Taurus X axis for Prusa MK3 and MK3s

by jltx Dec 24, 2018
Download All Files

Thing Apps Enabled

Please Login to Comment

What kind of support settings are best for the main parts? I did no supports except for one overhang, see attached. Is this going to work?

I printed with no supports, but what you have looks ok to me.

First: Thank you for your work. Did the Y mod and it works great! Now going for this.

After printing the stock version, I noticed that you did some modifications especially to the upper part of the idler in the non-stock version, but not the stock version (extra printable part mcap is still built-in with stock).

How different is the stock from non-stock? Could I just use the better printable non-stock version, even I am running the stock extruder? what are the disadvantages?

Thanks again!

Shoot. Looks like I accidentally included the Mcap in the stl. If you split objects in slicer does it separate? I'll have to fix that upload. Thanks for pointing out.

The stock is very different because I have to make room for the fan and PINDA holder on each side. This design started with and is better suited to skelestruder. I need to release the step so people can mod further. I have not kept up with Prusa's extruder changes in case something moved.

Any update on the stock self-test fail? I'm having lots of issues with the skelestruder and am strongly considering going back to the stock idler. Main reason I switched was because of the crashing.

Can you clarify? I’m not aware of self test issue when using with skelestruder. Why are you running self test and what is the fail?

No my issue was that the stock extruder was failing the self-test with the Taurus X-ends. Running the self-test is the only way to update belt status. So I switched to the skelestruder. But either something went wrong during the printing of the parts or I assembled something wrong and now it's making this weird resonance that I think is causing my prints to fail. I did some measuring with my calipers and the spider is about 1 mm shorter than it should be; from above, it looks like it's bending.
With all of this I've never been so mad at my printer for not working like it should.

Well, let's get this sorted out. First, self test is not required and most people agree the belt status is bogus. So don't stress about that. Just use the pinch test to see that the belts have enough tension. There is a wide range where things still print fine. Don't over tighten so it's like a guitar string.

What sort of resonance are you seeing? What do you mean by the print fails?

Video attached.

The video never got posted for some reason.

Yeah, I noticed that. Sorry. Apparently uploading from my phone didn't work well.
In any case, I figured out what the problem was after some disassembly. Turns out my wheel is not entirely even and was rubbing against the ecage-front. This had left a sizable build up of material. After I cleared that, it worked like a dream, mostly.

I seem to be getting mixed messages about the belt status. Regardless, I upgraded to the skelestruder.

I’m at work now and can’t get you a video of the resonance but I can tell you that the whole extruder motor seems to vibrate a lot. As I previously stated, the spider seems to be the wrong size and I’m certain that’s a contributing factor as the whole spider assembly and fbs arm from ecage front seems bent. I suspect that the resonance was what caused the filament to snap off at the bondtech gears during my last print. I got home and Godzilla was only partially printed with the extruder printing air. I took out the Bowdentube and saw that the tip of thefilament was right at the bondtechs but not actually moving. See the attached photo of infinite sadness.

Hello - is this compatible with the MMU2 extruder?

X axis has no interaction with MMU2. So yes. I am using it.

Can this be used on MK2S?

Off the top of my head I think so. Only concern may be if mk2 x carriage belt path is offset too much.

So if I have a stock MK3, an MK3S upgrade kit (not installed) and want to incorporate this mod, how do I do so? Upgrade to MK3S first? Do them both together? Is there still an issue with the "stock" extruder?

This is a very awesome build. I had it for 3 weeks now. Unfortunately the Z axis always crashes. My left z rod is so hard to turn manually but the right side was a breeze and moves smoothly.

That doesn’t sound good. What does Z crash mean? Any modifications to the Z axis?

You mean turning Z lead screw? Or sliding along smooth Z rod?

Comments deleted.

I upgraded all axis to Taurus along with skelestruder. I am having auto home issues. It does not want to return all the way to the left. It stops between 5-10 mm depending on belt tension. If I get any looser on the belt it skips teeth and I can move by hand with it on.

Can you see if anything is catching, like the noctua wire, that would trigger the stallguard? Or maybe a burr on the print catch the fan casing? Can you push it all the way left by hand with the stepper off? If so try returning belt to normal tension and when homing apply light pressure by hand to see if it will fully home. Then use move axis in menu to move it in and out and see if you can find what is binding. Make sure you are using the bottom belt channel on carriage.

Ok I think I have it under control. Something with the belts. With no belt it slides like butter. I feel like it’s the bearing in the idler and the belt doesn’t want to stay centered well. I would have rather used a metal toothed idler instead of the printed idler with bearing. Would be more stable I think. The two pieces of the printed idler want to separate constantly. But it’s homing now. Also the bearing on the x motor causes me to bind and not home all the way. I did some firmware changes on current and it seems to work atm. Will get back to you if I find anything else. I am currently underextruding with the skelestruder quite a bit but it’s not consistent. I calibrated the steps and it’s still inconsistent. Also have cold pulled and double checked nozzle. Extrudes fine on change and even on purge line but then becomes weak on first layer and final layers. Infill seems pretty good. Also getting inconsistencies in the walls. . Did you have to increase extrusion multiplier or adjust extrusion width to make up any differences from stock. Or increase flow?

