Maybe the World's Lightest Extruder

Quote
arsi

Quote
1) The knife edge had to be symmetric, or the extrude and retract would be different.

I have tried different knife shapes and the best solution for me is to thin the bearing flange to 0.2-0.3mm and perpendicularly ground to the required diameter
The force required for rotation has not changed...

But the main problem, which appears with all knife shapes, I have not yet managed to eliminate. The pressure on the filament from the hotend, causes a change in the thread pitch at different speeds of extrusion and therefore the number of steps per millimeter. According to this video, I'm not alone [youtu.be]

I have been using my extruder for well over a year now, and this has not been an issue. Tom and I have discussed it, and I think that using high quality (ABEC 5) bearings and precision shoulder screws is pretty important. If the bearings can wobble at all, there will of course be an effect on the extrusion and retraction.

Quote
arsi
But the main problem, which appears with all knife shapes, I have not yet managed to eliminate. The pressure on the filament from the hotend, causes a change in the thread pitch at different speeds of extrusion and therefore the number of steps per millimeter. According to this video, I'm not alone [youtu.be]

This is why RepRapFirmware supports nonlinear extrusion, to allow for the extruder steps/mm increasing with extrusion speed due to the increased back pressure from the hot end. See [reprap.org].

---

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

Quote
arsi
But the main problem, which appears with all knife shapes, I have not yet managed to eliminate. The pressure on the filament from the hotend, causes a change in the thread pitch at different speeds of extrusion and therefore the number of steps per millimeter. According to this video, I'm not alone [youtu.be]

This is why RepRapFirmware supports nonlinear extrusion, to allow for the extruder steps/mm increasing with extrusion speed due to the increased back pressure from the hot end. See [reprap.org].

Indeed, today I fired up an old geared hobbs extruder to check for this effect, and the same effect is there. I suspect that the results are so far down in the noise that no one has ever noticed it.
A slightly eccentric hob will far overpower any variation between hob tooth spacing.

Edited 1 time(s). Last edit at 09/18/2022 07:19PM by rq3.

Hi arsi,

Hum I've still a lot of job to get a fully usable extruder, but until now, with the tests I 've done, the cutting bearing makes a beautiful continuous thread mark, it is is very regular (visually seems to at least).

Which pressure do you mean instead, the pressure in between the wheels or the pressure to go through the heat block ? in my opinion, If it is to go through the heat block it means that your filament "glides" or "wears". Maybe the edge of the sharp bearing is not enough cutting, or your material is escaping like caoutchouc.

In my case the edge is sharp and it makes really like a clean microcut in the filament.

I'll work further and it, and report you the further results

---

ADG AKA GA3D_tech ------ [www.instagram.com]------ [github.com] ------ [ga3d.tech]

Hi everybody,

Before that I go further in my project version of the kind of extruder I've done a quick pulling test, to check the capacity of my design :

Youtube - Pulling test Rolling Thread Extruder GA3D.tech

I want to cry, only a little load on the filament is already making a lot of problems.

the filament can be cutted, or twisted, it does not matter indeed, it is just not going forward.

Thank you @arsi to have me warned about that.

I must take a step back to find a solution or at least make a break on the topic until I find a smart lazy efficient solution.

---

ADG AKA GA3D_tech ------ [www.instagram.com]------ [github.com] ------ [ga3d.tech]

But nonlinear extrusion would mean a new calibration for each filament and hotend temperature change.
First I want to try to include a bowden extruder in parallel. This could create a filament preload...

You're making a fundamental error in thinking. Been there, done that. The extruder doesn't PULL on the filament (except to get it off the spool). It PUSHES the filament in to the nozzle.

Edited 1 time(s). Last edit at 09/24/2022 12:32PM by rq3.

Quote
rq3
You're making a fundamental error in thinking. Been there, done that. The extruder doesn't PULL on the filament (except to get it off the spool). It PUSHES the filament in to the nozzle.

It's easier to simulate it that way. But it's about the same thing. I figured it out by measurements on a real hotend during extrusion. (50mm of filament at 1mm/s vs 5mm/s and the difference was 8mm)
A conventional extruder will easily rip the filament from my fingers when I hold it or loses steps, in this case unwinding the filament from the spool will have an impact on the pitch of the cut thread.. And on Monday I will measure it on my hemera, I never thought to measure it and I'm curious about the result..

Edited 1 time(s). Last edit at 09/24/2022 01:30PM by arsi.

Quote
arsi

Quote
rq3
You're making a fundamental error in thinking. Been there, done that. The extruder doesn't PULL on the filament (except to get it off the spool). It PUSHES the filament in to the nozzle.

It's easier to simulate it that way. But it's about the same thing. I figured it out by measurements on a real hotend during extrusion. (50mm of filament at 1mm/s vs 5mm/s and the difference was 8mm)
A conventional extruder will easily rip the filament from my fingers when I hold it or loses steps, in this case unwinding the filament from the spool will have an impact on the pitch of the cut thread..

No. It's not "about the same thing" at all.

@rq3

Yes you are right traction does not exactly equal compression.

But in my case, If I push or I pull the filament, I have only one "knife" stuck in the filament, if the effort is enough to start the cutting process : I cut the filament instead of pushing it.

with my design I must add more blades to elevate the effort needed to start the cutting the process. I could do a kind of tube with many sharp bearing but I must think how to make it simple and easy.

---

ADG AKA GA3D_tech ------ [www.instagram.com]------ [github.com] ------ [ga3d.tech]

Quote
rq3
The attached photo is of a piezo drive motor that juuuust barely works. It weighs 4 grams.

