Mendel Build Height

The Mendel wiki (http://reprap.org/wiki/Mendel) specifies a build envelope with a height of 140 mm. But with Adrian's geared extruder, I found that I can only go a little over 70 mm before the gear hits one of the frame bars. I don't know how high you can go with Wade's extruder but it looks like the gear which is rotated 90 degrees would also hit. Is anyone working on a solution to increase the height by either redesigning the extruder or the frame?

The motor hits the frame with Wade's. I only get 35mm, but that is because I have lost 23mm with my heated bed and about 10mm with my extruder's heatsink. That would give about 70mm with Wade's and I always assumed Adrian's would go higher because the gear does not stick out as far as the motor does on Wade's.

---

[www.hydraraptor.blogspot.com]

This is true. I beleive that you can get ~140mm without anything mounted on the carriage. To maximize your build height, you would probably want to move towards an Bowden Extruder setup like you find on a Ultimaker (http://blog.ultimaker.com/) This way you do not limit the Z height by the extruder height.
I am using a Stepper Extruder from Makergear (http://www.Makergear.com/) and i am able to max out safely at ~100mm travel in Z. This could be improved by compressing the springs on your print bed as much as possible while still remaining level.

I would still say that Bowden Extruder(http://reprap.org/wiki/Erik's_Bowden_Extruder) is the way to go.

The Bowden extruder is a good idea. Aside from the increased height, it's also easier to fit multiple extruders. One problem is the hysteresis resulting from the long filament and the sleeve between the gear and the hot end. Perhaps there are better cable guides out there using different materials or having metal lining like the cable guides in bikes.

Instead of a bowden cable, how about keeping the gear right above the hot end and driving it with a remote motor through a flex shaft like this?

[www.amazon.com]

This would have the same advantage and might be better in terms of hysteresis.

Since that item you linked is designed to work with a dremel I would imagine that it would not be stiff enough to predict the exact result when turning the stepper a few steps one way or the other.

Would the mechanism that allows the shaft to be flexible not also allow it to twist independant of the drive to some degree...?

I guess only experimentation would tell.

Thing is you are simply removing the motor but still need the gearing... erm actually I guess you would have the gearing with the motor and connect this directly to the filament feed...

Instead of the flex shaft, you could use something like a spline shaft to move the motor to the end of the X axis. The spline shaft would lie parallel to the X axis, and the pinion gear would slide along it to maintain contact with the main gear on the extruder. Of course, you don't need a proper spline shaft for this application. You could maybe use a small rectangular aluminum extrusion for the shaft, or perhaps some spur gear shaft.


I like the aluminum box tube a little better than the gear shaft. I think it will be easier to mate parts to. This idea will add a little drag to the X axis, but it also moves the stepper off the carriage while avoiding the issues of the bowden extruder. It would add a little weight for the Z axis to lift, but it shouldn't be much.

Or, you could just rework the frame vertices to get more vertical space, instead of constraining the already complex extruder design with requirements dictated by an inconvenient static structure.

Edited 1 time(s). Last edit at 02/03/2011 10:20AM by Dale Dunn.

AgeingHippy Wrote:
-------------------------------------------------------
> Since that item you linked is designed to work
> with a dremel I would imagine that it would not be
> stiff enough to predict the exact result when
> turning the stepper a few steps one way or the
> other.
>
> Would the mechanism that allows the shaft to be
> flexible not also allow it to twist independant of
> the drive to some degree...?

I used the dremel flex shaft just as example. You're right. I checked the one that I have and there is significant play between the two ends. I don't know if there are any that would be flexible enough so that the extruder can move relative to the motor and still be rigid enough to prevent twisting. Unless there is such a thing, the Bowden extruder might still be better.

Dale Dunn Wrote:
-------------------------------------------------------
> Instead of the flex shaft, you could use something
> like a spline shaft to move the motor to the end
> of the X axis. The spline shaft would lie parallel
...

This is also a good idea and would make it possible to mount multiple extruders. I thought about a splined shaft too but I was concerned about the additional friction on the X-axis. I also thought it would be difficult to line up the shaft with the gears over the full range of the X-axis. But your idea of using a rectangular rod might work.

Hmmm. Some Teflon inside the pinion might reduce friction some, if experimentation shows it to be problematic.

I was thinking about the multiple extruder possibilities too. Is anybody running multiple physical extruders yet?

The Prusa Mendel uses PLA bushings in place of bearings in many places. Perhaps this would be a good material for the pinion.

I don't see why the bowden extruder design on the Ultimaker (http://blog.ultimaker.com/) that seems to have room for 4 extruders can't be adapted to the Mendel.

Has anyone tried using hypocycloid gear reduction in place of the gears in either Wade's or Adrian's geared extruders? This looks like a practical way to increase build height by eliminating the large driven gear in those extruders. In this page, the author used a two-stage gear for the challenge but one stage should be enough for the extruder.

I and a few others have investigated differential hypocycloids, will experiment when my reprap is printing!

Check out this and this

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

Im working on a parametric herringbone hypocycloid speed reducer design in openscad (wow thats a mouthful!)

I have some issues at the moment though: note the render time!

Greg Frost Wrote:
-------------------------------------------------------
> I have some issues at the moment though: note the render time!

Looks like you need a more powerful PC.

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

Greg Frost Wrote:
-------------------------------------------------------
> Im working on a parametric herringbone hypocycloid
> speed reducer design in openscad (wow thats a
> mouthful!)
>
Great!

Maybe submit your file to the openscad guys as a bug report? That's a really unreasonable compile render time.

---

--
I'm building it with Baling Wire

How about posting your hypocycloid.scad file, maybe it can be optimized for speed...

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

Triffid_Hunter Wrote:
-------------------------------------------------------
> I and a few others have investigated differential
> hypocycloids, will experiment when my reprap is
> printing!
>
> Check out this and this


Your youtube links don't work. I've noticed that before as well.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!