How is the motor pulley binding?

No change to extrusion. What stepper are you using?

Comments deleted.

Are the problems for this build fixed? I want to do this mod to my printer but kinda hesitant now as i see unsolved problems(?) on the comments.

If you are running the stock extruder, you need to cut off a small bit of the idler side to clear the pinda. I am looking to make that better after I finish up some other work.

This mod only has issues with the stock extruder? I will be making your Skelestruder soon and then followed by this X axis mod. Will it be any issues?

That’s what I’m running.

After get hex nut in, i'm getting another problem.
Now Z endstop is not the Tarus X but the cable bundle.
I've printed the non stock version.
Also the right X endstop is pinda holder, is that normal?

It looks like your cabling needs to get tucked down closer to tail. Were you not seeing the same issue with the stock X ends? The dimension for Taurus is the same.

I redo the bundle and was it. It work now. Testing right now (calibration first)

Hex nut tolerance on idler and motor parts are undersized, hex nut won't fit, i end in use hammer to force push them but it crack by the sides, not because of hammer but hex push sides and break. I also tried a 3mm drill and unscrew position to file the slot, still hex can't fit. But otherwise square nuts slide very well on their slots, almost no push required to fit.
Can tolerances be increased?

Are you printing with 0.6 nozzle? have you checked your extrusion rate. They are snug to keep them from falling out but they should slide in. I press in with plier tips.

Yes 0.6mm nozzle, extrusion is calibrated, tried with PETG and ABS with same results. If i scale to 101% will do or will not be functional?

I manage to insert hex with help of a solder iron, for people who have trouble like me use a chissel tip and slow push hex to the place, don't hit the plastic, only the hex.
EDIT: Or use M2.5

I need to review the R3 extruder. I’m aware of clearance issue on idler side. You can just remove the bottom ridge below the pocket to allow pinda to move over further. Self test is not critical. You can fail that and still print fine. But check the homing is ok. I’m out for two weeks so can make any changes until then.

I see.
I’ll check the homing ok.

If I fail, I’ll be happy to go back to the stock X until you figure out the solution.

I changed 3 Axis to Taurus and every axis slides like on the ice btw ;)

Thanx for sharing great works.

Do you also use a 17HS19-2004S1 with the X axis like you did with Y?

No. I don't think that would be a good idea.

Due extra weight?

Correct. Y Bed has more mass + the print, so can use more power and motor is static. X is even lighter with skelestruder and motor moves with Z. So tempted to go smaller.

And what about the current value? Did you increase the current for Y? Or stock current will generate more torque with the new motor?

I have the bear extruder along with it's x carriage. I've printed the NON stock version of your x axis and it seems it's not compatible with bear x carriage (bummer :| ) so is the stock version compatible with bear x carriage? Or did I not mounted the NON stock version correctly?

It's not compatible as in the lower belt isn't where it supposed to be in the carriage... it needed to be higher ... i attached a photo of my problem.

This is not compatible with bear extruder, but most others including skelestruder. Use can use one of those with your best frame and these ends.

Looks like I can't have all the good things ... like I said bummer :) I will most probably print the skelestruder sooner than I thought.

I have a partial build (no motor or extruder yet, I'm upgrading to Skelestruder and Taurus y-axis at the same time, work in progress). Initial impression: pleasurable assembly, great design, feels stiffer than stock especially wrt rotation about the x-axis itself. Thanks for your excellent work on this - your designs exude technical mastery and aesthetic aplomb.

A couple of observations:

  1. The bearing clamps can't take more than about 0.5 turn tightening of the screw before the bearings start to bind. I've not mounted the motor yet, but I don't see how I'm going to be able to tighten the shared motor/bearing-clamp screw to suit both good fixing of the motor and suitable tightening of the bearing clamp. In general, I'm a bit concerned that such lightly-tightened screws might vibrate lose in use - perhaps use nylock nuts (would need bigger nut holes)? I'm using Misumi LMU8 bearings, which are the same dimensions as stock Prusa bearings.

  2. Regarding the smooth rod length adjusters: I had the opposite problem to the one imagined in the docs: my rods were moving outwards as I lower the x-axis. I checked the holes carefully, ensured full insertion, and disengaged the adjuster screws, but the assembled x-axis is still a little too wide. Given that the x-axis never goes too far down (to allow for extruder), the current misalignment is probably workable for the time being, but ideally I think the idler end holes could do with a bit more depth, and then use the adjuster screws to get perfect alignment. Note that I'm using your (jltx) fixed z mounts (https://www.thingiverse.com/thing:2779123).

