New Y gantry – stacking the pulleys again

After doing the groundwork it was time for me to take care of stacking those idler pulleys. Mounting holes (for the superstructure) were already in place and doing my research I had additionally discovered the word “Hülsenmutter” in the …at PuddingBaer91’s insistence for use with these… section of the description under  TronXY X5S New Y Gantry bracket by ljbrumfield. I can’t tell if “threaded sleeve” or “threaded socket” is the right english keyword for finding them – the german ones can be Gewindemuffe, Langmutter, Distanzmutter, Verbindungsmutter, Gewindehülse or something else – depending on your application. 😏
The important thing is: I found some cheap M4x30mm ones with an outer diameter of 5mm – prefect axles for the idlers! [1]

Having some leftover idlers lying around, those threaded sleeves could be immediately tested. I used my cordless drill and some fine sandpaper to shave off a tiny bit of the coating and the MR95ZZ (5x9x3mm) bearings I ordered seperately slid onto them just right.

When trying out one of my “toothless” idlers, I immediately could feel the bearings had to be replaced. A bit fiddly, but possible (not without completely destroying that old bearings, though).



But how to place those pulleys correctly? Uri Shaked strongly advised to put the “Vitamins” into the OpenSCAD project. So I added the Misumi aluminum extrusion 2020 profile by beverageexpert and also imported the Tronxy x5s reinforced left motor mount by xxdanmurphyxx, which I intended to use. I read Mark Rehorst’s CoreXY belt path explanations again and started to beef up the .scad file…


There are now quite a few “Display/Compile Options”, which can be active (1) or inactive (0).

Y_Gantry = 1; // Carriage for Y-Axis, lower part with MGN12
Y_Gantry_Idler = 1; // Carriage for Y-Axis, upper part with Belt-Idlers
X_Brace = 1; // Brace to support the X-2020 from below (optional)
Y_2020 = 1; // Aluminium extrusion, download seperately!
X_2020 = 1; // Aluminium extrusion, download seperately!
Motor_Mount = 1; // Motor mount (STL-import)
Corner_Mount = 1; // Corner pulley mount
Belt_Paths = 0; // Visualize belt paths and edges of pulleys
Axles = 0; // Visualize Axles and pulleys
ABS_support = 0; // Extra support for difficult materials (currently only Y_Gantry_Idler)
Logos = 1; // Embossed logos (=> viewing and compiling works quicker without PNGs)

The “Y_Gantry” part from the last post was re-designed for stacked idlers.
I’ve played Bridge Builder and added the upper frame, called “Y_Gantry_Idler” (and meaning idler holder). It is fixed with M3x16 secrews to the gantry plate (or M3x20, if you add the X_Brace) and holds the idler axle. I kept the “switch” for 400 vs. 450 mm linear rails on X and added another one to distinguish between the “Socket” (threaded sleeve, mounted with M4 screws – one M4x16 and one M4x20 on each axle) and plain “M5” as axle.

The “Y_Gantry” as well as the “Y_Gantry_Idler” parts now need support for printing some additional overhangs. For the “Y_Gantry_Idler” I had chosen some difficult material and added the “ABS_support” switch to add support and lilypads to the model itself.

The “X_Brace” isn’t in use on my X5S at the moment – more on that later.



“X_2020”, “Y_2020”, “Belt_Paths” and “Axles” only serve as visualisation, but I added a “Motor_Mount” to the package, that I re-mixed in Tinkercad. I really liked xxdanmurphyxx’s implementation of raising one motor in the mount (thereby reducing radial load on the shaft and the bearings), but I couldn’t find a hint on which mount to take for the other motor.

The mounts are the same – it’s up to the user to mirror one of them!



To keep the belt paths defined a “Corner_Mount” has to be added to the overall package. The rugged way  X5S Corner Pulley Guide by HanCheol was designed seemed to be a good approach, but as I wanted the positioning to fit, I designed it myself.
As this was the last part printed, I had to carve a bit, but the files I uploaded are already corrected to fit around the screws TronXY placed in the corners of the 2020 and 2040 extrusion profiles.



Fitting the parts together became a bit of a challenge for me, because of my exotic material choices. I came across Alfa+ / AlfaPlus from Filoalfa via It was used for the gantry plate. Basically it is an enhanced PLA (that info isn’t online yet, but they told me it is PLA via Mail), which can be “annealed” at 80°C for an hour, to raise the heat deflection temperature from 55°C to 108°C and improve it’s toughness. Look at their product video and you’ll see, why I ordered the 3x250g sample via Amazon.

Stiffness is looking good: This is Alfa+ (pink), dasfilament PLA (black) and Procatec PETG (green) under nearly similar conditions. The middle print was out of material before finishing, but the difference to PETG is obvious – even more, when you deform the square spiral flex test by Feric by hand.


