There have been a few people that have gone with dual bowden extruders, but most people wanted direct drive, which because of the weight of the motor, meant only a single extruder was practical. But with the release of the E3D Titan extruder, I think dual direct extruders are within reach. I built a carriage for the Titan and used a short body motor for it, which has more than enough power. I have been thinking of modifying the design to accept two extruders, but since I only have the one, I wouldn't be able to test it out.

 

So it looks like I had my temp too high when I tried printing with NinjaFlex earlier. Set my speed to 30mm/sec, nozzle to 220, bed to 50 and got this...



This is a pad for the foot on my printer. Came out perfect. So I think printing in flexible filament, at least NinjaFlex, is possible.

 

Looks like you got it dialed in, but I rechecked my s3d settings (again, for my Replicator1, using Sailfish) :
* Bed temp off, print on blue painters tape
* Extruder @ 240
* speed 30mm/sec
* Filament cooling fan off.
* 2 shells, 20% infill most of the time.
* Retract 2mm @ 20mm/sec
* Wipe nozzle 2mm.
* Only retract when crossing open spaces.
* Force retraction between layers.
* Perform retraction during wipe
* Only perform wipe for outer-most permitter.
All that retraction\wiping stuff has really helped reduce stringing on more complex prints.

 

Both versions are back up on Thingiverse now:
C-Bot Volcano Filament Cooler by AK_Eric

 

The temp difference could just be in how the hot ends are built and report temp. With my E3D's I use thermal epoxy to set the thermistor in place. I have seen a difference in temperature readings doing this way.

I had retraction at 1.5mm so that is close.

I have the Only retract crossing open spaces and force on layer change. I've never been sure what exactly the wiping does. Might turn it on when I print the next foot pad.

 

What is does is keep the extruder moving for a distance after extrusion stops: If there is any other filament in the nozzle during the retraction, it's pulled out by being 'wiped' across the toolpath in front of it by the given # of mm (like toothpaste on a toothbrush). I mainly use this to good effect when using really big Volcano nozzles that like to drool a lot. But it proved helpful when using ninjaflex as well, to reduce stringing etc.
The only downside I've seen is that it can slow down print times slightly, due to the extra 'wipe' motion, which is slower than printing without the wipe.

 

Went looking for a model to try printing with NinjaFlex and came across your squishBall. So set that up and started printing. Thought I had turned on all the wiping, but forgot to enable it, so lots of strings. Will print it again with wiping on so that will give me a good before and after look.

 

I'll second the 'no buildtak' warning. On another machine about a year ago I tried printing Ninjaflex without much success due primarily to the extruder setup. I was able to print a successful 2cm cube, but was unable to get it off of the bed. Ended up cutting it away with a razor blade, but could not get the last bit off of the bed. Finally had to use acetone to get it off and it ended up taking the surface of the buildtak along with it.

That said, on this new machine using the direct drive setup it works like a dream. Just have to make sure the PTFE tube coming from the cold end terminates in a wedge fitting snugly between the filament drive and idler. Give it any wiggle room and it'l go where it wants. Love it.

 

Missed that part about the Buildtak warning, but haven't had any problems with it on mine. I did have some ninjaflex not wanting to come off, but once I heated the bed up to 90, it came right off. A simple wipe with alcohol and the bed is fine.

 

This is a good one to print. Not too hard to pull off and it really showcases the properties of the filament. Printed a few of these and the kids stole them.

OpenBionics Ninjaflex Hand

 

Nice. I likely compounded the issue by squishing the first layer too much. Wasn't pretty!

 

Just out of curiosity, can ninjaflex be handled by the "stock" c-bot direct extruder?

 

I am using the direct drive that posted here a while back (by, I believe, Carl) and it works fantastically for Ninjaflex and everything else I have thrown at it.

 

Good to know. Thanks

 

Actually that looks pretty clean for OpenSCAD. To be honest though I'm not sure exactly what "good" OpenSCAD code would look like. I'm really hoping someday a more robust modeling language can gain traction.

