Taurus Y axis for Prusa MK3 and MK3s

by jltx Dec 8, 2018
Download All Files

Thing Apps Enabled

Please Login to Comment

Ich habe mal für den Taurus ein Kabelkanal gemacht und gleichzeitig eine Befestigung vorgesehen, viel spaß damit, sollte auch passen.

Great, the slots are for thin cable ties. :-D

There is a support on F knob you need to remove.

  • F knob? I dont have this stl file list. Was that the Y_foot_r2b?

"Need the two smooth idlers that came stock on X and Y axes. I'm sure you've already replaced these and they are just sitting in a drawer."

  • not really sure what does this means unless there is an image as guide.

"Smooth pulley has a small opening on one side and large on the other."

  • I don't see any Smooth Pulley in the stl file list. Only has tau_big_pulley.stl

Sorry I am really confused with the written assembly insttructions.

Oops, I forgot to remove that support comment after I redesigned the foot and simplified it. I removed that and cleaned up some wording . Thanks for pointing out.

The smooth idlers are the metal ones that the belt goes around on the end opposite of the motor, as it comes stock.

Thank you for clarifying it!

I'm having another issue;

"Need the two smooth idlers that came stock on X and Y axes. I'm sure you've already replaced these and they are just sitting in a drawer."

  • I believe i only have one on my Y-axis and one on my x-axis. If i take out the smooth idler from my x-axis then i won't have any idler for my x-axis. (x-axis is currently using the smooth idler)

  • this means that i need to put back my stock Y-axis again and order a new smooth idler.

Correct. Many people have replaced those or switched to Taurus. If not, they are easy to purchase.

On my MK3, the place where you screw the motor frame into is used to zip tie wires in place. Is there any wire management after this has been installed?

I just left that zip off. It’s tied up on the other side so still secure. You can make a cable hook that screws on using the motor mount screws.

"insert 2 M3x18mm screws into right side of motor frame base. Use 2 M3x12mm to18mm screws on left end of motor frame." What do you secure these with? I might be missing something, but I don't see the instructions for attaching the frame to the base. The center two holes have hex nuts to screw into. The ones on either end don't have anything.

I temporarily used zip ties to hold the ends of the motor mounts (and the wires) which seems to work thus far.

Is there another instruction for this mod that is easy to understand and follow like the instruction for the Skelestruder?

No, but it is actually pretty easy once you try it. You don't even need to remove any wires or the bed. Just read through the text carefully. Is there a step that is confusing?

I'm having another problem, when Auto Home it trigger endstop wrong and also its very easy to move the bed by hand when motor is ON. I'm using the custom motor 17HS19-2004S1. I tried to increase the current to 640 via Gcode but same problem. I also tried belt more and less tight with same results. https://youtu.be/_LJwG9E_t1c
While printing everything seens to work so far. EDIT: It also fails on prints.

That belt looks way too tight. Try loosening. The new M907 doesn't allow you to adjust holding current. Need to re-enable old M911/M912 in firmware. But I am not having that trouble so suspect it is the belt.

I tried but it trigger endstop anyway. Can you try move your bed by hand to see if you can move it easy?
Also what is the perfect tight value for Y belt? Using the self-test report value.
EDIT: looking at code it looks like M907 also affect holding current:
case 907: // M907 Set digital trimpot motor current using axis codes.

ifdef TMC2130

    for (int i = 0; i < NUM_AXIS; i++)
            long cur_mA = code_value_long();
            uint8_t val = tmc2130_cur2val(cur_mA);
            --> tmc2130_set_current_h(i, val);
            --> tmc2130_set_current_r(i, val);
            //if (i == E_AXIS) printf_P(PSTR("E-axis current=%ldmA\n"), cur_mA);

Yeah, looks like it does both. I just tried moving and it does not budge. I have to disable steppers to be able to move it by hand.

Are you using 625 bearing on shaft? Have you tried removing belt and make sure motion is perfectly smooth through whole range?

Even if you apply a heavy force by hand?
I'm using the 625 bearing. Motion is not perfect the same over all rail, there are places that i can fell more friction, but this issue only start after the upgrade, it had the same friction before and auto home never failed.
On assembly process i fell like the motor shaft was hiting the part wall, maybe that's one of the problem, can you please confirm the motor i use is ok? https://www.aliexpress.com/item/Nema-17-Bipolar-59Ncm-84oz-in-2A-42x48mm-4Wires-w-1m-Cable-Connector/32831072482.html?spm=a2g0s.12269583.0.0.13e618d2NKKnGO
Shaft Length: 24mm

That looks like the same one I am using, but I didn't get through Ali so not sure if it is a clone. Shaft length is fine and won't hit. I pushed pretty hard on bed, was afraid I might damage belt, and it did not move until I disable steppers. You could try swapping out for the stock motor and see if it does better. Then I would suspect the new motor. I would also track down your friction even though it may not be issue now, it could cause print artifacts.

