For a long time, extruder weight has been a compromise. Direct drive is heavy but grants greater capabilities to actual extrusion, while Bowden is light but sacrifices print quality and reliability. It’s time for a change. I call my concept a Valved Bowden, although truthfully, it’s a direct drive extruder being fed by a Bowden. Let me explain.

Direct drive weights are being reduced with gearing and smaller motors, but we are still a lot heavier than we could be. The main problem is that a lighter and smaller extruder has trouble supplying torque. My Idea was to find a way to assist a smaller motor. Bowden extruders can print without any direct drive assistance at all, so if a Bowden extruder is added upstream, an extruder with much less torque can be used and still produce reliable print results. I found a NEMA14 ultra thin stepper motor for PTZ camera - RobotDigg which weighs just 60 grams but still supposedly outputs 500g/cm of torque. That’s not enough for traditional direct drive, but could potentially work to control the end of filament and reduce slop. That’s why I call it a Valved Bowden rather than a hybrid design: it’s really more of a Bowden design. I’m going to work towards refining the idea, look for more updates soon!

---

Proud owner of a Flashforge Adventurer 3 and an Anet A6-L that's entirely in pieces.

I'm trying to imagine how this would work. Maybe put a slightly larger drive gear on the remote extruder (or set a smaller steps/mm) so that it is always trying to push or pull a little more than the smaller direct extruder, but...

there is limited space inside the tube for the filament. On a long extrusion without retraction the filament will "fill" the available space as the smaller extruder refuses to push more filament. The filament will be under "pressure" to keep going forward so when a retraction eventually comes, it may not be sufficient to relieve the "pressure" in the filament tube and the small extruder may not be able to retract against that pressure.

If you could have variable retract so that the retraction distance increases as the non retraction extrusion gets longer, it might be doable.

Edited 1 time(s). Last edit at 01/22/2021 12:35PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

The idea is that they both drive at the same time, using a stepper splitter. The small extruder will extrude when the large one extrudes, and vice versa. I'm getting the parts to do some prototyping, I'll keep posting.

If they are both extruding the exact same amount, the small one can't add/subtract from the extrusion. The only benefit will be that the already minimal force required to pull the filament from the spool will be reduced by the remote extruder.

What is the problem you're trying to solve and how does this scheme solve it?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

The one on the end adds control without majorly increasing weight. It's not meant to help with extrusion, but rather to reduce the backlash of the filament, reducing string and allowing printing with flexible filament. Hopefully.

How does it add control?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

I am planning on implementing a similar system, although weight reduction wasn't my primary concern, but rather a coincidental benefit. I was going to route my filament through a tube to protect it on its way to a direct drive setup, but I was hoping to automate the loading or unloading process by including something like an old MK8 located near the filament spool.
There is a couple ways of doing this, but without knowing specifications as to your printer's hardware or firmware these are just the options that I see.

I'm not familiar with a step splitter, but if it works the way I assume then I don't see why you couldn't use one. The only concern may be the current draw, but using smaller motors shouldn't be a problem.

If you are using something like a RAMPS 1.4 and Marlin, using the Mixing Extruder settings may be your best bet. This lets you use both extruder stepper drivers and allows you to tune feed rates between the two motors if you want to implement something like what the_digital_dentist recommended.

If you are handy with Marlin you could possibly modify the firmware (pins.h is where I would start) and use the Z axis driver (which has two 4 pin stepper motor connectors already) and swap the axis with one of the extruder drivers.

While I haven't had much experience with ReprapFirmware, you should also be able to map the E0 and the E1 drive to the E0 driver. Actually you should be able to do this with most firmwares, as they all should be able to handle mapping multiple motors to the same axis, like the 2 motor driven Z axis on a prusa.

Quote
the_digital_dentist
How does it add control?

I believe the concept is that the lost control due to drag and bending inside the bowden tube is regained by including a small stepper at the extruder. This seems reasonable to me.

When you push filament into a curved bowden tube, some of the force you apply goes into bending the tube, and the bending allows more filament to be inside the bowden tube. You can see that happen. Conversely, when you retract, some of the filament you retract is already in the tube (because it's bent, and the filament goes round the outside of the bend). So when you're changing extrusion rates, the pressure at the extruder will lag a little behind what your stepper motor is doing. I think that the "pressure advance" feature is intended to compensate for that. However, ideally the amount of advance needs to vary with the bend in the bowden tube, which itself will vary with the X,Y,Z position of the hot-end, depending on the geometry of the mechanism and spool location.

I applaud the innovative thinking and experimentation that is being seen here, and hope he gets great results.