  3. Regarding the BL adjusters: I had to use 20mm screws, 18mm were to short. If I use standard hex head screws, those screws became the contact point for z endstop against the top z mounts, which does not seem like a good idea to me. Instead I used low profile button head screws. I also used a light wave spring washer, lightly compressed. I would have used normal washers in addition to provide a better contact surface for the wave spring washers, but this raises the screw too much and risks becoming the z-stop again.
Prusa MK3 fixed Z mounts
by jltx


Bearing Clamps

I disassembled and checked everything. The upper (as printed) internal surface of all the bearing clamps had some roughness due to overhangs. It would appear that this was causing some localised pressure on the bearings when clamped. After filing down any roughness, the bearing clamp screws can now take a couple of turns without causing the bearings to bind. This feels a lot better. I suggest checking this carefully and assessing the bearing fit on each side by sliding up and down on z rods and gently tightening clamps prior to full assembly - it's much easier to assess one side at a time.

Smooth Rod Holes & Length Adjusters

As per Jon's suggestion, I checked the upper (as printed) surface of the smooth rod holes. There was some roughness from bridging. It was hard to tell, but perhaps it was worse on the idler end due to the incomplete bridging surface (hole for length adjuster screw). Perhaps it would be good to have one layer which is unbroken for the bridging, and then drill or push out the single layer that would be blocking the length adjuster screw hole. I used a 7mm wood drill (which has a mostly flat cutting surface) with a cordless drill to carefully abrade any protrusions. Measurements show that this relieved about 0.2mm on the idler end. This was enough to make the x-axis the right width.

Backlash Adjusters

See photos below. I decided to try putting washers between the BL flange and the end body, using enough washers so that screw holes would line up when the flange was in contact with the washers. With the x-axis mostly assembled, one end at a time I then checked that I could feel the slack between the end and the lead screw by applying pressure up and down on the end (you can feel and perhaps hear a slight clunk). I then gradually tightened the BL screws until the slack was gone. I then checked that the ends were not binding by turning both lead screws by hand. With this arrangement I was able to have the BL screws moderately tight, with no fear of coming lose, and not relying upon precise screw adjustment to get everything working. I needed one washer on the motor and, and two on the idler end, probably due to the different angle of the BL screw holes on each end. Finally, I powered on the printer and checked that the steppers were moving the x-axis reliably, and that the z top end stop was hitting the body of each end, not the BL screws (button head screws are essential for this).


The x-axis feels incredibly solid and rigid, with no z-axis slack/backlash at all. The most obvious flex now comes from the left side of the z frame (perhaps it's half Bear time?). Next step is to build the Skelestruder...


  1. that last motor screw/clamp is to be left loose until the end. As you point out, you should not screw down the bearing clamps too tight. I limited the range on purpose because I have found that some people jump on the end of breaker bars to tighten screws, despite my pleas. That said, 1/2 turn seems smaller than I planned. I'll look. Anyway, that top bearing will sit there nicely from gravity until you are ready. You can also use a temporary nut to hold it until you mount the motor. Once you secure it, I would not worry about it coming loose. There are three other dedicated screws and this one has a long shank which will help resist vibration/unthreading.

  2. good point. I will open up idler end a bit more to allow compensation either way. I wonder if you got some bridging issues at the ends. Can you check?

  3. agree. If you rotate the adjuster one thread, it may sit a bit lower, depending on where you are in the twist. Button head is a good idea. Let me know how the washer does. I don't have a better idea atm.
  1. Ok

  2. I'll check later today re bridging issues.

  3. The screw holes in the adjusters appear to be pretty much 45deg out of phase with the holes in the idler end body when the parts are fully mated, so I'm not sure that shifting thread will help, but I'll try. Was this the design intent, or did you intend that the holes would be aligned when fully mated? If the holes were aligned (and if the elongated holes were perhaps a bit longer), then fine rotation of the adjuster would set the vertical position, and the screws would hold it in place (with some trial and error to account for compression during tightening). Most anti-BL designs seem to isolate the concerns of setting position and holding in place once positioned (eg, opposing nuts), or alternatively constraining rotation and applying counterforce (linear slide/split keyed shaft and spring). As it stands, with the holes 45deg out of phase, there's a gap between the adjuster flange and the idler end body, so the screws are setting the vertical position of the adjuster, and attempting to hold it in place, and unavoidably applying counterforce at the same time - which seems like a hybrid of the two approaches, where the adjustment of the screws has to satisfy all the concerns at once. I might try putting some washers between the idler end body and the BL flange, and tighten (moderately) against the washers, rotating the BL adjuster before tightening to get the preload right.

addressed these issues in r1b

Cool design, once again! I'm printing it in a contrasting black and orange scheme.

I'm having a hard time picturing where the tension arm, swing, knob, and screw system fits in. Could you please post a picture of it, as well as the motor-side of the axis?


Here's some pics from my partial build (no motor or extruder yet, I'm upgrading to Skelestruder at the same time, work in progress).

Great looking prints. Thanks!

agree. Motor end looks really clean. I'm still using my beta prototype. Think 8-bit version. ;-)

I'll try to snag some pics for you. Hopefully it will be obvious once you have all the parts. My printer is in partial disassembly atm.

When printing BL adjust, should fan speed be 100% on all layers, or left off for the first n layers?

I think I left it off for the first 3 layers.

This looks so awesome I can't wait to try to fit to my wombot drafter.