CoreXYU -> CoreXYE

I prefer light setups for 3d printing. Unfortunately direct extruder (the best way to feed a filament) goes against it therefore I'm looking for ways around it.

Looking at successful CoreXYU designs I think it should be feasible to use belts from the second X carriage to drive extruder gear instead. Such setup should be more accurate than remote direct drive (faster retract) or flying extruder (shorter filament path).

Interesting idea, I can't see why a second or third belt run shouldn't be able to drive an extruder. That being said remote direct drive extruders do work very well, they retract fast enough if setup correctly.

You'll have to figure out how to ensure the forces are balanced or the force driving the extruder might skew the gantry.

Edited 1 time(s). Last edit at 08/30/2017 10:35AM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

It would definitely be possible, and would probably work very well. My main concern is that you are adding a lot of complexity to the design to move only 200g off the carriage (maybe even less if you net out the weight of all the additional moving components).

You have to weigh the cost/benefit of a fixed extruder motor vs the cost/benefit of making the frame and linear guides more rigid. Its not obvious to me that a ~150g weight savings would produce better results than switching to MGN rail or moving to 40x40 extrusion.

I can identify one obvious problem with the idea, which is that you need a continuous loop of belt for the E axis. Its trivial to find a 6m length of open-ended belt, but I'm even sure if its possible to acquire a >2m loop of GT2 at any price. Even if you grab the largest standard closed-loop belt available you will have hard time building a printer with more than 8"x8" travel.

That's a good point about the continuous belt but there are techniques for splicing belts together.

I'd agree with your argument that whilst it's a cool concept, it's a complex way to save 200g when cable extruders are readily available and very capable.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Can you use gears and such instead of a long continuous belt?

You can get 2mm pitch belts longer than 2m, but there will be only a couple of specific lengths to choose from. Slightly more selection, maybe, if you are willing to go to MXL belts.

Avoid MXL pulleys and belts unless you are willing to pay for pricey Western-manufactured pulleys. The Chinese so-called MXL pulleys I have come across have the wrong tooth profile. The MXL belt has wider gaps between teeth than the width of each tooth, therefore the teeth on the pulleys needs to be wider than the gaps, otherwise you get horrendous backlash.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42
Avoid MXL pulleys and belts unless you are willing to pay for pricey Western-manufactured pulleys. The Chinese so-called MXL pulleys I have come across have the wrong tooth profile. The MXL belt has wider gaps between teeth than the width of each tooth, therefore the teeth on the pulleys needs to be wider than the gaps, otherwise you get horrendous backlash.

You probably want to avoid Chinese MXL belts as well - they are noticably less stiff than US manufactured belts (Gates, etc.).

My next question would be, if it makes sense to go for a dual extruder right away?
The saved weight should be good for something else than just speed.
Another problem not addressed yet is the orientation of the drive gear. The drive gear shaft is horizontal, while idlers and pulleys are vertical.
LoboCNC has made a 90° hotend once, but I don't know the pros/cons of that. ( horizontal filament path= vertical drive gear )

Edited 1 time(s). Last edit at 08/31/2017 01:42AM by o_lampe.

Hey, that hot end that LoboCNC made would be a perfect companion to the Sidewinder (working title) extruder. I will investigate. But sorry, not relevant here.

I am not sure if the whole assembly will be as light as a RDD extruder. The weight of a RDD extruder ranges between 40 grams to 27 grams, depending on which cable driven extruder you are talking about.
And remember, this is the whole extruder, including the hob and idler and bearings. The same stuff you will still need if you use a belt driven extruder.

Nevertheless, go for it! It is a great way to expand the shared knowledge and we can all learn something new. It is nice to be able to experiment and test out new ideas, so why not?

One question, if the extruder is extruding and the whole print head is moving in the x direction, will that influence the extrusion rate? Or is the belt moving along the x axis at the same rate? (clearly I have not built or used a CoreXY printer yet, it is next on my list)

---

Lykle
________________________________________________

Co-creator of the Zesty Nimble, worlds lightest Direct Drive extruder.
[zesty.tech]

If the belt path for the extruder is anything like a corexy belt path the firmware would have to factor the x and Y movement into the number of steps the extruder motor turns as the head moves. I'm sure this is something DC can do in his head, for the challenge of it. It shouldn't be that hard to work out how many steps to add/subtract to ensure the extruder pulley turns the right amount.

Edited 1 time(s). Last edit at 08/31/2017 08:06AM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
o_lampe
Another problem not addressed yet is the orientation of the drive gear. The drive gear shaft is horizontal, while idlers and pulleys are vertical.

I had a healthy sleep that night and it dawned me, you maybe could add the extruder belt path(s) to one of the sides or back of the cube? That way the horizontal orientation works. Still I have no clue, how it would look in detail.

I think the most straightforward way to do this would be to run a standard H-Bot or CoreXY path on top of the regular one. Mentally replace the motor in this picture with your extruder, I just drew on top of something from google.



It wouldn't be hard to make the belt-driven E axis present the same shaft/hole pattern as a Nema17 stepper, allowing the use of regular extruder/hotends.


Technically the arrangement only requires three motors, but a four motor design probably makes sense. Four motors can apply double torque to the Y axis, which is useful because the Y axis is heavier and needs more force to match X axis accelerations. Four motors also makes the design symmetrical which is nice.



Note that the motors can be on the same side of the printer but the belt anchor points should be on opposite sides of the carriage. I believe a pair of H-Bot paths can offset their respective racking forces, but I'd have to think about it.

It seems that a Zesty Nimble would do what you're looking for:
Removes motor mass from carriage, while still being direct drive.
Zesty Nimble, by Zesty Tech

Sure not totally the same, it uses a flexible drive cable, rather than adding another belt loop :p

Maybe using spectra line instead of belt would allow simpler design (even better if tying it into a loop using a knot would work).

If it turns out to be an issue, pinch type extruder (two wheels driven by the belt turning in opposite direction) would cancel out the torque applied onto X carriage.

Of all things I think the most important would be to assure that extrusion force does not exceed max force needed for reliable X and Y movement since moves are interdependent (E needs to skip a step before X or Y). Therefore, torque applied to extruder's gear would probably have to be lowered using geared extruder.

Edited 2 time(s). Last edit at 09/04/2017 11:31AM by newbob.