Need a hint for a printer design


I'm constructing a new 3D print with coreXY
So my question is: Should i stack the X rods horizontal or vertical? What are the pros and cons? (I take rods because I already have them... )

Thanks for your hints,


If I understand you correctly, you're asking about pairing rods for the X dimension to add stability. My guess is that it's the same lag due to gravity in the middle regardless of the orientation, to be honest.

I would orient them to minimize the amount of complexity.


Yes, I just need two because one would not work at all. I should have attached the image before. This is my current design:

I think it would be more easy to attach hotends when I arrange the rods vertical.


Oh. I must have misunderstood you before. I thought you wanted two horizontal rods each in front and back for the sake of stability. Ignore that, then.

I'm guessing that most of the work earlier on this was for CNC.

Have you seen this?


Nice work in the video.
My thought go into the direction of practicability of the different designs. When I have a look at my MK2, it seems easier to mount a hotend, but in decreasing the height.
I will have to think about it, especially because it has not to be finished tomorrow. :grin:


I love how fast the printer was in the example video. I may have to consider doing one of these.


It's pretty solid speed for rapids, but the main problem with speed is usually the hot end keeping up. It's nice and rigid, though, since that build uses 2x2 extrusions, so it's a pretty good design for dimensionally stable and accurate prints at the hot end's maximum flow, provided the backlash is worked out of the system. It could probably benefit from preloaded cross bracing, though, which would make it a space frame and thus much stiffer...that'd definitely support super high speeds (with a volcano, if you want to break 100mm/s regularly).


In fact, speed is not that important for me.
For me, a reliable machine is that, what I want to have.
Just start, control the first layer and let it run.
I fact, my old K8200 that I sacrificed for this new project worked so fine, that I even did not need to check the first layer.


For me, I want to build humungous parts and if you don't crank up the speed the damn thing will never finish. I would compromise on print quality for a very big printer.


Yes, that's a fact I have to consider too.


Same for me (large to very large parts), though I don't really want to sacrifice quality, hence the value of a very rigid frame that won't lead to quality losses at speed. CoreXY machines have some nice advantages over FFCP or enclosed Prusa-style machines, but if you're going to build a large machine, keep in mind you will need wider belts than your typical printer. CoreXY uses very long belt paths to begin with, so a larger printer will have a correspondingly epic belt length. 5 or 6mm GT2 belts will introduce backlash on a huge machine, as they will stretch far more than you are accustomed to, if properly tensioned, since they will be driving not just the mass of the gantry and extruder, but also almost twice the length of belt. More mass, over more belt, yields more stretching (primarily from the extra length of the belt, not the added mass, but the extra mass amplifies the effect).

The solution is a wider belt, which resists the stretching forces better, or a chain drive, which will resist stretching far better, but is heavier and will require larger motors and may require larger linear motion components, depending on your design. You'd need a pretty huge machine before the gantry mass required a chain drive; as long as you use thick enough belts with kevlar or glass threads (steel will fatigue and cheap belts with nylon threads will stretch enough to kill your tolerances) and your mounting components are all rigid, you should be able to build a pretty large CoreXY printer that's fast and has decent tolerances.

The thing machines seem to screw up most is the Z axis, IMHO. With all the focus on cheap construction in the industry, really solid Z axis designs are less prevalent than you'd hope. Prusa-style Z axes are rarely preloaded, so they don't scale all that well, and FFCP-style Z stages (CoreXY machines need a Z stage or have to move an entire XY frame vertically, so they fit in here) are notorious prone to lateral wobble and sagging under large amounts of weight due to having only 2 Z axis rods, close together, near the Z screw.

The best large Z stages use either side rails or 3-4 rods, though I've seen a few that used 2 beefy Z rods, with longer linear bearings and a lot more mass in the stage and mount. The beefed up designs have a lot of mass in the Z stage, though, so while they stabilize the stage and minimize play and cantilever droop (because the stage has so much more mass that even high infill prints that are huge don't have nearly as distorted a mass ratios to the stage), they need much more holding force in the Z axis and often have glacial Z moves due to getting the holding force by using larger ratios on moderately larger motors instead of massive motors with lower ratios. The side effect of the ratios is the option of thinner layers if the Z screw's tolerances are tight enough, but thinner layers add more time, so it's a tradeoff to if you actually use that (and most of those machines aren't using screws with tight enough tolerances anyway).

Now if only we could have our cake, eat it, and have another cake baked before we're done with our slice, too... :wink:


I've been auto-playing all the videos in the (concrete) home-building machines, always impressed with the weight that's been jogged around like that.


Yeah, they're pretty neat. I've thought of buying some land and making a machine that could fab rebar reinforced segments in place and cut, place, and weld angle, channel, and I beams...toss in a few containers if you want (but you don't have to) and you could fab a pretty nice home for pretty cheap (all materials, virtually no labor), and have it be virtually everything proof.


...and then make a million by designing and renting out the equipment.


Well, it won't be free to build, so why not make the money for a vacation home off it? :smiley:


How about... a combo container + 3D concrete? Or combine pre-cast huge concrete pipe plus 3D printed concrete?


@Ewald_Ikemann Have you seen this?


Well, combo container and 3D printed reinforced concrete is what I was implying. More as an option than anything else. The cool thing about my idea is that it's capable of arbitrary form reinforced concrete casting, because it's capable of doing supports. Obviously you can't do true bridging with concrete without using support, but with a machine capable of arbitrary support and arbitrary reinforcement, you can print/cast safe concrete structures of arbitrary shapes and sizes in place, in an automated manner.


If you want to do that and still have the speed advantages of CoreXY, I'd suggest you modify parts so that you can use larger extrusions...single 20x20 extrusions are not as rigid as ideal for that kind of machine. I'd do at least 2x2 if you're using 20mm extrusions or, if you want to use singles, I'd use 35 or 40mm profiles.

Just my personal opinion, though, since I like to keep my options open for speed and scalability.


BTW, 72" ID, 8' long sections weigh just over a ton each. They are reinforced, too, so I'd imagine they cost more than a container, though they reduce your insulation costs pretty substantially, I'd imagine.