Hollow shaft extruder
-
@breed I've build one version of the screw extruder, too. The "skrewder" has 3 screws and their bearings are able to tilt in the holders. I was able to manually load the filament by opening the screws. Then wrap the screws around the filament until they have sufficient grip.
I almost bought me a lathe to make these screws from metal, but it occured to me that the thread pitch and the cant-angle wouldn't cut a single thread in the filament. They only scratch the surface, which is not as reliable as the threading blades of the VDE-100.
It would be a game changer if the screw pitch and cant angle would match. (mass production-wise)
Especially if we could use standard threaded rods. -
@o_lampe How did it work out when you tried driving filament with it? Did it scrub real bad? The problem is that the lead angle for the threads is different between the screw which is say 9-10mm in diameter and the filament which is 1.75.
-
@breed I didn't really try driving it with the motor. The screws are resin printed. But I got a pretty good idea about how tight the wrapper must be to grip the filament.
Much tighter with more threads, obviously.
Maybe It would be a good idea to form a funnel with the three screws, so they can slowly dig deeper in the filament. -
@o_lampe if you think about it, the cant of the screw already does because of the threads lead angle.
-
@JoergS5 said in Hollow shaft extruder:
One topic would be to solve the twisting force on the filament. If you're using two screws, finding a solution to rotate both in opposite direction would be preferable in my understanding.
The same applies the the VDE100 but the torque is so minimal that in practice this is a problem that doesn't need solving. Just the back-torque from the melted end of the filament and the torque coming from the filament winding off the reel are enough to counter the feed torque. Even with TPU! Having a short path through the heatsink also helps because that reduces the amount by which the filament can wind itself up before it gets to the melt zone.
Trying to drive the filament in two different directions with counter-rotating threads / bearing edges will result in (at best) two crossing helical threads on the filament. I have not tried this but my guess is they'd interact in undesirable ways.
-
@breed said in Hollow shaft extruder:
I put my old screw extruder design I had been working on a few years ago and an updated design made for a stepper motor that Ive been working on lately
I reckon that would work with the stepper motor we have specced for the VDE100.
In earlier experiments, rq3 who pioneered the VDE100 used a similar design with threaded rollers. And the fuselabs design does as well. On which topic, does your design by any chance predate 28 October 2019 when fuselabs filed their patent application? And did you public anywhere?
-
@tombrazier It was during COVID but don't remember when it was. I don't think it was that early I'm gonna guess late 2020 early 2021. It was originally for the anycubic predator which came out 2019 I think. It's basically a thread less ballscrew.
-
@tombrazier I made an extruder that used counter rotating nuts to move the filament. The problem with that was the two counter rotating nuts (or whatever cuts threads into the filament) will never grip the filament at exactly the same strength, so one will dominate and the filament will try to twist with it. I found it mainly to be a problem with retracting the filament. If the filament rotates a little when you do a retraction, you won't get the retraction you are expecting. I think you need a bend in the filament path at the input of the extruder to minimize that problem.
See https://drmrehorst.blogspot.com/2021/08/an-old-project-snakebite-extruder.html
-
@mrehorstdmd that looks alot like I was designing originally. Except it had double 3 screw assemblies inside of the bevel gears. I ran with it in cad for a bit but it ended up so big and more than twice as heavy. I was just looking for a vertical extruder with a small foot print to fit on the smart effector. I had double timing belt design I started also that had a fairly small foot print but I felt like the torque required to run it would require too big a motor and defeat the purpose. I'm really excited about the work these guys are doing on this VDE100. The going to a low count of "threads" by using the flange bearings allows a high lead angle that would not be possible on a normal threaded screw without scrubbing. Plus with the high lead the rpms come down into the realm of a regular stepper. The math on the screw design I was working on originally was like 6 revolutions per mm or something like that. Which is possible with bldc motor and I originally was trying to do the flex shaft remote direct drive thing I had first seen at Mrrf2019 which would like the high rpm low torque thing. I was worried though if the filament would carry enough heat away or if it would gum up the screw.
-
Does anyone still print with 3mm filament? Why? Why not? Seems to me like the VDE100 would be a little easier to make if the parts were a little bigger. It would potentially take longer to melt I guess but it seems as if hotend tech has outstripped extruder and cooling techniques at this point.
-
Please, can some one explain (for me, suffering acute dyscalculia)
how many rotations of the shaft is needed for tom´s above linked FreeCad Model?
I´ m not an engineer
My Bearings MF-85-2RS have diameter of 9,14mm.
How do "pitch circle diameter" of the earing axes has to be set.How deep should the sharpened flange cut in the Filament.
The Axes are in 20 ° .
15 ° would be very engThank you,
Gruß, zero K -
@o_lampe I've been thinking about what questions I have for the BLDC mounter VDE. In the end I reckon it all comes down to one question: how fast can it push filament through a hotend before the motor stalls assuming you run the motor at low enough current that it does not become too hot and (for me at least) with a 12V supply voltage.
-
@mrehorstdmd I have seen a number of posts and videos of the snakebite extruder in recent days. It looks like it was a great contribution to the rapidly growing pool of FDM designs and ideas in the last decade. Nice work IMHO.
The key difference, I think, between VDE-like designs and the screw designs of 10 years ago is that the screws freely roll over the filament in these designs, thereby almost eliminating torque on the filament. In principle, one might even be able to drive the screw at the right speed that torque on the filament is entirely removed but I suspect that would be overengineering.
-
@breed A 3mm hollow shaft stepper motor would be harder to source but many gimbal BLDCs would be be great. 3mm would also mean you could use larger flanged bearings and get more overlap with the filament for each bearing. For me the main reason not to try 3mm is everything else in my system assumes 1.75mm.
-
@zero-K Short answer: you get about 1.55mm of filament feed per rotation at 20° and 1.75mm filament. But depending on the specifics of your actual system, it will vary slightly. I find it's best just to measure it experimentally.
Long answer: will have to wait, I don't have the time to answer it right now unfortunately.
-
@tombrazier I only tested the sFOC setup for a few short prints. But I can tell, it doesn't get hot because of the close loop. There is no excess energy, like with static current steppers.
For 12V you should look for a motor with ~300kV. Mine has 140kV and runs on 24V. I had luck with it, because it also has low rotor cogging.
The rpm-limit is also determined by the sensor and the MCU clock. I can tell more, when the optical-disk sensor arrives. -
@tombrazier said in Hollow shaft extruder:
you get about 1.55mm of filament feed per rotation
That's about 2065 steps/mm for a regular stepper setup. Three times more than the 710 step/mm on the Sherpa mini and other extruders with the round NEMA14.
Maybe the stepper can be much smaller then? I have a super small NEMA08 stepper with hollow shaft. -
@tombrazier
I got two of this,
14HS14-1254H
from Stepper-Online, about 16 €, plus delivery and Tax.
It has 8 mm Shaft and 4 mm hole,
Hopefully the Power wil be OK.Gruß zero K
-
-
@tombrazier that 1.55 mm of lead is interesting. Should be more like 1.91mm. I wonder why? Are you finding that your thread pitch on the filament is staying consistent with varied speeds and loads?
