April 16, 2025
Description
UPDATED: No drilling of the gantry cover is required if you print the revised front left Core XY mount.
NOTE: This is a full replacement of the stock part cooling fan and stock part cooling duct.
I just want to start by saying that I love my Prusa XL 5 tool head; but it's a love/hate relationship. If you don't want to read the back story, just scroll down to the instructions.
The designers at Prusa made some seriously flawed decisions when they designed the tool heads. The most outstanding being the decision to use a 4010 radial fan for the part cooling and the decision to make the supply voltage for the part cooling and heat break fans 5V when 24V was already available on the Dwarf board.
As a result, the part cooling is very poor because the poor little 4010 has the static air ability and air volume moving capacity of an emphysema patient. The issue is further exacerbated when the printer is located in a hot environment or fully/partially enclosed.
I live in ‘The Land Down Under’, where room temps generally exceed 32°C and get as high as 42°C. I have found that there is a significant drop away in cooling performance once the room temp climbs past 26°C.
As my base to work from, I have a Prusa MK3S in a custom enclosure and print all the time with door closed. The chamber temp sits ~46-50°C from passive heating, but with the stock 5015 part cooling fan and the brilliant fan shroud from RH_Dreambox; I can print any material, at any speed the printer is capable of, with overhangs in excess of 70°, and achieve consistent, flawless results.
https://www.thingiverse.com/thing:4124587
It was extremely frustrating when I came to the realisation, after many kilos of filament and over two hundred hours of testing; that my expensive, and long-awaited, XL couldn’t hold a candle to what I could produce on my MK3S. Even slowing the XL down to speeds below 40mm/s and reducing print temps on PETG to 225°C made little or no difference. Overhangs of 45° or greater were completed to a poor standard or just outright failed.
It was around this time that the 3D printing community, also experiencing the same issues, started releasing their own cooling shrouds for the XL. I tried all of them and all yielded the same, or in many cases, worse results. Many caused the fan self-test to fail as they restricted the airflow so much that the fan could not achieve the required max RPM.
In fact, the stock fan shroud will cause the same issue and Prusa Support regularly have customers remove the shroud to get it past the fan self-test stage. This is all the confirmation I needed that the stock fan and shroud combo are an epic failure. They rapidly released a ‘user pays’ upgrade for the MK4, which suffered the same deficiency. Surprise, surprise, it's effectively the same cooling solution as used on the MK3S, albeit with a slightly bigger 5020 fan instead of a 5015. Prusa went on to make it part of the standard build for the MK4S, which closely followed, and it's also the basis for the CORE One's part cooling.
Unfortunately, due to the clearances on the XL, they have painted themselves into a corner. In my honest opinion, this is why they have not released a solution/upgrade to resolve it in the 2+ years since the XL's initial release.
So, Project ‘Mastodon’ was born and I set about carefully measuring the clearances on the tool heads, etc. There are a lot of moving parts to take into consideration and clearances are tight! Life was made significantly more difficult than it needed to be thanks to Prusa’s increasing move away from Open Source by way of them still not releasing 3D models of the XL, schematics of the Dwarf board, etc.
I contacted Prusa Support to see if I could get a 3D model of the Nextruder. All I wanted was a cutaway of the actual heat block so I could use it in CAD for working out clearances. I didn't need any details of the Nextruder itself. I was happy to sign any required NDA, etc; but that was a ”No”.
Given that dead end, I reached out to the community to see if someone could do a high resolution 3D scan of an XL tool head.
While I was hoping and waiting for someone to do the scan, I spent a significant amount of time trying to source the highest output 5V 5015 available that also had an operating RPM range that would match the stock XL 4010 so that there would be no errors or failed self-tests. This ended up being harder than expected. There are tons of awesome 12V and 24V 5015 fans available, but almost no high RPM 5V ones. I ended up building an Arduino based test rig to test fans and bought a laser tacho to verify tacho readings. In the end, there was only one on the market that met the criteria from a company called MX Research.
And yes, for those wondering; I experimented with a DC buck converter to step the voltage up to 24V and a MOSFET to translate the PWM for the 24V side.
Grant at 3D Musketeers came to the party with the 3D scan of the Nextruder about 6 months later – thanks dude! 😊
Had it not been for the lack of files and support from Prusa and the delay in getting someone that could provide me with a 3D scan of the Nextruder, this project would have been completed around a year ago.
Finally, I was able to start working on the problem, and oh boy was it more involved than first anticipated!
I had a working prototype by the start of October 2024. Initially I ran into issues with the clearances off the left side gantry. A 5015 fan is 5mm wider than a 5010 and I quickly found I didn’t even have half of that to play with. Weight was also a major consideration.