Sorry for pulling this old topic. But I am curious about setup with piezo. You played enough with VDE, so you can have some feeling here. Let assume, if we have piezo based VDE, with blade plane angle about 10 degrees. Alone it will no work. But will it be able to control passthrough of filament, if filament comes to VDE already with force, which is enough to extrude? Will it be able to block passage of the filament, if it comes with this force, but motor is not rotating? (I assume that 10 degree of angle is not enough to force motor to rotate, isn't it?)

Because, if it can work as, effectively, controllable filament gate, then we can try to build really lightest and actually working setup, with piezo-based VDE as gate, and controllable force delivery via bowden from strong extruder (details are in my post in hollow shaft extruder thread on duet3d board).
In this case we separate system into parts, local one gives precision, and remote provides force.

BTW, am I right, assuming that this motor is something like RAS4P2009 ? I bet it is not cheap way
But, in any way, any tiny motor, which is capable to be used as stepper (with encoder), and torque which is enough to deal with friction here, might be enough, isn't it?

Edited 4 time(s). Last edit at 08/02/2025 04:42PM by antst.

That's a great idea. The bowden motor will need an encoder to run in closed-loop torque mode, but the precision motor can run open loop for simplicity. The bowden motor can read the same direction signal as the precision motor, so it switches to pulling when it's time for retraction without any need for firmware modification. I'll reassemble my NEMA11 VDE and give it a try.

One issue is that the bowden motor can't use a toothed gripper since the indentations in the filament would likely cause trouble for the VDE. Something like Proper Printing's belt extruder should work.

Edited 1 time(s). Last edit at 08/04/2025 12:11AM by dekutree64.

Ideally, you need torque mod WITH force sensor on VDE side, to account for delays in bowden tube. Instead of torque calculations (what is in normal torque closed loop), you need to use force sensor measurements on VDE side of bowden (or, ideally, on both sides of bowden, but this is next step).
SimpleFOC supports that. For force measurement one need to use strain gauge (most likely KFGS-2-120-C1-11) on special fixture.
Fixture, I think, is rather simple, but must be non-flexible material (better aluminium or carbon, but PLA for start). I'll draw later.
Unfortunately, ready to use ring load cells are damn expensive.

Also, one need to patch firmware to act on red/green light from bowden controller, especially if VDE is acting as pacemaker/filament gate with very weak but very light motor and low angle blade tilt. So, you avoid situations when VED is trying to push filament opposite to mechanical force on filament at VED entrance (because, there is a delay on force sign change, due to compression/decompression of filament in bowden).

So, to achieve really proper work of system, is a bit more work than merely add torque-controlled bowden extruder with sign change.


And yes, I was thinking also about belted one, like proper printing

Edited 9 time(s). Last edit at 08/04/2025 09:16AM by antst.

Damn, I am going to vacation in 5 days, and will not receive before that components I need, to make testbed and do some measurements!

Something like that

Bowden extruder: Some really powerful and fast BLDC, weight does not matter. Preferably belted extrusion.
BLDC controller: B-G431B-ESC1 (has CAN bus support, which allow us to enable communication with printboard in future)
Force sensor: fixture on the base of (1, 2 or 4) KFGS-2-120-C1-11.
UL(ultralight) VDE: low tilt angle for blades, very weak motor (as planning this around pieso which cost like 1K is pointless, I am going to investigate what I can do with something like 1804 motors).

Edited 2 time(s). Last edit at 08/04/2025 09:57AM by antst.

I don't think it would need a force sensor or anything fancy. The bowden compression delay shouldn't be a problem since retraction moves don't require much force. The bowden motor only needs to respond fast enough to stop applying force in opposition to the VDE before it starts moving after a direction change, and to get the filament fully compressed before the VDE finishes de-retraction moves.

And I don't think it would even need a controller as fancy as B-G431B-ESC1. Since the bowden motor will only be applying high force at low speed, it can use voltage control without current sense. Current/torque will decrease when it speeds up during retraction moves, but again it doesn't need to apply much force.

But after giving it some more thought, I think it's more trouble than it's worth. Since VDE is backdriveable, we can at most double its usable force without risk of the bowden motor overpowering it. This approach would be better with a worm gear or something that can't backdrive. But then a little motor may have trouble turning the worm against the friction of the bowden motor pushing on it.

Quote
dekutree64
I don't think it would need a force sensor or anything fancy. The bowden compression delay shouldn't be a problem since retraction moves don't require much force. The bowden motor only needs to respond fast enough to stop applying force in opposition to the VDE before it starts moving after a direction change, and to get the filament fully compressed before the VDE finishes de-retraction moves.

And I don't think it would even need a controller as fancy as B-G431B-ESC1. Since the bowden motor will only be applying high force at low speed, it can use voltage control without current sense. Current/torque will decrease when it speeds up during retraction moves, but again it doesn't need to apply much force.

But after giving it some more thought, I think it's more trouble than it's worth. Since VDE is backdriveable, we can at most double its usable force without risk of the bowden motor overpowering it. This approach would be better with a worm gear or something that can't backdrive. But then a little motor may have trouble turning the worm against the friction of the bowden motor pushing on it.

1) Assuming that pacemaker is weak and light, starting retraction when force is still there will not work. Same either extrusion after retraction, when force didn't built up yet. And, in general, having precise force control can be this setup even better than normal DD.
2) Choice of B-G431B-ESC1 is for simple reason, CAN bus. Although, not CAN-FD... It is not settled yet, in other words
3) Important part of Idea is to use low blade tilt VDE. So it is not backdriveable! This is important moment.