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 Toolchanger System V2 (E3D-Plate Compatible)

by poofjunior Aug 16, 2019
Download All Files

Thing Apps Enabled

Please Login to Comment


Your work is fantastic, congratulations.

One detail I'm curious about is why you traded the zesty-nimble extruder for a bond tech ??

I ask this because I was thinking of including a zesty-nimble in my project and you went through the experience that I don't have.

A few comments, I found some PTFE tubing on Amazon that has a 1mm ID and a 2mm OD that I thought I might try in place of the spring guide. Second comment is on the spring for the lock actuator, do you think that a tension spring with a .2mm wire diameter and a 2mm OD will work if cut a little shorter to get close to the same tension?
Those are the last two pieces of my puzzle that I have to get working, everything else I have adapted to work with my printer. Can't wait to get this thing running test!!

Hey Poofjunior,
Great project.
I read in your FAQ about a bowden PTFE tube being unsuitable and a spring cable is better, however have you considered bicycle cables? From my quick look they are cheaper, basicaly the same thing and more available.
Anyway, good luck with your project.

Thanks BluntP!

Yes, back when I was playing with animatronics a bit more, I experimented with Bicycle, Bowden, and other various tube housings to try to get something stiff in the axial direction, but flexible in other directions. Bicycle brake cable is just far too stiff for a reliable actuation by small motors. (It turns out that our grip strength for squeezing bike brakes is actually quite strong!)

In the end, I settled on spring guide to build this:

and there's really no comparison; it's just soo much better than anything else. I know it's yet-another vendor to purchase from, but at $1.75ish per foot when bought in 10-foot increments, it's darn cheap. Heck, it might work so well, you might try using it for other applications like this:


Anywho, I hope this helps.

Marionette Probe

Thanks for the reply, I should have figured you already tried that.
Good to see you found the perfect material for cheap.
Again Best of success!

Many thanks! I'll try to make this clearer in the notes. Apart from that, spring guide is really cool stuff! I'd love to see others do some fun projects with it either in or outside of the 3D printing space.

Stoked to see the progress on this! My machine design incorporates the V1 of your tool changer mechanism, but I will be modifying it to incorporate the improvements V2 brings. Aside from having compatibility with E3D's tool plates, what would you say are the major improvements?

I can see that there is now a limit switch integrated (where I was planning on using a BLTouch; I have to say your solution is sleeker!); it seems like overall the carriage is thicker as well, so I would assume that greater rigidity has been achieved.

Many thanks! There are quite a few updates since V1. The changes to the carriage are mostly size-related to accommodate E3D tools and simply bigger tools in general. At the end of the day it's a better standard size to work with since I hope (and partly expect) E3D will sell their tool plates. (They are primarily an original-equipment manufacturer for 3D printer parts after all!) The parking post has also changed as a result, but I have also added screws to adjust the horizontal and vertical position of the post. That change is an incredible pinch-point-reliever compared to V1 when it comes to dialing in the software-and-hardware location of the parked tool.

The locking mechanism is now using an all-mechanical variant as opposed to the TMC2660-stall-detection method on V1. This system is external to the carriage and can be used interchangeably on V1 or V2. It is far more reliable and requires basically no tuning to get it working. It just takes more parts and time to assemble, but the up-front cost is totally worth the gain in reliability.

I admit that I shamelessly adopted the z-leveling limit switch location from E3D! It's a fantastic solution.

I am building one of these. Please let me know when more is available.

Hey, great question! The short answer is yes, with some redesign effort, a 59Ncm should be enough to lock tools. I actually started with a Nema17 motor, but I moved away from it because of heat issues. The motor was at max torque holding tools, and the heat was causing the mating plate to warp. The nema14 with gear reduction gets about the same amount of torque, but it doesnt heat up anywhere near as much as the Nema17 variant. Unlike the original version, I can run this version all day and night without warping issues.

Id call this version of the design a proof-of-concept thats not cost optimized just yet. When I get more time, Ill try to reimagine a cheaper solution. Or some folks in the community would be welcome to tackle this one, although i encourage folks to thoroughly test their solutions with some cycle tests before proposing models as improvements. The tricky bit about toolchangers is that they cant work 99% of the the time. they need to work beyond 99.9% of the time. Getting this system to run reliably over thousands of changes is what kept me from posting partial solutions that "mostly" worked for months. While a geared Nema14 motor might cost a bit more, it's totally better than dropping a tool and damaging a nozzle and heatbreak every couple hundred changes.

I hope this helps!

There is also the STEPPERONLINE 17HS19-1684S-PG5 (48mm stepper body length) and 17HS15-1684S-PG5 (40mm stepper body length) utilizing NEMA17 motors and other than higher inductance on the shorter-bodied model and higher holding torque on the longer-bodied model, they have identical electrical specifications. Currently in the US on Amazon they are $45 and $42 respectively, they take a slight bit higher current (1.68A max) and utilize larger diameter gearboxes and 8mm shafts. Is there enough room to allow for such modifications without significant redesign? I'd imagine offsets would have to be made to accommodate the stepper mount and shaft alignment if the gearboxes are different diameters between NEMA17 and NEMA14. Ideally the best would be if the mount could take either option with an optional printable sleeve or something, but the shaft attachment would also need optional prints for the different shaft sizes. I'm curious to know your thoughts on alternate options. I'll likely order the larger NEMA17 motor this weekend and attempt to replicate your relevant parts in OpenSCAD so they may be parameterized easily. The main reason for me going with a geared NEMA17 at this point is because the NEMA14 models ship from the UK and NEMA17's are one-day shipping on Amazon to the US.

If you're willing to sacrifice a bushing (very doable if you drill out the hole to 8.1mm to reduce friction), then yes, you could rework this setup to use an 8mm motor shaft and NEMA17 motor. You would need to rework the fixed pulley bore size, the floating pulley bore size, and the clamping plate feature.

Minimum holding torque requirements are about 65ish Ncm. I've found that, in practice, using 59 Ncm works inconsistently and is more sensitive to rough spots on wedges in the tool plate when it tries to lock it. Using a motor with higher torque also helps keep the part cool and prevent any warping. This is why I opted for a planetary setup: less overall current to achieve a certain torque rating.

Sounds good. I ordered the NEMA17 5.1:1 stepper, will build out the models in OpenSCAD this coming week. The code may not be pretty, but it should look pretty close to identical. If I have trouble recreating the parts, do you mind if I ask you for dimensions/radii for the fixed and floating pulleys' curved features? I'll be measuring via calipers against printed parts, as I likely can't run the CAD software you used. (Running Linux and no Windows machines in the house)

Can do; just keep me updated. As a heads-up, I've been trying OnShape (cloud-based, free public files) recently, and it's actually quite good. For light editing, it could be much easier than recreating the part. Bonus: I can post up the original Solidworks files, and you could import them natively.

Feel free to keep up this chat via DM.

So I see this requires a 5:1 planetary reduction NEMA14 stepper. How much torque is required for the locking mechanism? Is it possible to re-design for a high-torque non-geared NEMA17 motor to minimize cost? 59Ncm NEMA17 motors are super cheap, especially if bought in bulk for a new machine. I realize it's a difference of maybe $10-30 US, just curious about your thoughts on the matter as far as practicality. Also curious why you chose a NEMA14 size as well, I don't see those too often.

I have no E3D tool plates.so.... thx!

Thank you for posting this.

Why is there a need for a planetary stepper?

Hey thanks for posting and for include STEP files!!