It have the correct label from STEPPERONLINE at back, looks like real deal. That's strange, i even try to increase to 35 holding and running current on firmware and nothing change. X axis is a lot harder to move when holding but still possible here. I will take care of it and test some things. Thank you

And since printing small patterns with PETG is a nightmare, here is a clean version of a pulley.

After assembly everything the self test always fails at Y. I did a cube print test and it print accurate 20x20mm x and y. What can be causing that issue? Is my assembly correct? My belt is tight up to produce a base tone on top side.
I've made a video: https://www.youtube.com/watch?v=UmwaLNbl4mE

Build looks good. Like the colors. Lucky you to have real Gates belt.
Odd about self test. I will check firmware to see what it’s looking for. Length is correct. Printing should be fine. I never run self test so maybe I’ll fail. ;-)

I got the belt from this seller: https://www.aliexpress.com/item/GATES-LL-2GT-3D-printer-gear-synchronous-belt-GT2-Width-6MM-9MM-High-quality-wear-resistant/32853559638.html best 3d part supplier i have found on ali, even the bondtech clones are great.
That can be the case, i can decrease Y max pos on firmware but don't seen a good solution for everyone. This mean if i do a printer reset or wizard i will get stuck on self-test, which is bad.
I also have the custom motor like you, running fine without tuning steps.
Other thing i notice is the rear idler have some space to twist when bed is moving, is not critical but not good for my eye

do you have pronterface or octoprint hooked up? It will spit out the length it measure during selftest to the terminal. Curious what you are getting. I think it is likely about 1-2mm shorter than it expects. They add margin to allow for slop in bearing position (which is dumb and what I am fixing here). Since you are no longer bashing on the bearings, you should be getting a tight ~215mm. Looks like it thinks at least 216mm. Doesn't actually affect anything since it is just a report back to you to let you know what it found during self test.

Nice i didn't know about that debug text to terminal.
Recv: X AXIS SG1=271
Recv: Y AXIS SG1=236

One thing i notice is that prime line is a bit upper than before and also measure points of PINDA looks like at a offset compared to original position

That's more odd. 236 is long. So it's failing on the long side (max=225). Your X is also long (max=260). X is tested before Y. Does only Y fail?

X passed and its stock parts, it only fail at Y.
Did you test your self-test?

Yes, I passed. And I see part of the problem. You read the SG1 values which is the belt tension (StallGuard) it measured. Right after those in the terminal output you should see two axis measured length for each. Mine were 255 for X and 215 for Y.

Well i know there's other message but it output a ?.
Recv: X AXIS SG1=273
Recv: Measured axis length:?
Recv: Measured axis length:?
Recv: Axis length difference:?
Recv: tmc2130_home_exit tmc2130_sg_homing_axes_mask=0x00
Recv: Y AXIS SG1=239
Recv: Measured axis length:?

Ok i reflashed firmware and now it output the values:
Recv: X AXIS SG1=269
Recv: Measured axis length:254.610
Recv: Measured axis length:254.770
Recv: Axis length difference:0.160
Recv: tmc2130_home_exit tmc2130_sg_homing_axes_mask=0x00
Recv: Y AXIS SG1=232
Recv: Measured axis length:215.250

I thought you were stock otherwise. Looks like have modified firmware and you missed: "c. modify platform.txt to enable float printf support:
add "-Wl,-u,vfprintf -lprintf_flt -lm" to "compiler.c.elf.flags=" before existing flag "-Wl,--gc-sections"
"compiler.c.elf.flags=-w -Os -Wl,-u,vfprintf -lprintf_flt -lm -Wl,--gc-sections""

Anyway, 215 is spot on. So why is it failing? Did it not give you a second measure? it measures each axis twice and compares the difference. You are missing the second measure for some reason.

This last self-test was with stock. And yes i was custom before but same behaviour. The measurements are exactly like shown on video: It goes back - front - back and fail no other values come to termial than that.
I have done about 10 self tests and they all fail, Recv: Measured axis length:215.250 always same value
On firmware: #define Y_MAX_POS 212.5
212.5 to 215.25 is a bit of a difference while X pos is on spot #define X_MAX_POS 255

I see in the code they add 8mm to Y_MAX_POS, so 220.5, then allow 5mm margin, or a minimum of 215.5. You are 215.25 so it fails. Mine was actually 215.6. You could probably pass with adding a piece of tape on one end. The self test is not really important though.