The problem with baking your PLA (generally – not Alfa+ specific) is, that it deforms your parts. As the plastic is frozen into place while printing, it is put under internal stress, which you can whitness best when looking at warping ABS parts. If you “anneal” your PLA in the oven, the same happens here. Additionally X/Y and Z dimensions react differently – one plane shrinks, the other one grows!

The X-Brace couldn’t be added to the whole package anymore, because of the way everything was deformed. Next time I’ll try boiling water for annealing…



Apart from that, the fluorescent pink overdrives my PiCam. 😀



The other exotic material choice was a special offer. At 1kg PC-ABS for less than 33,-€ I had to try out the printability of Filament-PM / Plasty Mladeč’s blend. “PC-ABS” should mean there is more Polycarbonate in the mixtrure than Acrylonitrile butadiene styrene. Nevertheless my E3Dv5 with it’s standard temp sensor (295°C maximum) could easily handle it at 250°C and the build plate could also reach the recommended 110°C (I guess my maximum here is 120°C). What I expected was that parts print more bitchy than with regular ABS. And I was not disappointed. Left/green: Procatec PETG, right/grey: PM PC-ABS (first prints – start of tuning the parameters, temps of PC-ABS are 20°C higher than embossed).


The smell was disgusting and the parts were warping heavily. Overhang quality was subterranean (no fan allowed) and as PC is hygrsocopic and the X5S in the cellar my surfaces were pretty on not a single spot…
…but heat deflection temperature of this material is 98°C and it’s stiff, tough and strong (and really comfotable to machine). I had to use screwdriver and hammer to break away supports, where a scriber would have been enough for other materials.



So how did everything turn out so far?

Can’t complain. The amount of failed/unusable prints wasn’t THAT big this time… Hey, I dialed in two new materials and tried out a new design, so that’s not much:


…and I also had the opportunity to de-rust and re-grease my MGN12 linear rails. Speaking of them: Don’t forget that you need an alignment tool! They are included in ljbrumfield or DKDaigle’s parts on Thingiverse and I’m too lazy to copy them. There is also a standalone version.



The yellow and pink Gantry Plates had to be mounted to the 2020 profile with a hammer. So there is no play at all.

They skewed in the oven in the same way. That means my X axis was also skewed (but not as bad as the photos show – the camera lens adds some extra distortion).



To pull them straight again I used re-purposed my old Tronxy X5S Z level by Veterimario77, a screw clamp and my hot air gun. The remaining skew was taken out of the mechanics by brutally tightening the belts until everything was angled properly again. The screw heads of the Z axis rods are a useful reference point for that.

Hopefully noone else tries out that “annealing” process on parts like this. 😉



The rest works the way I wanted:



As I still have to design a new X-carriage to match the new belt paths [Update: The X-carriage straightening the belts is now documented here and can also be found on YouMagine], it’s no “Benchy time” yet. But the speed test already looked promising.




This is the test model I use for acceleration, jerk and maximum speed in general:

My Y-Gantry parts and OpenSCAD code are here:

Libre3D seems to be down for maintenance at the moment, so I’ll add that link later…

[1] Vovo Schrauben Shop, Ebay (germany)

3 thoughts on “New Y gantry – stacking the pulleys again

  1. Hi Becks

    After thoroughly reading all your posts about upgrading Tronxy, I’m eager to go back to the drawing board and add / test a couple of changes to my own design. You are very knowledgeable in different areas, corexy theory, materials and design. Hats off!

    There is one thing not entirely clear to me: what would be your final material recommendation for your prints? There’s pros and cons to each choice, I get that, but considering all facts, what would be your choice?

    Thanks for your time and effort put into your blog, very much appreciated!

    PS: I’ve been working on a novel tool changing mechanism for a while now, then got frustrated, now my hopes are back up. In case you’re interested in my results, let me know. I’m happy to share and nerd about it.

    Liked by 1 person

    1. Hi Claude,
      my material choices for “end use parts” would be PET-G in most cases. And ABS/ASA, if you need it to be stable at higher temperatures (e.g. for the part cooling duct or for the motor mount, if you use a heated chamber).
      PLA is good for beginners, but as a hot summer day can warp it (and annealing will definitely warp it), I’ve almost stopped using it.
      Discussions about stiffness of the materials themselves are mostly useless, as long as the construction of the parts allow the walls to be at least 2mm thick. Experimenting with flexibles gives you a good feeling about the influence of wall thickness on part stiffness.

      Regarding the sharing of results: More is always better and every hint maybe helpful to someone. Especially, if those experiences are available on the “proper internet” and not some gated community (Facebook group). I’m always trying to keep at least my YouMagine / Libre3D part descriptions up to date, even if at a certain point I’m not satisfied enough to document something here.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.