Tool switching and kinematic couplings

[Ryan Carlyle]

E3D has a kinematic coupling on their tool-changing motion platform now, right? https://e3d-online.com/blog/2018/03/21/tool-changer-q The coupling gives 100% repeatable alignment for each tool relative to the XY carriage. So you can record fixed XYZ offsets for each tool and get repeatable tool switching. 

What I want to do is come up with a different kinematic coupling arrangement that ensures the nozzle tip is always centralized as well, so there's no XY offset to measure/record, only different Z heights. A Z height offset can be probed out with any kind of Zmin sensor that triggers off nozzle contact. (Piezo mount, FSR, nozzle hitting a microswitch, etc.) So if your tool changer coupling positively positions the nozzle's XY in relation to the coupling, you eliminate a ton of setup work for the tool changing system. In fact, you could probably make it plug-and-play: install any new hot end (or plotter pen, or drag knife, or...) and the system measures the Z offset itself and is ready to run. 

A standard mill type cone-in-cone taper profile can do this, but that's not a great solution here, because of the need for high fab precision and complex automated drawbars or whatever. Drawbars would get in the way of the filament path, too.

My thinking for a first concept is to design a kinematic coupling that self-centers the coupling on a groove-mount or similar cylindrical point in the tool. Preload via magnets or whatever.

Groovemount diameter isn't quite right? Want to grab a dry erase marker? No problem, it'll accommodate different diameters via a slight rotation of the tool in relation to the carriage. All you should need is for the tool/nozzle tip to be centered in relation to the cylinder you're grabbing.

I think this should work. Might be missing something, I just thought of it. Depending on the fab precision of everything, you may still need a small offset calibration (like with a vernier scale dualstrusion test print) but it'll get you darn close right off the bat, which is worth something. 

[Whosawhatsis]

Maybe I'm not seeing what you're trying to say, but in that diagram, it looks like there would be an offset along the direction of the adjustment screw as the diameter changed... I also don't see how the triangle relates to the diameter of the circle. It doesn't seem to be the same in each version of the diagram, nor does it seem to scale with the circle.

--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To post to this group, send email to 3dp-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/d775e15c-9aa2-44be-af8b-f9d183f6f3ec%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Ryan Carlyle]

Yeah, not my best drawing. Needs a 3d render probably. 

1) The drawn blue triangle is showing the ball positions for the carriage side of the coupling (equilateral triangle). Imagine a ball at each triangle corner. (I should have drawn the equilateral triangle as stationary and rotated the groovemount clamp.) Those carriage coupling balls sit in grooves on the removable clamp side of the coupling, two on the V-piece and one on the adjustable piece that clamps the groovemount.  

2) When the adjustable piece is tightened, the incomplete equilateral triangle formed by the three kinematic grooves always shares a center with the groovemount cylinder. This occurs so long as the clamping pieces remain in a 60-60-60 angles and have the same offset from the clamping surface to the groove. That gives you a triangle of grooves that changes size to match the clamped object.

3) The carriage side coupling is a fixed-size equilateral triangle of balls. The removable side coupling is a variable-size equilateral triangle of grooves. When these interface, they will share a center.

The variable offset you're seeing along the direction of the adjustment screw is not seen at the nozzle. The hot end clamp moves off-center with different diameters, but the hot end stays centered on the carriage. 

To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+unsubscribe@googlegroups.com.

[Whosawhatsis]

Ah, I get it. Might be hard to keep the triangle planar and equilateral as you tighten and loosen, but otherwise it should work.

To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.

To post to this group, send email to 3dp-...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/7ad71132-067a-4cf2-b0fa-3c0edb886b1f%40googlegroups.com.

[Ryan Carlyle]

Yeah, that's my main concern. I was originally thinking the adjustable clampgroove piece could use a V-notch to centralize and align itself off the groovemount cylinder (which in turn centralizes and aligns off the stationary clamp part) but that would make the groovemount-clampgroove radius vary with the groovemount diameter, thus undoing the intrinsic self-centralization. 

I think if the adjustable clampgroove piece V-notch is sufficiently small relative to the diameter of the groovemount, it will be "accurate enough" but I haven't put a ton of thought into it. 

One alternative is to mount the adjustable clampgroove piece on a parallelogram linkage. That's not a perfectly accurate or mechanically simple solution though.