Wait, that was dumb, you need longer. :-) so just shave a hair off an end or the belt clamp.

Ok i trim with sandpapper, it now works.

Recv: Y AXIS SG1=232
Recv: Measured axis length:216.370
Recv: Measured axis length:215.730
Recv: Axis length difference:0.640

BTW what are the correct values for belt tension?

I'm in assembly process and i have everything done. How do you route LCD cables now? From top of motor to the VSlot?
Also cables from motor that were ziptied to the frame holes how do you fix them at the middle?
I've used nylock nuts to fix the left and right 4 screws of the motor frame, is that the right way?

I didn’t have to move my lcd cables. They should still fit.
Are you talking about the wire from the right Z motor?
Nyloc or regular hex nut. Either should work.

I have route my LCD cables like that. I tried to put under the plastic tab from motor but it cracked. I think is not a problem

Taurus hard feet look very low profile for my printer. Using them the printer is not balanced now.
EDIT: It seens my frame is not 100% leveled, one coner foot stay in air, i temporary fix with some pads

Love the look of this. Would love to see it made compatible with the bear frame.

I'm planning to do your fantastic MOD for my MK3 next week. Thank you for your great works :)

BTW, could you tell me what the filament guide you use is? I mean the rotating thing on top of the gentry.

That’s just my filament runner design.

I just finished the job. Looks so awesome!

But there's a problem.

Y-axis length errors on self test. Is there something wrong with my work?

Only one suspicious thing is the mounting direction of belt clamp

I made it!

There were two problems:

  1. Wrong direction mounting belt clamp
  2. Too much belt tension

Your clamp is backwards, which would just be a potential clearance issue with motor frame, but would not change overall Y length. Unless it hit the motor, but you would see that stop way short. So that’s really odd. Does it report a value?

You’re right. Clamp was not.

It was tension.

Now everythings going fine. Thank you :)

Here’s the cube.

Gold one is before and silver one is after. It seems there’s a siginificant improvement.

I noticed almost all parts parts got updated and dated 1-18-19 but no changelog, what changed compared with before parts?

Nothing. I had to reload them and they got a new time stamp. Same parts.

Hello, I've remixed current version of gear 608 bearing pulley. The old version didn't fit well on my 608 bearing and did some weird noises while printing. Here is my version.
NOTE: print with 0,25mm nozzle with 100% infill

What did not fit correctly? PLA or PETG? Is your extrusion calibrated?

Be careful with GT2 profile as many people get the pitch line wrong which will wear the belt.

I've calibrated everything but two parts were splitting so I've connected them in fusion and now it works nice.

I attached more photos of the back of the printer showing the back of my frame. Checked the other two printers and they both look like the one in the photos. I got the printers already assembled so I don't have any Ideea why I have a different middle frame part :(

I wonder if Prusa has gone to a different frame design. They did not update their dxf file as you can see. The good news is you are probably fine without those screws. But, I wonder if the position of the other screws has changed which would explain your clearance issue. I'll ask around on frame change. I would need a new dxf file.

I checked with a few folks who have printers before and after your build date and they all match mine and the official DXF. So I don't know what you have. But since you got it from PR I'm sure they stand by it. Anyway, are you able to run with the tweaks you made?

I just checked their new online 360 degree viewer and there is no deep notch. So either I have an old frame or some prototype/testing one. Anyway as you said i think it's no problem without those. Maybe this is the same reason that the belt clamp was hitting the rear pully bearing. Time will tell ;)

Check out step 3 https://manual.prusa3d.com/Guide/4.+Z-axis+assembly/106?lang=en

I would chat Prusa and see if they will swap out that mk2 frame for mk3.

I see :( I searched and couldn't find any advantages of mk3 middle frame vs the mk2 .. which is true i have it on all my 3 printers which should have an mk3 frame.

I will contact them just to see what they can say about this.

I have run my printer with your y axis and z axis and it feels like a different printer. It's more stable and it moves more fluently - they y axis at least. I would have mounted the x axis also but i later found out that it wasn't compatinile with my bear extruder.

I will post a make after i receive the threaded rods for the z axis (hope to be tomorrow).

They drilled this frame to fit MK3 extrusions, so it is some interesting hybrid. What is the spacing between the inside left (einsy side) and right (psi side) of the main opening? It is narrower on mk2 so wondering what you got. If it is wider like mk3 then this is more strange.

