jhertzberg

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.

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.

Wow! Rails? That's one sure way to keep the bed stable!

If you are looking to print architectural models at double the x and y, but don't need more z, you may want to look at a gMax. It was designed by an architect specifically with this in mind.

The z screw controls the height and holds the weight. The smooth rods prevent wobble. You may want to either add one linked leadscrew opposite the current one, and add two more smooth rods in the front corners, or put leadscrews in the front corners, and smooth rod front middle. The delux approach would be to run each of three leadscrews on separate steppers, use a sensor, and microstep to automatically level the bed.

At some point, do you think it makes sense to move the CoreXY gantry on the z axis to simplify making a level (and hot) bed?

That's really good quality for flex! What were your print settings?

A note of caution. I also used the bearing spacer, but with Nick Foley's direct drive Twister Blocks). It worked fine for the y rod. However, the clearance at the front left corner of x rod before the pulley of the y rod is ~13mm, but the bearing spacer protrudes 16mm. After failing in my attempts to trim and ream my two spare spacers, I had to cut a spacer out of doubled over quarter inch plywood. I presume you avoided this pitfall by mounting the x motor on the front right corner, since the rear has space no longer used by a motor pulley. I redesigned the spacer for 12mm, but it's such a hassle to unmount the motors that it will go untested until my next big upgrade. (BTW, your motor block link is broken. It has a space on the end.)

Is that your Merlin hotend that you are using with it?

feohnsturm, I love this! There have been threads about this on the umforum a few times (eg. here) but up until now most people have just tried with dremel type flex shafts. I also like the reuse of your modular print head. Will you be sharing the design? With all the hot end options you keep making, it can become a UM standard. Can you describe the flex shaft provided by Mutley3d, how heavy, how long, how rigid? Is there any appreciable resistance to head moves that tighten the bend radius?

I hadn't thought of the weight driving down on the motor, though I wonder if a lot of weight pulling down on the motor shaft is any better for it. I'm hoping that heated bed kit gets announced at the NY Maker Faire next month. I want to see what the company solution will be before doing anything more with my z stage. I have been thinking of moving my electronics to back as well. Underneath my machine is about the dirtiest place in my home office, not the best place for an electronics cooling fan. It would also make it a lot easier to fiddle with the electronics if I didn't have to upend the printer every time.

I love the idea of using an oldham coupling. It moves the complexity of dealing with inexact alignment out of the nut trap, where space is tight. A few questions regarding the couplers you went with: What benefit did you get from moving the z motor to the top? Cooling? Gravity assist in keeping the leadscrew plumb? More z height? What material did you go with for the center disc of the coupling? McMaster has both Acetal and Nylon. My gut is to go with Acetal out of concern that Nylon might be too "mushy" for very small movements like with .06mm layers or jorised walls. Did you go with set screw, or with clamp-on couplings? My stock double set screw couplings seem to loosen up every few months. Should I assume that clap-on couplings won't suffer this as often, or am I just being wishful?

I suspect that Ultimaker may announce the UM1 heated bed upgrade at World Maker Faire in NYC on September 20th. If they do, I'll buy it! Then, I may enclose the chamber as well. I already switched to direct drive, moving my x/y motors out of the chamber.

I never tried kapton. (I'm cheap.) I always hated trying to get air bubbles out from under blue tape. I usually resorted to piercing the tape with an Exacto knife to let the air out, which then increased the likelihood that the tape would pull up when removing a print. With the glass, I'm not using any tape. I paint the glue onto the glass with a two inch house painter's brush. It takes a couple of hours to dry, but that's okay since I have two glass plates and the glue is good for several prints before retreatment. When It gets too messed up, I wash the plate and repaint it. I keep a sealed jar of diluted glue by the printer. Any PVA based glue, such as wood glue, will do.

My UM1 is spotted with upgrades, but I'd have to say that the upgrade that adds most to my print quality is replacing the acrylic bed with glass and using Murat's three point leveling with magnet mounts: http://www.thingiverse.com/thing:84754 I use diluted Elmers Glue-All on the glass for great adhesion. I have two pieces of glass, so I can print on one while I am removing a print from and preparing the other. The magnets on the three point set-up let me swap the beds without having to re-level. And the bottom of an object printed on glass is, well, like glass.

Thanks. I looked at the OsiriX manuals table of contents online, but I don't see where it exports to STL. Is it through a plug-in?

Thanks, that's an interesting group.

For discrete small prints, I would think that any number of the head mounted http://www.thingiverse.com/search/page:1?q=syringe+extruder&sa= would be the way to go. For larger prints or a more continuous process, perhaps a head mounted http://www.thingiverse.com/search/page:1?q=moineau&sa= would be better. Regarding other peoples MRIs, I found that the http://3dprint.nih.gov/ is great, but I am trying to print the CT of my own tib/fib spiral fracture as a thank you gift for my surgeon.

What software are you using to convert DICOM CT files to STL? I have tried invesalius and 3DSlicer, but I'm still having to do a lot of cleanup to get manifold models.

It appears to be a Chrome compatibility problem. I'm on latest version of OS X. I just checked, and it works fine on latest Safari and Firefox, but not on Chrome.

You can try friction welding. Take a short length of filament and use it as a bit in a rotary tool. http://hackaday.com/2012/12/31/make-your-own-plastic-friction-welder/. Wear eye protection.

I determined that the UM1 leadscrew was 3mm pitch via the observation that when I had a bad z wobble, the banding period was 3mm.

From the Trespa US web site: "Trespa® Meteon® is a decorative high-pressure compact laminate (HPL) with an integral surface manufactured using Trespa’s unique in-house technologies, Electron Beam Curing (EBC) and Dry Forming (DF). The blend of up to 70% wood-based fibers and thermosetting resins, manufactured under high pressures and temperatures yields a highly stable, dense panel with good strength-to-weight ratio’s." Considering the composition, how do you think it would perform also as a print bed for Nylon? I tried Garolite LE, but the piece that I got from McMaster-Carr was too warped to be useful without serious flattening and bonding to a metal base. Are the Trespa samples that you received uniformly flat?

Jeremie did a nice writeup on his lessons learned in printing nylon: http://www.tridimake.com/2014/01/how-to-3d-print-nylon-and-trimmer-line.html The takeaway seemed to be to dry it out, but not too much, or you will have delamination issues.

A couple of folks on these threads have tried the extra wheel idea. An inexpensive rotary encoder could at least let you know if the filament is grinding. For the feeder, I don't think that a simple rubber wheel will give you enough traction, especially with slippery filaments like nylon, but Ian has a great idea for a filament tractor: http://umforum.ultimaker.com/index.php?/topic/4393-ultimaker2-feeder-system-improvements-and-ideas/?p=52246