Screw Direct Drive Extruder!

Hello everyone,

I don't know if anyone has brought this up already. A search did not turn up anything.

So, I have been looking at all kinds of different compact extruder designs for a small X carriage and lo and behold:

(from BFB 3D Touch)
They use a pulley transmission to drive the threaded rod, but theoretically, if one was to attach the rod to the stepper directly and push the filament against it really hard, the generated torque should be enough to push it into the hotend. As the contact surface area is a lot bigger than with a hobbed bolt, filament strip could also be reduced.
Here is another picture with the pinch wheels in place:


If I am seeing this correctly, a ultra-compact extruder could be possible.
What do you think about this?

Here is the source by the way:
3D Touch extruder loading tutorial
3D Touch extruder update documentation (PDF download)

That is the way Reprap used to do it. See [hydraraptor.blogspot.co.uk] for my last design like that.

It gives very good grip on softer filament like HDPE and PCL. The downside is that it is very inefficient power wise. A lot of torque goes into cutting a thread, rather than driving it forward. It also makes a lot of dust and tries to rotate the filament, which can lead to tangles.

BFB use a custom machined screw, which presumably works a lot better.

---

[www.hydraraptor.blogspot.com]

Oh, I didn't know about that. bummer.

I guess the 3d Touch has an improved design, but the Rapman 3.1 has an earlier screw drive design and I have all the problems nophead mentions.

Over time the performance had got worse, turns out the hotend was jamming. I now have a new hot end, but the extruder performance is still not good, I suspect the thread is wearing out but that is just a guess. It has only ever worked reliably at fairly low speeds. Increasing the pressure on the roller bearing provides more grip, but the inefficiency of the design requires a lot more torque from the motor so the motor skips. Reducing the pressure allows the motor to run, but ends up carving a groove in the filament and then there is no drive on the filament at all.

The principle of having a high contact area seems good, but I think because the screw rotates against the filament it creates too much friction.