Prusa XL Core XY Rework / Optimization (for new O-Rings / Washers)

Optimization of all Prusa XL Core XY parts for better idler alignment, tighter tolerances, better printability, ...
17
55
2
446
updated July 25, 2025

Description

PDF

Preface:

Since the beginning my XL was troubled with many problems that also included severe VFA's (you could even feel the ripples), unequal belt tension or misaligned x-carriage and unable to calibrate IS as either one or the other belt were said to be too lose, no matter how tight they were.

My XL was even sent to Prusa at the beginning of the year to fix all the issues including the ones stated above, but was sent back to me then without having changed anything, telling me everything is fine (and for my other issues they don't have a solution, but that's another story).

So it was obvious to me that at some point I have to try and fix it on my own once again (that's part of the Prusa experince since the MK4 I guess) and take the core xy apart and investigate.

Getting the XL Enclosure was finally the right excuse I needed to reprint all the core xy parts in PCCF and also see if I can improve them.

 

Findings:

It all started off that I wanted to use a toothed idler to conquer the VFA's so I first went with the solution from @INVESTIGATE but I was a bit surprised about the comment to print the parts at 0.25 instead of 0.2 and this got me thinking.

Next step was to import all the necessary step files into Inventor, take a closer look at the parts dimensions and make an assembly…

…yeah, and there I found the issue, the models were not designed with 3D printing in mind, they were all purely mechanical and theoretical, some with 3 digits after the dot…

…so it actually depended on how you orient the model on the print plate and how the slicer interprets different heights in relation to layer height if you then got an accurate model or not…

…additionally the planes for top and bottom belts and the corresponding idlers didn't match up at all, they had quite an offset, especially when printed.

Also one thing that needed to be fixed were the way too big tolerances (which I'm sure they did as they were necessary the way they designed their parts) and the oversized holes for the idler shafts, which were wobbling like crazy. This was especially noticeable on the core xy back parts (motor mounts) and I guess the reason for this was the possibility of the top part being misaligned.

 

Goal:

  • rework all 3D models with printability at 0.2 layer height and orientation on the print bed in mind
  • align all idlers so they are at the same level (top and bottom) as good as possible and match up with the top and bottom belts (the exact position for the belts is determined by the tool holder on the x-carriage)
  • reduce overall play to a minimum and get precise holes for the idler shafts

 

Y-Carriage:

 

Core XY Back:

  • adjusted idler positions to be in line with the y-carriage idler positions
  • reduced the hole size for the idler shafts to ø2.8mm and added a point angle to them → use a ø3mm hand drill to ream out the holes to a perfect fit for the shafts (or a ø2.9mm hand drill if you want to hammer them in)
  • gap for idlers adjusted for the right amount of play
  • added a printable pin to perfectly align the corexy back with the corexy back top part
  • removed features that are no longer used
  • added a chamfer to have enough material left in the area of the point angles (it also helps a little to stiffen up the part)

 

Core XY tensioner:

  • there's now a Left and Right corexy tensioner → was necessary in order to have the idlers on the same plane as the idlers on the y-carriage
  • gap for idlers adjusted for the right amount of play

 

Tip:

  • When you're done with the assembly, check position of the pulleys on both stepper motors and align them properly, both were quite off for me as well

 

Results:

  • both belts have now the same tension (82Hz) and the alignment of the x-carriage to the end stops is spot on
  • IS calibration worked first try and resulted in MZV 35Hz for X and Y axis
  • VFA's have significantly improved (see pictures before and after, both parts printed in Prusament PETG Jet Black), but are not completely gone (think that's impossible) → they are acceptable now and almost as good as what I'm getting on my MK4S after some modifications

 

Are toothed idlers necessary?

I'd say not 100%. I tried toothed idlers on my MK4 after I had already done the modifications to reduce VFA's and I couldn't find any noticeable improvement on top of that with the toothed idlers installed. The before and after is quite similar to what I got on my MK4 so I'd say belt alignment and properly aligned idlers and pulleys have way more impact on reducing VFA's than anything else.

 

All parts except for the pins are printed in Prusament PCCF → make sure you have your filament (extrusion multiplier) dialed in! 

Pins are printed in Prusament PC Blend, but PETG or even PLA should be fine as well. If they are too lose, just slightly increase the scale by 0.5%.

Open the 3mf files as project, you only need to adjust your printer and your filament, everything else is set up and ready to go.

 

 

Tags



Model origin

The author remixed this model.

Differences of the remix compared to the original

  • Optimization of all Prusa XL Core XY parts for better idler alignment, tighter tolerances, better printability, ...
  • Detailed info's to be found in the model descritpion

License