MakerBot Print is our newest print-prepration software, which supports native CAD files and STL assemblies,
allows you to interact with all your printers via the Cloud, and many other exciting new features.

Download Now


Cable-Driven Tool Changer System

by poofjunior Jan 19, 2019
Download All Files

Thing Apps Enabled

Please Login to Comment

Could you point me in the right direction on the firmware setup for this? I assume you are using a duet?

Very nice design! I will start building one this weekend.

I love this! Well done. Is there a particular reason you went with 10mm balls? Just brainstorming ways to reduce weight :)

Excellent, question do you have the pen holder flexure part ???

Which parts are used for the cam lock? Thanks

This is a great idea! I've started adapting it to my printer using an MGN12 linear rail. Instead of a remotely driven locking mechanism, I chose to go with a small stepper (NEMA8) mounted on the carriage itself and will use stall detection (hopefully). Still a work in progress, but it's getting there! I'll probably post it as a remix when I get around to it.

why not use a servo? wouldn't it be lighter? and instead of using the stepper's stall to control lock pressure. maybe a semi flexible locking pin. just thinking out aloud :)

Eh, just figured I'd give it a try. I might end up going with a servo instead if I can figure out a way to get a linkage or gearing done in the space I have for it.

I also have been working to adapt to a 12mm rail. This is how I know that the 2mm PTFE tubing doesn't work.

@poofjunior Some of your BOM links are the wrong size; for instance

4x M2 screws, 6mm length 91292A107 - is 18-8 Stainless Steel Socket Head Screw, M4 x 0.7 mm Thread, 6 mm Long.

Did you mean 91292A831?

Thanks for clearing this up. I didn't auto-generate the BOM from the CAD files, but I will for all future designs so this doesn't happen again.

Great Idea! I think I'm going to give this a try for my tool-change 3DP design.

Out of curiosity, did you ever investigate using a single cable like a bike derailleur?

I know you cant push a rope.. but I wonder if a strong enough spring assist would make it work well.

Will ptfe work? 1mm od 0.6mm id 5 meter on amazon for $6 as opposed to $17

There is too much give in the PTFE tubing for this to work with it.

Great solution! Thank you for the hard work you put into this. I think i'm gonna try and adapt this for my printer build.

Hi, I love your X-gantry, looks like carbon Fiber....Where did you source it ? if you don't mind sharing :)


Thanks, jmpaquin!

This is my take on a CoreXY. It seems to be working pretty well, but it has some kinks that make assembly a bit tricky. I'm planning to release an updated version, so stay tuned! As far as where I got the Carbon Fiber piece, I'm getting my designs cut by Nick at CNCMadness.com. He generally does quadcopter frames, but there's no reason he can't do other things. (Simply send over a dxf.) My crossbar was about $35 and came back to me within a week.

Thanks for the reference.

Where are the files for the E3D V6 mount shown in the video?

At long last, a draft is up!

Cable-Driven Tool Changer Zesty Nimble Head (Draft!)

Ok, this is going to take a bit longer than I expected. I've got a full-on printer in the works with this design, so I'll try to release them as a set, but it's going to be a bit farther into the future.

Thanks for asking. I'm making some minor modifications to the V6 mount. Doing a test print by tomorrow, so it should be up within 2 days. As a heads-up: I'm using the new-ish (and hopefully here-to-stay) threaded hotend from E3D.

Looks like SW2019, is it?

Very good eye! You are correct!

Very nice engineering. You are way ahead of my tool changer efforts. Unrelated but may I ask where you sourced the carbon fiber plate for your cross rail support. Is it 6mm thick?

Many thanks! I got my carbon-fiber plate from Nick at CNCMadness.com. He's got a 12"x24" routing table and the right tools to cut through composites. Simply send him a dxf, and he'll cut it out for ya. The price is pretty good too! (I think mine was around $50, which is fantastic since I don't have any tooling to do this sort of a job where I live.) My crossbar plate is 4mm, but I'm thinking about bumping up the gantry size and using a 6mm plate on a larger build.

Thanks for the quick reply. Once I get my tool changer working (or switch to yours) I want start assembling my large 7 tool position bot with a 550mm span on the x and 12mm rail. 6mm carbon seems to be the way to go for such a long gantry . I'll check out CNCMadness.com. It sounds like his pricing is as good as I have been able to find just for raw plate. I have followed all your social posts and can't wait to see video of your bot printing. Is that a DIY cable drive extruder I see in your Jan 10 tweet? Thanks for sharing and documenting your efforts. Keep up the great work :)

Hehe, a 550mm span version is in the works for me too, although it'll be a couple months before it's mature. I'm excited for ya! Having cut carbon fiber as a resource is super useful. I do have a cable drive extruder, but it's the Zesty nimble flavor (still untested). I figured I can't redo everything from scratch all at once ;)

I am for sure making one of these for my future build,very cool stuff, thank you for your efforts. I'm curious what was the design reasoning for using a pancake stepper for this vs a something like a servo that only rotates 180 degrees? I know you plan on adding stall detection which i don't think you can do on a servo. Thanks again for sharing your work!

