Tilting Dual Extruder

[BONE]

Has anyone done something like for their Rostock? http://www.spiderbot.eu/en/blog/dual-he ... lease.html

[dtgriscom]

Looks interesting, but I'd be concerned about using those limit screws to limit the servo motion. I've posted a comment on the blog about that.

[Polygonhell]

I'm not sure how they can patent it given it's exactly what some of the commercial dual extruder designs do. Although the one I saw used a somewhat different mechanism for the actual tilt.
The more I play with dual extrusion the more I think the right solution is a single nozzle with multiple paths feeding it, much like richrap mixing extruder or something as simple as this https://reprappro.com/2014/04/15/multim ... -research/ I think I saw something similar on this forum somewhere.
Using a single nozzle just solves so many of the issues with dual material printing, the one big downside is if you wanted to use radically different temperatures for the materials, the material change would be costly.

[RegB]

I agree, both richrap and Boyers have it "about right".
Not sure that switching colors needs purging to be done on the infill or at a "sacrificial dump" part beside the real part(s).
With paths of known length and cross section (known volume) switch ahead by the amount of material that is in each pipeline - a fairly simple "look ahead".

[Generic Default]

[img]http://forum.seemecnc.com/download/file ... ew&id=3001[/img]
[img]http://forum.seemecnc.com/download/file.php?id=3002[/img]

I made that around a year ago to try to fix the problem. It merges up to three different filaments in one hotend. At the time, the software wasn't very supportive of it but now Cura would work with this very nicely I think. I still got a few working prints of two types of nylon at the same time being printed. Up to three filaments in one hotend, one filament in the other. At the time I was designing the mount for the hotends I had available. Now I can just make my own hotends entirely!

This has the limitations of a single nozzle size, the need to heat up or cool down if different plastics are being used, and large retractions+large purgings on each change. So it can be optimized with a shorter hotend to waste less plastic and be faster, but it will always be slower than alternatives.

Since then I've come up with a few solutions, one of which I can't describe here for another week or so. I believe that a multi-nozzle hotend with leveled nozzles and independent temperature control and filament feed will be the best solution.

Furthermore, a multi-nozzle hotend like my Tri hotend can be merged with the multi-input thing shown above. For example one nozzle would print dissolvable PVA at 200 degrees, one nozzle would print trimmer line at 270 degrees, and the third would print multiple colors of ABS through larger retractions with the multi-input thing at 245 degrees.

Colors can be mixed this way too, but the hotend would need a mixing chamber. More products for this kind of stuff are coming later this year and next year. My company specializes in this particular field as the name implies! Expect some mind blowin' inventions in the near future.


Slightly off topic, but multi-nozzle extruders are becoming more and more relevant now that there are more and more types of filaments available.
Just yesterday I was filament shopping on Amazon. I saw polycarbonate, PC-ABS, PETG, and even Acetal POM on sale for use with our printers at $50 per kilo. Plain ABS is down to $20 per kilo and the quality is decent (at least that's my experience with the cheap stuff so far). The biggest advantage of all for multiple nozzles is support material. It can't be understated!


EDIT;
The patent office has recently passed several 3d printing related patents through even though there is extensive prior art. If anyone patents this design I'ma be pissed.

[McSlappy]

Is there no end to your amazing inventions ?

[Generic Default]

Hopefully not. I bet that other people in reprap have been making similar filament mergers since like 2007. Justin@Saturday made one that looks like it would work a lot better than mine, and his works for the newest E3D v6. Check out his thread!

[teoman]

You can achieve the tilt differently aswell, all you need is a 4 dof robot. (adding a 4th tower).

That way you can have your nozzle tilt aswell...

I am currently designing such a mechanism for a completely random personal project.

[Generic Default]

Teoman do you know the kinematics needed for a 4th tower? I was thinking that even with three towers you could have a tilting effector plate, but you could only use one ball joint on each side. Four towers would constrain the mechanics I think. Stuff would have to be well-calibrated. Still, it may be worth doing and it would have huge benefits. The only change would be the extra tower and the firmware. Magnetic ball arms provide all the motion you would ever need.

[Polygonhell]

3 towers with one balljoints each is under constrained, the effector would be able to rotate around the vertical axis, moving in Z as it did.
With one balljoint each link only removes a single degree of freedom from the system, so you need 6 arms to be completely constrained. In the standard rostock design each arm pair removes 2 degrees, so you can get away with 3 of them.
With 4 arms you'd have to arrange it so you had 2 arm pairs opposite each other then 2 single ball joint arms opposite each other at 90 degrees to the first pair, I think that's sufficient, to allow constrained motion with tilting.

[teoman]

Delta kinematics are slightly tricky.

With 3 motors you can have a maximum of 3 degrees of freedom which we all would probably prefer to be X Y and Z. (Others being yaw, pitch and roll).

Our deltas kill some of those degrees os freecom on purpose using double/parallel arms. If the arms were single with one universal joint at the end, then there would be nothing (except gravity) from the end effector (nozzle) from rotating in the z axis and moving up at the same time.