I've printed this in prusa white petg (not prusament) and after assembly the bed clamp slighty hits the top extruder motor screws - i replaced the screws with flat head ones but the clamp doesn't have enough space to slide freely because of the rear bearing ... it hits the fins of the bearing. I will try to sand a little from the top and bottom of the clamp to make extra space tomorrow.

Also I can't put the pair of screws on the left side of the motor as the holes of the printer body are a little to the right. I've put only the pair of screws under the motor and the ones on the right and i think it's enought the hold the motor frame but it's kind of strange to have screw holes and no screws ;) I will try to post some photos tomorrow.

Anyway I am following you for some time and I like your designs very much. I've printed the Z axis and mounted (partially because the threaded rod 380mm threaded rods didn't arrive yet) and the X axis also but I have not mounted yet. The skelstruder is on my list but I need to get some experience first.

Thanks for all the work you've put in the taurus & prusa projects. Regards

Had a thought. Did you mount the belt clamp so that belt slides in from left (motor side) or right? It needs to be to the left. I realize I didn’t say that in instructions.

That’s very odd. The screws should all line up. This is mk3, not mk2.5? When did you get it? What is diameter on your pulley flange?

It's mk3 delivered in may 2017. The distance between the center of the holes on the left and the one under the motor is about 28.60 mm and the diameter of the pulley that I took off the x axis and is the same as the one that i had on the y axis is about 17.90 mm ...

I had the bed clamp sanded until it didn't hit the rear pulley. So maybe you can release a bed clamp that has the bottom part (the one facing the motor) a little smaller in height.

I attached a photo how it looks now. For future reference the front pulley needs to be vertical or 10 degrees from vertical right? I wanted to be able to tune the tightness of the belt more and mine is now 20-30 degrees from vertical .. will this do ?

I like that white. I need to get some. Your idler angle is fine. A little steeper is good, but you don’t want vertical. Pulley size sounds right so I don’t know why you are having clearance issue. Do you have any mods that may have altered bed height, like leveling nuts? You did have belt slot in from motor side?

How far off are the screw holes on the left? Just a fraction too much for screws to pass through? The screws you have should be enough, but I don’t understand how that is off. I used the cad file from prusa git and mine fit perfectly and others have not raised the issue. Curious to understand cause.

I do not have any mods on this printer apart from the y axis smooth rods holders, but they do not alter my bed height. Anyway as I told you after a little sanding it slides great. Now regarding the screw holes position on the motor frame I attached a photo so you can see where my screw holes are ... also I have 2 other printer (shipped together with this one .. may 2017) and the screw holes are in the same position as this one. Maybe you can make the screw holes wider so it can work with my kind of holes and yours ... just a thought ...

I am mounting the X axis now and maybe tomorrow the threaded rods will arrive so I can finish the Z axis completely.

Best regards

You have something really strange there. That frame looks nothing like mine, or any one I have seen, and does not match the CAD file I got from PR. Look at the deep notch behind the extrusion. That does not exist on mine. Those screw holes are in a completely different position. I'll see if I can find a DXF that looks like yours so see if I can figure out what is going on.

I can only find the MK3 version. But you can see here this does not match what you have: https://github.com/prusa3d/Original-Prusa-i3/tree/MK3/Frame

I really want to do this mod but the instructions are quite confusing. Also - what non-printed parts will I actually need? I ordered some 608 bearings but it seems you need toothed bearings (I ordered smooth ones) is there a link to each part you can post so I'm sure I'm getting the right parts? For the 625z bearing can it be smooth or does that also need to be toothed? how many teeth? etc.

edit: also belt length needed. I have a brand new GT3 belt sitting in my drawer but I don't want to go ahead and print all the parts and take everything apart to find that the belt is too short. It's standard length belt for the stock Y axis. Do I need longer belt?

Yes, the bearings are smooth. You will print everything else that is needed. I think it will be clear once you have the parts and follow through the assembly. You can do most of it off the printer first to see what you are doing.

I am not an expert on belts but I understand gt3 is not compatible with the motor pulleys, etc that are gt2/2gt, due to different tooth profile. May want to check into that first.

The bearings in your photo look like they have teeth (red) ?

restored. look at big_pulley stl and see if that clears up your question.

How's it going with the 48mm stepper? I'm thinking of ordering. What's your reason for trying increased torque vs stock motor? What current value are you using to drive it?

Fine so far, but all I did was straight swap it. bed has most inertia so wanted to have more torque available. So far motor runs way cooler. Have started tuning it yet.

This was one of the best projects I have printed. Thank you for putting this together. Now on to X and Z!