Hi NathanMichaelMoore,

Both, when done correctly, can get excellent results.

When a servo does the locking, the lock stopping point is position based. Over time, the two contacting surfaces can wear out, and that position would need to be adjusted. In that sort of system, you can springload the interface, applying a preload force on the two contacting surfaces. (I think this is the mechanism that E3D uses.)

When a stepper with stall-detection does the locking, the lock stopping point is torque based. In this case, even if the contacting materials wear out over time, the stepper would, in theory, always lock down on the tool with the same amount of preload force.

The reason I opted for the second is because of the materials. Because my locking system is almost entirely 3d printed, I need to assume that the contact surfaces will wear out over time. (The actual wedge-lock on each tool is a consumable.) For that reason, having something lock down with a set amount of torque will still yield reliable repositioning as the surface rubs away.

I hope this helps--and feel free to ask more questions if it's not clear.

I do have another question for you, is there a special reason for using delrin for shoulder screw reinforcement vs just making it part of the base geometry like shown in the photo? This way it reduces the BOM a little bit with less parts. Maybe you tested this and it's just too weak as a printed part?

No special reason for those being a different material. I've made 3D printed variants that work just fine too. As for integrating it into one part, It's possible, but I haven't given it a try. (I don't think there's anything too troublesome about integrating them, so if you get some nice prints, let me know!)

One of my original design goals was to minimize "fabrication process complexity." To that extent, none of the 3D printed parts need any support material as long as they're printed in the given orientations. I think that idea bled into having that feature have 2 additional parts.

This was the exact technical answer I was looking for, thank you.

that’s what I’m looking for
Can you tell about how you control the motor to lock the tool holder?

Sure! My plan is to drive the stepper motor with a TMC2660 (TMC2130 also ok) driver. This chip comes from a family of chips that have a stall-detection feature which can sense when the motor has stopped rotating. The Prusa Mk3 actually uses this feature for homing each axis. My plan is to detect tool locking and unlocking with this stall-detect feature. In my system, rotation has 2 locations where stalls have been added. The first is on the pulley side. The pulley itself has a hard-stop that it will bump when the pulley is in the unlock position. The second is present in the locking feature. The lock twists into a wedge feature until the motor stalls.

Currently, I'm adding this into Smoothieware as I write this, but I wanted to release the CADs first just in case anyone else was excited to get a head-start into making parts.

OMG i love this :-D
i wanna definitely build this and ill give you a nice tip then :-D

which printer build do you recommend for this? a hypercube?

Thanks a bunch lordgort!

Right now, it's tuned for the printer I'm building, which just has an MGN9H rail turned upside-down. Anything with that MGN9H profile (M3, 15x16mm hole pattern) should work.

That said, do you have recommendations on popular printers that I should adapt this design for? No promises, but it might be a good use of my time to re-imagine this build for a different platform to encourage folks to use it.

Just design your own! You do a great job on designing this tool changer so why not a whole printer ;)

Aw thanks! It's in the works! I'm having some issues with making my bed-leveling design "easily reproducible" without machine shop tools, so that's partly what's holding me back.

Great plan. I have been working on the idea of a multiple head tool changer for over a year now and have seen several great ideas. I like your the best so far as it incorporates several design ideas I had already been playing with. I was also inspired by E3D's toolchanger system, but the cost and unavailability right now make it out of my reach. I have a giant hypercube 400x400x450 ready for cutting and assembly. I also am in mid design of a giant printeRbelt conveyor printer that is 500mm wide 300mm high and as long as I want to make the part. I want to put multiple tool heads on both of these. Luckily I have all of the aluminum extrusion already but the electronics are holding me back cost wise. I look forward to further updates. Keep up the great work. I will start getting parts and prints done from your article. Thanks again.

Ed Stoddard

Thanks, edstoddard! Yes, I can see you adapting this one to either of your printers. Currently, it's nicely modeled to be mounted under an MGN9H rail, but it shouldn't be too tricky to remix it for a different form factor. Everything here is an off-the-shelf part save the locking mechanism, which has one machined rod. I'm willing to bet you could print a solid version and get rolling though. (If you start making progress on this build, let me know, and I may be able to send one of those T-Locks your way.)

Cheers--and thanks for the kind words.

I signed into this Painfully slow website, just so i could say
Awesome Design

Thank-you for your time and effort

Many thanks, Wallys3D! Id love to see folks give it a go and tell me what they think.