I’d love to get some opinions on this… One of the members at @Milwaukee_Makerspace has been working on an entirely new (as far as we know) method of moving filament.
Any thoughts on if this will work well?
http://milwaukeemakerspace.org/2014/03/snakebite-extruder-testing/
This looks very promising
Interesting. I’m not sure twisting will be eliminated though. The first nut the filament passes through will basically thread it, and the second will be trying to cross-thread it. They will exert different torques on the filament. Just how different will be determined by testing.
I don’t think it should be necessary to have this much grip though. Are there any materials that behave well on the printing side of the nozzle with this much back pressure?
A big question is how it handles variations in filament diameter.
I just worry about the speed. My retractions are already set about as fast as the direct drive motors will go.
@Dale_Dunn Since he mentions a “left handed tap” in his parts, I assume that one of those nuts may be tapped backwards.
I do wonder if both nuts even need to turn all the time, which would simplify the design. eg: Have a semi-fixed nut on entry and a moving nut at the bottom. The semi-fixed nut gets turned to “prime” the extruder (till it gets to the nut) by hand, then gets locked in place with an arm for normal operation. Not sure if it’ll work as well but it’s worth trying and could cut it down to 1 drive gear (could even mount the motor at right angles to the current placement, use herringbone gears to cut down on slop, etc).
Also, if it has the torque that he’s saying it has, you could probably use a NEMA14 motor to make it even smaller. Can we get NEMA14’s with bigger step angles than 1.8 deg that have the same torque, or does torque drop off as we get a bigger step angle? This would improve the speed, which might be an issue for retraction.
The early RepRap DC motor extruders relied on similar worm-gear theory but were made obsolete by extruders based on hobbed bolts.
Now, the article starts out with One of the biggest problems with FDM 3D printing is hot-end jamming. , which i’ll have to disagree with. Using a wade’s style extruder and a homemade hobbed bolt, i haven’t had a single extruder-related jam in the last 20kg of plastic i’ve put through it. Once you figure out the optimal tension on the idler, you’re pretty much set for life. Someone once did a couple of tests with a wade’s and found out that it can produce upwards of 200N of force.
But i do agree that the pinch wheel extruder isn’t all that ideal - for once, it could use a bigger contact area with the filament and better self-healing capabilities. It’ll be interesting to see how the snakebite extruder does in real use. It certainly looks interesting.
@Thomas_Sanladerer It’s definitely got a lot more contact area with the filament compared to the old worm gear methods
Also another thought: Because it pushes the filament through the nuts, this means if you have variable width filament (and well, it all varies, it’s just a matter of how much) then you may have issues. For generally consistent filament it should be OK, but since people have been working on methods to measure the dia of the filament as it goes in, it is worth noting. That said, even various manufacturers can have differences between rolls.
@Mark_Rehorst , one issue mentioned was the variance in filament. I think you said you were also working on a way to measure the filament, and I assume this would perform a stop print if filament too big showed up. I had some filament that went from 2.8mm to 3.2mm and it jammed things up, but luckily my gear just chewed it up and my print was ruined, but no damage was done to the machine.
Speed is also critical for ooze-free retracts. Would the nuts grind up the filament on prints with tiny details and lots of retracts?