A different approach to ext...

In which your narrator goes nuts...

As you know I've been designing Tommelise 3.0. T3 will be able to be configured either as a printer or a milling machine and is looking to cost about $150 in parts for the printer option.

Obviously, I've got to have an extruder if I'm going to do printing. Heretofore, I've been torn between building a "me, too" pinch wheel extruder or a more robust variation on the Mk II. With what seems like everybody and their Schnauzer building more flavours of pinch wheel extruders than Ben and Jerry's have flavours of ice cream and the formidible Nophead exploring variations on the Mk II, I was left to either wait for the dust to settle or somehow come up with some new wrinkle to those two initiatives that would add something to the state of the art.  I frankly couldn't see what I had to offer in that regard.

The dust hasn't settled yet on extruder designs. A clear favourite hasn't been clear to me at all. Most pinch wheel designs, while simple, require heavy NEMA steppers. I haven't been too happy with that. Nophead's, pinch wheel design and his variations on the Mk II, while light, require machined parts. I'm not happy with that, either.

That left me wondering whether there was a third way of designing a workable extruder that was both reliable, light and required no machined parts that couldn't be milled or printed on a Reprap machine.

I got to thinking about threaded nuts, for some reason. The problem with threaded rods is twofold. First they have to have bushings. You have to mill those and the damned things wear and create metal dust which gets into your extruded plastic if you don't break the system down and clean it quite frequently. Second, the threads curve away from the filament, which means that the threads have to cut quite deeply into the filament to get a proper grip.

A nut, on the other hand, curves around the filament, giving a much larger contact area. In the past there was a notion that you ought to be able to thread a nut onto the filament. The problem with that approach is that the filament tended to turn with the nut. I came up with a different approach.

I used a nut considerably larger than the filament and then milled a filament guide {the black HDPE part just above the ruler} that pushes the filament against the threads.





The nut, I embedded in a gear and drove that with another gear connected to the drive shaft of the gear motor. The bushings and drive shafts are milled out of HDPE and are made with large diameters to keep the contact area between shaft and bushings large and the force per unit area small. Given that HDPE is about as slick as Teflon friction isn't a problem.

The black filament guide is connected to the Z-axis and the filament pump floats free.  The two #8 threaded rods with wingnuts and springs provide the compression required to let the filament engage the threads inside the nut. 

• at 280 grams, it is light.
• it requires absolutely no especially machined metal parts.
• it can be milled, printed or laser cut {most of the parts for this last}.
• the feed rate can be increased by two-thirds simply by swapping the 1/2-20 fine pitch nut for a 1/2-12 coarse pitch nut.
• it can process both very stiff and highly flexible filament with equal ease.

If you had to you could probably cut these parts out of HDPE with a scroll saw.

I will be seeing if I can get this extruder running properly and assess it's robustness over the Summer.

Late addition:  I was able to get down to the hardware store this morning and get the requisite #8 nuts, washers and lock washers that I needed to complete the extruder filament pump.  The compression springs that you see in the picture are adequate and the extruder pumps ABS at 5v.  I was also able to acquire a spring scale like Nophead uses at the hardware store as well, so I should be able to get an idea of how well it pulls very soon.