What is the best customizable 3D printer system these days?

[strantor]

the "rota-printer" that shares more in common with a CNC lathe than it does a router.

If this is what you're talking about then you understand what I would like to do.

The video shows printing worm gears onto a shaft, which is exactly what I would like to do for starters. It is inspired by this:

Not exactly the gold standard in surface finish. I have made improvements with acetone treatment on ABS:

And by machining on PETG:

But it would be so much better to print these in the correct axis (the rotary axis) like shown in his video. This rotoprinter what I consider a stepping stone between 3 axis printing and 4/5 axis printing.

The quality of printing in the video isn't great but hopefully it can be improved.

 

[Lo-Fi]

That's it! It wouldn't be easy, but making a printer like that is not impossible. The tricky part is going to be getting the software to generate gcode as any of the standard slicers are a flat no-go... I suspect you're into the realms of general CAM software, but you'd need (or want) it to be able to add all the 3D printer specific stuff in, along with the tool path. It's not trivial!

 

[strantor]

That's it! It wouldn't be easy, but making a printer like that is not impossible. The tricky part is going to be getting the software to generate gcode as any of the standard slicers are a flat no-go... I suspect you're into the realms of general CAM software, but you'd need (or want) it to be able to add all the 3D printer specific stuff in, along with the tool path. It's not trivial!

I have been thinking about how to do it and I think if you modeled it as an unwrapped shape you could print it with ordinary slicer software. Just the Y-motor will be a C-axis but neither printer or slicer will know (or need to know) that fact. It may need some manual modification to the G-code to increase extrusion ratio as the extruder gets further from the center of rotation.

I imaging this will result in a gnarly seam where the two ends of the software-unwrapped/real-world-wrapped part edges come together, but maybe not.

Other ideas include unwrapping it more than a single circumference. I haven't thought that all the way through.

 

[Lo-Fi]

That would work for a single layer, but you'd have to do some calculations and change the movement, extrusion and all sorts of other things as the diameter/circumference changes and the layers get bigger in one dimension!
You could get rid of most of the seams by utilising the continuous rotation, but that's not available if you treat the cylinder as an unwrapped panel, sadly.
Shame the chap who made the rotoprinter didn't publish any of his work beyond some videos.

 

[GrayTech]

I'll concede that for the vast majority of things that get made on 3D printers (knick knacks, toys, ornaments, etc) a 4th axis would not be of much benefit. But for those of us trying to print functional parts that will see real use, it could be a big deal.

It's always been a nagging gripe in the back of my mind (the kind I keep to myself because I know nobody cares) that we call these things "3D" printers when in reality they are 2.5D printers. They make 3D parts but the operation isn't 3D, it is just a list of 2D operations separated by Z steps. Bed height correction doesn't count. You've probably never seen a 3D printer go from X10Y10Z10 to X50Y50Z50. The word "slicer" says it all. These parts will always have a grain like wood and will always be weak in one axis. And will always require supports for overhangs.

Adding axes could change that. Imagine a sphere printed on a normal 3D (2.5D) printer; it's a stack of disks starting with small disks, going to larger disks, and back to small disks, probably with a bunch of wasted material for support to keep the thing from falling over. Now imagine that sphere printed on a true 3D printer. It would have to start at the end of a rod probably, and look like a lollipop. It would start out as a little bud at the end of the rod, then grow from the inside out. Instead of slices it has layers, like an onion. Each of these onion layers are printed in a different orientation so there is no axis that is weaker than any other axis. A very orderly and uniform ball of yarn. Now take that sphere concept and apply it to anything else.

There is a good deal of situations where this either won't work, or generating the tool paths required to avoid collisions will demand such a powerful slicer (unraveler?) that it will be more trouble than it's worth. For simple parts though, well... I want it bad enough to try making it.


I am intrigued by these as well. I am just waiting for an excuse to get one.

Thanks for the recommendation! Marlin is what my old Prusa uses. Is Arduio IDE still the tool used to flash the flash the firmware?

Can you post a picture? I am not sure what you mean by "not restricted to linear motion."

The print head hangs on 3 (6) arms and can move in any direction. Not bound to xyz axes. It's not modifiable in the way you intend.
For a picture or video you can google "anycubic predator". Can't seem to load anything here myself which is a shame.

It occurred to me that the rotational printing you wish to do is simple enough if you replace x or y axis with rotation. Still a 3 axis printer essentially.

Sent from my SM-A528B using Tapatalk

 

[strantor]

That would work for a single layer, but you'd have to do some calculations and change the movement, extrusion and all sorts of other things as the diameter/circumference changes and the layers get bigger in one dimension!
You could get rid of most of the seams by utilising the continuous rotation, but that's not available if you treat the cylinder as an unwrapped panel, sadly.
Shame the chap who made the rotoprinter didn't publish any of his work beyond some videos.

I discovered something rather titillating. Apparently you can generate additive g-code for multi-axis machine right inside Fusion360, and the feature has been available for over two years now. Not just XYZC but think of robotic arm toolpaths. This bullet at the end makes me think that currently it only supports welding/laser tools, but I think that the G-code it generates should be close enough and/or mendable with a Python script.

Fusion 360’s new Multi-Axis Deposition toolpaths aim to remove some of those limitations, enabling users to fully use their machines to create 3, 3+2, and multi-axis toolpaths. You can now:

• Create and store process information in a deposition tool (currently limited to DED-type tools).
• Create planar or non-planar part slices from a defined base surface.
• Define a raster infill and perimeter pass build style.
• Control the tool axis for deposition.
• Visualize their toolpath layer by layer.
• Play through simulation or produce a deposition stock model that you can input to a subsequent milling setup for part finishing.

Immediately after reading this (like 10 minutes ago) I jumped on Fusion360 and tried to follow these instructions to get started. I was unable to do it as some of the mentioned options were missing, but after acknowledging the "update available" notification that I've been ignoring for ages, the options appeared. I am now diving head first into this to see if it can work. I've never used fusion 360 to generate G-code, so starting out with something complicated like this I'm sure isn't recommended, but screw it.

EDIT
From here:

The Additive > Multi-Axis > Feature Construction

strategy lets you generate a multi-axis additive toolpath. The toolpath is primarily for driving machines using directed energy deposition (DED) technology; however, with edits to the post processor, it can also be applied to other multi-axis deposition applications, for example, fused filament fabrication (FFF) and concrete printing.

So, it can (allegedly) be used for 3D printing, no need for Python Band-aids.

 

[Lo-Fi]

You'll find this interesting:

Robot CNC 5 axis

Including Pentarod, Open5, CoreXY 5 axis. 5 Bar Parallel Scara

docs.duet3d.com

 

[strantor]

You'll find this interesting:

Robot CNC 5 axis

Including Pentarod, Open5, CoreXY 5 axis. 5 Bar Parallel Scara

docs.duet3d.com

From your link:

A CAD, CAM program, slicer or postprocessor will create G-Code which can be executed and is used to control the 5 axis CNC axes.

That's what I'm trying to do right now. I got all excited about the Fusion360 thing, went down the rabbit hole of tutorials, and wound up here:

 

[GrayTech]

Another thing to keep on mind is that there are different flavors of g-code. Commands and functions differ by flavor.

Sent from my SM-A528B using Tapatalk

 

[Lo-Fi]

Oooof, that's spendy! This is the problem when you're moving outside hobbyist trodden paths... Unless someone pulls their sleeves up and carries people along with them into an open source project (and this someone could be be you, of course), the commercial option is the only one.

The key to it is keeping people who know it can't be done away from those doing it.