I took inspiration from the Voron StealthBurner hotend and removed the outer cover from the 5015 to reduce the overall thickness and reduce weight. The mount for the fan becomes the outer cover. Also, a key design consideration was to not modify the printer in a way that was not easily reversible, whilst also being achievable by any user with tools they already own.
Next was the challenge of the shroud/duct itself. This is extremely difficult due to the fan outlet being offset from the centre of the Nextruder, how close it is mounted to the bottom of the hotend, the clearances for docking/undocking and the need to clear the spring loaded ooze stopper on the dock.
After MANY, MANY revisions, hundreds of hours in Fusion 360, hundreds of hours of testing and many kilos of filament; I have the current design and it has been running on my machine for 3 months now with hundreds of hours of production printing. I have been able to print many parts that were previously unachievable on my XL.
The final package is only 3g heavier than the stock setup. I highly recommend upgrading to the awesome “Skelettonized” tool head parts from Kevin248 as they reduce the overall tool head weight by 14-16g.
https://www.printables.com/@Kevin248_103670/models
The cooling is still not quite as good as my MK3S, but it’s close. I believe this is down to a combination the RH_Dreambox shroud being refined using CFD and my design being the result of some very limited CFD (4 runs), a lot of eye-balling and ‘feel’ as well as water bowl testing; but ultimately it is just the major limitations of the area it mounts and the clearances I had to work with. I’m sure with full access to CFD, it could be tweaked to be better.
Anyway, that’s the abridged version of the journey. It’s been long and emotional with an enormous amount of time and effort and money going into it. Just this write-up and the following instructions took ~6 hours.
With that out of the way, on to the instructions! :)
This is not designed for the single tool head XL. I do not have access to one, so cannot guarantee it will fit. If it does, please let me know.
Some of the parts you print will look slightly different to those in the images below. The photo series was taken during disassembly/assembly with early prototype parts.
BOM:
PC or or ASA should be used for the fan mount and duct due to the heat the parts are exposed to. PCCF or ASA-CF is recommended for the duct due to the large suspended section of one of the outlets.
I highly recommend the Silicone Carbide Nextruder nozzles from Phaetus. They're cheaper than the stock brass nozzles and you can kit out a 5TH XL for less than one Diamondback nozzle.
https://www.aliexpress.com/item/1005007877919205.html
I have provided pre-sliced, pre-supported models for printing on the XL. You may need to adjust the filament settings, etc, to suit your printer.
For the fan mount, a 0.25mm nozzle is recommended due to how thin sections of the walls are. You will probably get away with a 0.4mm and Arachne, but the resulting print may not be as pretty and may have some holes in it.
If you really want to ‘go it alone’ on your settings, load the pre-sliced version as a guide.
I recommend printing the new Core XY Mount in PCCF or similar. Use a 0.4 or 0.6 nozzle, 6-8 perimeters, 10 top and bottom layers with 20-30% infill. You can also install M3 heat sets instead of the two square nuts.
The spacers can be printed in PETG or whatever you like. I would never recommend PLA.
You don't need to be print the ‘Front Spacer’ or ‘Drill Guide’ if you print the new Core XY mount instead.
NOTE: You only need one set of the spacers. That means you Alex! ;)
Printed parts and fan
Gantry cover
Unfortunately, despite all the design work, we still end up ~1.8mm too wide overall to allow the printer to successfully home the X axis. The new fan mount will now contact the sheet metal cover on the left gantry before the tool head hits the end stop on the X axis. This results in a crash during homing operations. I wasn't kidding when I said the clearances were a major problem.
To get around this, we need to space out the left side gantry cover.
The print heads don't need to be removed from the printer. Lay a sheet of cardboard or a towel on the build sheet to protect it and then just manually undock the tool. You can completely remove the tool if you like. It's just 2 bolts on the flexible plastic support strip for the cable loom, unplug the cable look and disconnect the Bowden tube.
All done!
Now you just need to run a belt tension calibration
https://help.prusa3d.com/article/adjusting-belt-tension-xl_401793
Phase Stepping calibration
https://help.prusa3d.com/article/phase-stepping-xl_681760
and an Input Shaping calibration to account for the weight changes and you're away.
Prusa don't have an article for Input Shaper, but the routine is found in Setting > Input Shaper > Calibration
You can also run a fan self test to confirm all the fan runs and pass the test. Control > Calibration & Tests > Fan Test
As I said, this was a massive project and it was all done for free. I make no money out of it and there are no affiliate links in any of my designs. If you do do this mod, please be sure to give me a review and post your make. This will help me get some Prusameters and collecting them is the only thing I get back for this.
Happy printing and be sure to check out my other XL mods.
https://www.printables.com/@IPIND3D/models
License:
Creative Commons — Attribution — Noncommercial
7