’t tested that yet. I’ll try to post pictures tonight so you can see how they’re made. One side of the hub + interior bearing is just a metal sleeve and acts as a thrust bearing. The other side is just a washer. The mechanical engineers can tell me if I should flip it, but I have the sleeve side down.
I designed for two rods because I’m using carbon fiber tubes, not metal, and I was worried about needing more rigidity over that span (~.5m). It may turn out I can do with one.
After taking apart/reconstructing this a few times I’ve found some tricks for aligning it, but was thinking I might have to add something for fine-tuning. I’ll know more once I get it running.
Forgot to mention I’m also going to try a new kind of bearing that’s basically a plastic sleeve. Will post pictures later.
@Chris_Horton good on you for trying so many new things. I’m interested to see how the cf tubes turn out, what sort of dimensional/straightness tolerances do they have?
If it doesn’t work out, my Ingentis design uses single 8mm shafts for the X and Y cross, spanning about 400mm so you could happily use 10mm shaft for 500mm span. Wrt to getting everything aligned. I use printed spacers to lock the crossbar to a parallel structural member, tighten up the belts then repeat for the other axis. This works well as long as your frame is very square.
@Tim_Rastall Good question - I haven’t checked them scientifically (they’re from a hobby store, so no tolerance data included), but they roll together smoothly, suggesting they’re fairly straight. The proof will be in the printing (or not).
Your technique is similar to what I discovered. I loosen the pulley on one side so the two carriages can move independently, slide the crossbar all the way to the end, and then tighten everything back up.
I also designed new belt tighteners (half is acetone welded to the carriages), here: http://www.thingiverse.com/thing:181349, which makes belt adjustment easier without requiring me to cut them yet.
I imagine that with 2 sets of rods for X and 2 sets of rods for Y, you can place the hotend and extruder above the rods and just stick the hotend nozzle down through the hole or your could hand it down from them or somewhere inbetween. There is alot of flexibility there.
Do you know if you could put on side panels (for heat retention) and how you might do that?
@NathanielStenzel Re the hot-end, exactly. See my latest post for the extruder holder prototype.
Regarding side panels, I’ve got moving parts all around the top of the frame, so it can’t be totally enclosed just using the existing frame. My existing Printrbot is in a PVC “cage” with large size oven bags cut & taped around it to make an enclosure. I may make a larger one for this, eventually.
Ah. I see now. The two identical parts go back to back at different 90 degree angles. Cool.
While side panels may seem impossible to fully enclose, I think you may be able to get it down to a few 1 inch holes which could be taken care of by a skirt or a bristels like the ones in this hand broom.
@NathanielStenzel im using t slot for my new build which is similar to @Tim_Rastall s ingentis I am waiting to receive my plexi that is 6mm and if you put your frame together like I have you can Mount it in the slots like Window frames the top is open unless I make s box but just a few corner notches and holes drilled for the rod holes is all that’s needed
