Proposal for drip-free dual extruder system

[Tim]

I think that those of us who have beta tested the dual extruder system agree that it (and basically every other dual extruder system, with the exception of the new E3D "cyclops", and that has its own limitations) has a substantial drawback in being very drip-prone. Software (namely Cura) that can generate drip shields and wipe towers can help, but these additional structures use additional plastic, increase the risk of print failure, and do not work for every situation.

I have a proposal for a mounting system that I think will be drip-free, and which uses no additional motors or electronics. See the attached diagram. The key point of the system is that the extruder motors and hot ends are in a separate mount that rotates on a hinge that is a bit above the motors. The motor mount is held between this hinge and a gear rack passing under the stepper motor gears. The motor mount can swing from side to side. When it swings left, it reaches the end when the right extruder is pointed straight down. In this position, the right extruder is "active", and the left extruder is "parked". The parked extruder's nozzle rests on a wipe plate that prevents it from oozing while parked. When the motor mount swings right, it reaches the end when the left extruder is pointed straight down, and the left extruder is "active" and the right extruder is "parked".

The trick that makes the system work is that the gear rack is designed such that only the parked extruder's stepper motor engages the gears. The active extruder's stepper motor can turn freely for normal extrusion and retraction. When it is time to change extruders, the following things happen (which are coded entirely in g-code): (1) The active extruder retracts a small amount of filament (say, 2mm or so) as it would normally do. (2) The parked extruder extrudes about 10mm of filament. The act of extruding swings the whole motor mount, because the parked extruder's stepper motor is engaging the gear. This brings the extruder out of park. (3) Both extruders run at the same rate, with the (previously) parked extruder extruding an additional 10mm of filament, while the (previously) active extruder retracts an additional 10mm of filament. In the middle of this action, both extruders are briefly engaging the gear rack. At the end of it, the previously parked extruder is free of the gear, and the previously active extruder is engaging the gear. (4) The previously active extruder retracts a final 10mm of filament. This brings it into the parked position and puts the other extruder into the active position. (5) The newly active extruder extrudes 2mm (or so) as it would normally do to get started again.

The 20mm or so retraction is large but not so large as to be ridiculous; as it is, I'm regularly setting the retractions up into the 10+ mm range to keep the thing from dripping all over the place (which it does, anyway).

This system has the additional advantage of keeping the unused extruder out of the way of the plate where it won't be in the way when only one extruder is being used.

I can't think of any disadvantages that are not already present in the existing dual extruder setup (e.g, the right extruder can't move all the way to the left side of the plate). The only unknown is that I'm not sure how well the stepper motor gear can be made to engage the gear rack; I think the position of all the gear teeth should be known to within the precision necessary to keep the gears from binding.

Any comments would be appreciated.

Note that I don't actually have the technical expertise (or equipment) to build this thing. . .

[jsc]

Here's a similar rocker system that somebody built: http://forums.reprap.org/read.php?70,196591

I think something else that might help is to drop the idle extruder temperature way down and use a fixed wiper, but waiting for all the temperature changes will slow your printing speed way down.

[jimc]

i saw a youtube video prob 8 mos ago of someone would crudely built this. i dont think it was the one in jin's link but who knows and i dont remember 100%. that vid was the last i ever saw of it. there was another one about 6 weeks ago on kickstarter but it didnt cap off the unused nozzle. not sure what the advantage was and i cant remember the name of it. i know when it popped up on kickstarter everybody was making a stink that it invaded stratasys patents somehow. shortly after it disappeared. not sure if if they took it down or if i just cant find it anymore. in any case though tim's idea is what is really needed. im all for it

[pyronaught]

Didn't Stratasys take out something like 300 patents so that pretty much anything anyone does can violate one? That company just annoys me. I was watching a documentary on the development of consumer 3d printers that showed Stratasys harassing the Formlab guys with lawsuits. The CEO of that company just came across as a douche.

[Tim]

Good point about the patents. . . made me think to put my name and a date on the figure, just in case.

Also I realized that having the clamp for the hot ends bend in the middle was just silly, as it's a lot simpler just to drill holes at an angle.

I looked up a number of the Stratasys patents, but they don't have any with obvious titles such as "dual extruder system" (in fact, the word "dual" does not appear in any of their patent titles). Which means that one would have to work through all the claims of each one to figure out if anything is relevant. I have yet to see any other idea that makes use of the same extruding/retracting stepper motors to move the nozzle into and out of position. All other designs I've seen have a bunch of extra (moving) parts. If this idea is workable, its beauty lies in its simplicity.

[Tim]

I think something else that might help is to drop the idle extruder temperature way down and use a fixed wiper, but waiting for all the temperature changes will slow your printing speed way down.

It has been pointed out many times that we need both hardware and software/firmware improvements to make dual extrusion practical. There are, in fact, many cases in which one extruder or the other will be off for a long time, such as if you are printing supports that are just around the bottom of the print, or finishing a part with lettering or a design in a different color on top, or something. In that case, the software should be able to figure out roughly when to start pre-heating the extruder so that it can be ready when it's needed, or to shut the extruder down and let it cool when it is no longer needed. Any cooling at all will make the plastic more viscous and reduce the amount of drip, so as long as the software can predict the heating/cooling ramp (which is consistent enough that like the PID settings, it can be done once with an automatic calibration), the heating/cooling can be done without adding any time to the print at all.

[pyronaught]

Personally I'd like to see a dual extruder design where the extruders are oriented front to back rather than side by side, since I don't want to lose any print volume in the X direction. Since the M2 already has the bed longer in the Y direction there is more print volume to sacrifice on that axis. Mechanically it is more difficult to design a dual extruder head like that though, which is probably why all the machines have them side by side.

[Tim]

Here's a similar rocker system that somebody built: http://forums.reprap.org/read.php?70,196591

I think the one Jim is referring to might be the D3D extruder, which is a rocker system but which uses a single stepper motor. It looks like it does not use a servo. The one you posted a link to is very similar to mine, except that it requires an additional servo. However, looking at their video made me realize that I don't need to move the heads so far that one moves into the position of the other; they just need to move far enough to get off the bed and onto a wipe plate. The rest of the distance between the two can be taken care of by the X axis stepper motor and software in the usual manner, by specifying the offset between the two. So the total retraction/extrusion length for each motor doesn't need to be more than a few mm.

Thanks for the link. . .

[jimc]

i just had to look that up tim. thats not the one. the one i saw the kickstarter for was a rocker type system. it was very narrow and didnt rock back and forth very far like yours. it had a nice black cover over the hot ends so you only saw the tip of the nozzle. wish i could find that. your right though. it doesnt need to move far. shorter throw is a quicker switch as well.

[jimc]

hot damn!! i found it. in the video they have it on an m2


http://3dprint.com/22573/praxis3d-x-truder-kickstarter/