Great! Glad you like it. I am happy now that I fixed the foot. The whole thing came out better than I expected. It started as a dare. :-)

Can give a bit more margin to the pulley? Even 0.4mm can't print it all close on 2nd or 3nd layer. I'm almost able to print everything with 0.6mm nozzle. If you can just increase the bottom diameter for it complete the perimeter will be great. +0.4mm should be ok

Many people have printed this on 0.4. Try the one from the Taurus X which has a different pattern. May work better for you. But still needs 0.4 for the teeth I’d imagine.

No, i was able to print the teeth with 0.6mm, i just need to close that loop. See img.

Interesting. But try the X one. I’m not near my computer.

Yes, that’s what I’ve been saying. They are identical, just different patterns.

I was able to print, and belt have a good trip to teeth. My first try it break when insert bearing, so I scale to 101% and now it inserted good. I will print the rest and report back. Thanks

I'm a noob. The details say I need a new GT2 belt. Can someone point me at what I need to order?

I think it's this stuff in 6mm. If so how much length should I order?


Sorry, missed this. Wow, that’s the real thing. Hard to find normally. Are you planning to redo X as well? Get enough for both. I got 6 feet to be sure.

Thanks. I think I have to redo X to get the second idler.

I've 0.6mm nozzle, and there are some parts that doesn't print and show on slice, is it possible to do a 0.6mm nozzle variation for small parts?

I can upsize the lash a bit to work with 0.6. Any other parts?

Oh, wait, I don’t think you’ll be able to print the pulley and resolve the teeth well enough. Check that first.

yes up size everything will not work since it will need a good margin, only perimeters must be thicker to work with. But i can try reduce wall to 0.5mm or even 0.4mm and print to see if that works, a experience

teeth will not print at all, to much thin, even with "thin wall detection" it can't print.
Walls must be > 0.6mm to print with 0.6mm nozzle. 1mm wall to give a margin is the best value and also works with 0.8mm nozzles
I've uploaded screens from first and top layer. Even the support from knob is missing

Pulley teeth need to fit belt so will be challenge with 0.6 nozzle. Printer came with 0.4 nozzle, do you still have it? I am reworking the foot.

I do have, but is a bit sad keep exchanging nozzles. There are any hardware i can buy to replace that teeth bearing?

The rear idler lash/strap is... entirely ostentatious. I love it. I'm printing all the main parts in black - except for that thing, which will be bright orange. Thanks again for these designs!

Can you share your print setting with slic3r or gcode file ? I print "tau_f_knob" fail 3 time
Thank you.

How is it failing? PLA or PETG?

I print it with PLA 0.4 nozzle, 0.15, 2 perimeters, 5 top 5 bottom, no thin walls, no brim.
I will print it again and take a picture for you. It not stick to the bed after 5 layer something like that.
Can you share some tip to print it ?

I redesigned the foot so it is easy to print and uses no screw. Unfortunately you would also need to reprint the motor frame.

let me print it again and see what's up. What you did sounds fine.

under printing / advanced, try increasing overlap to 50%. If you are just getting a few bridges not connecting, that can still work. You can also try turning on supports. if it is worse let me know. I can increase the center support.

Can we make a remix of this for the Bear Frame mod? Looks like a real intriguing design

It already works on the half bear upgrade. It won’t work on the full bear because there is no cross member.

No i mean can we mod it to work for the Bear, are you willing to share the source files?

Well, sure, except you need Greg to add a cross member first. I'll talk to him.

That's easy enough for me to add with some spare 2020 and some 90° corners, i'd just need to mock up the alignment and merge in the v-slot channels instead of it mounting to the flat plate.

You extrusion guys speak a different language. :-) I am trying to improve one thing on the frame, then I can release.

Lol! Ok great thanks, you're awesome man!

haven't forget, but I see tech issues with bear port. need to investigate.

Any news on this?

Hi jltx! I hope you don't mind but I'm trying to take a stab at getting it to work with the full bear frame too (gives me something to do.) I don't want to be counter productive though if you're working on it or if there's something wrong with it. I figured I'd try to convert the pieces over into Fusion 360, get things lined up, and then try to modify the motor mount to mount onto a 2020. I can also shoot it over to you when I'm done. You're designs are awesome and I'm hoping to be able to use them :D

No worries man :D my bear build's on hold till all this holiday printing is done LOL

Nice design!

I keep thinking that inverting the belt so the teeth are at the other side (outside), should be an improvement. Then except the motor pulley, all the other pulleys would be toothless. Much easier to print and you will get rid of the belt-teeth/pulley match problems.