You definitely shouldn't be embarrassed to post this code though -- it's much better than most I've seen.

 

I'd gladly take your help with the generator if you want to poke around and add stuff. Just need git, some acceptance that the code is not pretty and a little patience.

I have the the inductive sensor and servo probe moved to the front carriage now, as well as moved the hot end to the side. It has already shrunk the width by a solid 1.5 cm.

Started working on the fan duct again last night, but am struggling with the best approach. Might just leave it and move on to adding the direct drives and bl touch probe.

 

So I got my first successful print last night! After switching to grid-based leveling and making a better mount for my inductive probe I was able to get it to actually print *on* the bed! Unfortunately it was a bit trapezoidal with little twists on the end. I'm guessing it was a belt tension problem, but didn't have time to try again.

Is it normal for it to shake if not on a perfectly level surface? I'm worried that either a.) it's not square enough, or b.) it's not rigid enough. I printed most of the brackets with eSun PETG, which is *really* springy.

 

@Chris Roadfeldt
One option to consider that worked really well for me -- make your inductive probe attach to the hotend collar. I moved the holes on your inductive sensor holder to do this mainly so I could retrofit it to the "standard" direct drive hot end mount, but I found it was also nice because I only had a Y offset, and no X offset.

I have planned on submitting a few pull requests at some point.. But right now I'm focused on getting my printer built

 

Right on, I look forward to seeing that.

I did a similar mod to mine when I built it a few weeks back as well. I moved the probe to the right of the hot end. Had to mod the mount to do so, but it fits great. I believe I have 2 or 3 mm of Y and about 20mm of X offset. Much better than the 65 of Y and 25 of X with the probe on the back end.

I will update git with what I have. Heads up that I am verifying the fan duct outlet positioning. Something isn't right in the visualization and I'm trying to verify if it's my code or the volcano model I am using.

 

Anyone with an E3d V6 w/ volcano, can you verify the distance from the top of your v6, the aluminum top, not the ptfe tube holder to the tip of the nozzle on the volcano? I would but mine is in the middle of a 24 hour print...

 

I've done a preliminary version of the E3D Titan Dual carriage. I haven't printed it to test and I don't have another Titan to use for a test fit. I still need to make the back. If anyone is interested in testing this, and have two Titans, I can send you the STL files. The second extruder adds and additional 60mm to the carriage.

 

I had thought about that option. The alignment, as far as printing, shouldn't be much harder. The nozzles could be lined up so there is only a Y axis offset instead of the typical Z. Fitting a fan would be the bigger issue. A way to mount it on the side would probably be an option. A modified version of AK Eric's fan shroud that would cover both hot ends.

 

I've done this with my probe and with the default frame size, you will loose some printable Y dimensions. I believe it was around 30mm of Y loss.

Another option is a hybrid setup. Keep both hot ends on the same carriage side, run one extruder on the carriage and then one in a bowden setup.

 

Right now I have more X that I can give up than Y. I'll probably leave this on the back burner for awhile. I have another printer I am working on that needs my attention.

 

I've noticed I have nearly 100mm extra Y space and at least 50mm extra X space.. I'm wondering if this is common? I used the recommended extrusion, and I even have a build plate 1" larger in both directions than I built for. I have the build plate pushed back to reduce weight on the front, but there's a ton of extra room there.

 

Tried with the wiping on and it was just as stringy as with it off. Need to play around with it some more. Need a smaller model to test with that doesn't take so long to print.

 

Did you set it to retract while wiping?

 

I have about an extra 30mm on my X. When Y is at zero, the nozzle is at the front edge of the plate. When all the way back, there is still about 20mm, but the side carriages are hitting the end. I have exactly 200mm on the Y.

 

Yeap. Need to play around more with the settings and try to get it dialed in better. But at least I am getting a good bead when extruding.