Flexible shaft - the best of both worlds?

[pm_dude 27]

@ovidiu That's weird. It shouldn't happen unless the screw holding the knot is loose or missing.

 

[BasVanDeursen 0]

Hi!

I've never mounted a hobbed wheel onto a UM2, I've only tested it's grip in a test rig that I had built for that purpose. If your knurled wheel keeps getting loose you could secure it onto the motor shaft with a drop of loctite.

Cheers,

Bas

[eldrick 65]

Meanwhile, back on topic, here's a first cut at a Flex3Drive adapter for the UM2, by Erik van der Zalm: https://www.youmagine.com/designs/ultimaker2-flex3drive

 

[pm_dude 27]

page not found

 

[IRobertI 520]

Link fixed. I want to kick the guy or gal who made the URL detection for this forum in the ass, because it's bloody useless.

 

[twistx 0]

Bought a flex3drive kit today to fit on my UM2. I'll give Erik's design a try.

[muttstang 8]

Could you use a set of 45 degree gears to minimize the need for lubrication? Probably would need an extra gear reduction to get 40:1 though.

might be a pick your poising kind of thing there.

 

[ultiarjan 1,223]

I Like to give Eriks design a try, what do I need in terms of software changes? what is the HEX file that comes with his files ?

 

[ultiarjan 1,223]

answer already in the YouMagine comments ...

 

[muttstang 8]

Here's a thought. What if the motor was mounted on a carriage that aligns itself with some form of telescoping rod with the extruder head? That way there is no relative rotation induced in the flexible shaft. Implementation may be tricky but I think this would get rid of that error.

 

[jhertzberg 19]

That would be interesting, but it I imagine that there are a couple of problems to overcome:

Slop - The more segments the telescoping rod has, the more slop there will be. I imagine that in order for rotation to be transferred between segments, they will each have to be keyed. If the key tolerance is too tight, the segments won't easily slide out from each other. If loose enough to slide, there will be slop. BTW, you could also consider just using a single splined shaft with ball splines, though that would be pretty expensive.

Backlash and/or software complexity - The simple way to mount the rod would be on a pivot assembly at the rear of the printer and a similar assembly at the head. If the motor pivots too, there will be a fair amount of lateral force exerted on the rod by the mass of of the swinging motor. If you mount the motor stationary and use bevel gears to turn the shaft, you are back to a relative rotation problem, since even when the motor is stopped the shaft bevel gear will rotate as it travels across the motor bevel gear, keeping the shaft pointed at the head. This rotation is opposite to the rotation needed to keep the base head bevel gear still. Then it's a trig calculation to determine how much to modify the extruder motor motion.

Weight - You need to keep the new head weight less than just putting a small geared NEMA 14 motor on the head.

 

Here's a thought. What if the motor was mounted on a carriage that aligns itself with some form of telescoping rod with the extruder head? That way there is no relative rotation induced in the flexible shaft. Implementation may be tricky but I think this would get rid of that error.

 

[jhertzberg 19]

If you are really worried about the relative rotation issue, you could instead mount the motor horizontally, and just do the gear reduction with regular gears on the head rather than the worm gear. You have to rig something up to keep the flex shaft from sagging into the high points of the model or into the belts.

Or, you could use a pair of bowden tubes to confine a loop of #6 beaded chain (it fits in a 4mm ID tube). Mount the motor horizontally, driving a beaded chain sprocket in a housing that holds the near ends of the two bowdens, and mount the head-end sprocket and housing in any orientation you wish.

 

[muttstang 8]

Or just increase the gear reduction some more depending on if the motor is capable of the RPM's at whatever the load ends up being. The bigger the gear reduction, the more that the relative rotation will be minimized. You can pretty easily get some 64 pitch worm gears and see a 60:1 reduction without a lot of size. You will likely have to buy the gears as the teeth would be getting a bit small for printing, especially with the higher friction loss that you get from worm gears

 

[foehnsturm 969]

There is an improved version of the original mount on https://www.youmagine.com/designs/modular-printhead-ultimaker-flex3drive-merlin now. Less parts and print-optimized.

 

[ccw1984 1]

@twistx, @UltiArjan, @EldRick

So does anyone have this up and running on their Ultimaker 2 yet? I would be interested in getting one if it's not too difficult to install and we have some success stories to back it up.

 

[Blizz 53]

Very curious about this one too.... Although getting the replacement pieces for my printer would be even nicer at this point

 

[BasVanDeursen 0]

Hi everybody,

At the moment I have Erik's design running on a UM2 However I did make a custom shaft for the hobbed wheel. The standard shaft that comes with the kit is about 10mm too short for Erik's design.

At the moment we are experimenting with this.

Kind regards,

-Bas

 

[zumfab 3]

I am keeping my eye out for experiences, I am very curious to know about the improvements and drawbacks this development brings!

 

[IRobertI 520]

Indeed, I'm looking forward to see what further experimentation brings as well. If there aren't any huge drawbacks I will for sure upgrade to this system as I really like the concept.

 

[izzy 33]

Hi Guys just 5 months late into this topic

It doesn't look like its adding too much extra mass to the print head for the inertia, and looks a neat solution, is there enough room for this in a Dual print head system that may arrive in the next 3 ish months.

Izzy

[meduza 191]

I was thinking about this relative rotation issue, shouldnt we be able to compensate for this in software? (or is the electronics not powerful enough to do this?)

 

[Anders Olsson 136]

I was thinking about exactly the same thing the other day :smile:

It should not be that difficult to calculate how much the shaft has rotated based on where the head is, really.

I guess it could be a feature in Cura, lite a "slice for flexible shaft"-button, in case it is too complex to integrate in the firmware.

 

[ultiarjan 1,223]

I guess if its flow compensation based on head location it can be done with a plugin... not saying i can do it

[aviphysics 9]

I was pondering that you might be able to use counter rotating shafts to reduce or eliminate the problem. The gearing could be design so that when both shafts turn the same way, the screw doesn't move, but it does when they are counter rotated.

Sort of like an automobile differential hooked up backwards, so two drive shafts and one wheel. Would need one more gear in there to invert the direction of one of the shafts. Would still only need one stepper motor. I think you might need a differential at both ends.

 

[restinpieces 3]

I can't get mine to work while printing. I built mine from scratch and the problem is as soon as I start printing the motor skips and vibrates violently. But I don't understand why. I made changes in the Config and I removed micro stepping. Every method of moving the filament (e.g. sending G1 or moving it with the rotary switch) works except when I try to print :(