Okay so I just sat down to start mocking up my first pass at a CoreXY design with linear guides to support all the movement and the only thing I hadn’t decided upon is whether to use stacked belts or crossed belts and in my research I found the link below. Is there currently an open source firmware that would support this because this is absolutely amazing and would definitely be preferable.
TL;DR basically happened upon an open source (with not much documentation yet) Independent Dual Extrusion printer
@Cristian_Nicola I originally was going to do two heads controlled by one CoreXY gantry but this would definitely be preferable. I really like the design of the Eclips3D2 but there are a few design elements I want to change like using only linear guides to support movement and no rails. Plus it’s only single extrusion.
I would LOVE to make a dual CoreXY if I can find a decent firmware to support it which is where the designer of the printer above seems to be stuck.
e3d has dual extrusion hot ends (chimera, cyclops) and you could alway use one of them Y spliiters for filament. designing and implementing a dual corexy gantry is too much effort for too little benefit.
Pro tip: “if you build it, they will come” does NOT apply to firmware support for wonky new printer builds. (Learned this the hard way myself.)
You shouldn’t build something you can’t write code for yourself unless you have a commitment from one of the major firmware devs to help. There are a lot of reasons why these builds might end up being a lot more complex to control than you expect. Control of over-constrained mechanisms like this is particularly challenging when it comes to adding additional synchronous motor channels beyond the usual XYZE. The number of motion motors is pretty deeply baked into the code in most firmwares. Even if the motion controller is easily extensible (like MachineKit) you’re still going to have to figure out stuff like homing routines.
@Cristian_Nicola I think I’d rather use dual Aero heads instead of their legends heads. I ordered the multi-material kit for my Prusa so part of the reason I’m making a custom printer is so I can use PVA to support PETG and the Cyclops and the Y splitters are counter to that purpose. The Chimera is nice but I like the cooling better on their circular heat sinks.
@Ryan_Carlyle Yeah I’m coming to realize that this won’t work on any firmware that exists, especially after talking to @raykholo . I think I’m gonna go for a drive system similar to BCN3D’s Sigma but I’m a little dissapointed I’m gonna have those extra motors as moving mass. Still not sure, and I’ll probably end up swapping back and forth between IDEX and standard dual over the next couple days. If I end up going with IDEX I’ll be using Marlin.
@Adam_Steinmark You can “wire up” ANYTHING with MachineKit’s HAL system basically as an elaborate config file – that’d be the first thing I would suggest here – but it’s got a steep learning curve. And it’s already kind of difficult for MachineKit to control machines like CoreXY where the relationship between actuated joints and coordinate axes is weird.
Plus, an XY bridge gantry is just so much simpler for IDEX that it’s hard to justify any other approach. Just make the Y axis nice and beefy and the mass really isn’t a problem.
@Ryan_Carlyle Yeah I think I’m gonna do the bridge gantry and maybe use the Ultimaker style of movement so I can keep the X axis motors off the moving bridge. But that would require using rods which wouldn’t be the end of the world.
@Adam_Steinmark Dunno about a UM (cross gantry) IDEX setup… you’ll need the X stage to have twice as many parts so you can control a separate X cross rod. I was meaning Replicator 1/2/2x style bridge gantries. There you just need an extra motor and belt for the second carriage on the X bridge.
@Cristian_Nicola Too much room for error in grabbing the toolheads. I’d sooner do the UM3 style of switch heads.
@Ryan_Carlyle Yeah I know, like the Sigma, but that would require the motors to move with the bridge. That plus dual direct drive is just too much weight.
@Adam_Steinmark weight is only a problem if you under-spec the frame, linear hardware, or Y motor. In my opinion those are much easier problems to solve than figuring out how to build a double-cross gantry.
@Adam_Steinmark You can use 9mm wide genuine Gates GT2/GT3 belts if you’re worried about stretch. Don’t forget the bridge gantry has two Y belts sharing the load and they’re not very long. UM gantries have a non-trivial amount of moving mass in the rotating rods too.
Don’t get me wrong, I am absolutely a believer in CoreXY and cross gantries as superior mechanisms, but the vanilla XY bridge gantry really, really makes sense for IDEX.
@Adam_Steinmark Well, you’re going to need twice as much X hardware, so four X belts and six X rods… unless you only use one rotating rod per X stage on opposite sides of the gantry, which should work. But in either case you’re dealing with height offsets on the carriages to clip the belts and more bearings to support.
It’s a whole lot of engineering on a gantry package that’s already highly-optimized for space utilization and parts count. Nothing insurmountable, just more work.
@Ryan_Carlyle Doesn’t sound like a huge deal. I’d still use linear guides to support each head instead of rods so only 4 X rods using rods on each side. I’d have to do the same sort of system for the Y axis so another 2 rods there. The hardest part is gonna be like you said, making everything work with the height offsets but it shouldn’t be too bad. I may simply pose the question to the community, whether they’d rather see IDEX with less moving mass and more complicated gantry or just stick with the moving X motors. Price isn’t too much of a factor for me, I don’t really have any delusions that this won’t be a super cheap printer.