The Papilio

Current state: Alpha release and testing

Papilio, Latin for Butterfly.

What is the Papilio Extruder?

No metal teeth biting into the filament, but soft rubber belts pushing the filament down to the nozzle. This is the concept of the Papilio: An Extruder which uses regular 2GT timing belts instead of metal gears to feed the filament.

You may wonder why we need an extruder that uses belts instead of gears? Guess it’s fair to say, that we all know the woodpattern on flat surfaces or the small bites on the filament when the metal gears grab the filament. With the belts there are no teeth damaging the filament, but soft rubber pressing against the filament that also act as a perfect guide for the filament all the way through the extruder. As all my projects the Papilio will be a 100% opensource project.

How does it work?

Powered by a Nema 14 the Papilio uses a 50:10 (or 50:8) reduction to drive two tiny 2GT timing belts. In order to keep everything as minimalistic and cheap as possible the Papilio uses very common parts like the BMG Drive Gear assembly, GT2 Idlers and Pulleys.
The Nema 14 doesn’t make it just super compact but also very light weight – only 140g!

The Papilio uses a Dual-Drive System so the filament gets pushed on both sides by the two timing belts.

Alpha Testing – How can I try this out?

The Papilio is currently in a closed Alpha-Testing with limited/controlled access to the files. This is to get a controlled launch of the Papilio and improve the design before it gets published. If you want to join the Alpha-Testing I highly appreciate your willing to help.
As for all my Beta testings the Alpha-testing will happen on my Discord – check the channel “How to become a Papilio Alpha-Tester” for more information. Note that the release is limited to given number of testers to keep things organized.
Click on the icon to join my Discord Server:

Are there any limitations?

  • Using the BMG drive shafts makes the assembly quite thick. While the extruder body itself is only 21mm thick, the shafts stick out additional 7mm, giving the whole assembly a thickness of ~28mm. Using custom hardware or (SLA) printed gears could make the assembly even thinner. Due to the thickness the Papilio doesn’t work with the popular Voron Stealthburner at the moment – I’m using a Papilio with a Stealthburner with different hardware already, but this isn’t part of the Alpha release yet and still needs some fine tuning.
  • The Nema 14 is a small but not so powerful motor and at the same time the belts add a lot of friction. Therefore a transmission was needed that generates the required torque to push the filament. While the LDO1 Nema 14 often gets called a prototype the Moons Nema 14’s are a very different story – small but powerful. With the 10T Moons Nema 14 I get a flowrate of ~12-13mm³; with the 8T Moons Nema 14 ~22-23mm³ depending on the filament (tested with a Rapido UHF). I’m currently working on different Papilio versions which will use different and/or custom made hardware to get a higher gear ratio and therefore more torque. I will keep you updated!
  • So far my testing with flexible filament has been semi successful – the open design of the extruder gives the filament too much room to escape (Not only the front but also the back towards the Nema). I already have ideas how to solve that, but so far I mainly focused on daily printing with non-flexible filament.
  • Since the Papilio is a printed extruder you need somewhat of an accurate printer – nothing too crazy, my Ender 3 printed dozens of Papilios already but you should be familiar with tuning procedures like this: Ellis tuning Guide.
    1 Therefore the LDO Nema 14 is not supported.

What’s planned already?

  • Papilio Pro: A version which uses custom made hardware to get a light weight and very powerful extruder
  • Papilio Lite: Inspired by the LGX lite this version will use some of the LGX internals to gain a smaller formfactor for Stealthburner
  • Papilio Bowden: Of course there will be a bowden version of the extruder for your bowden setup – powered by a Nema 17

How the Papilio was born


Papilio Pro

The Papilio Pro – we will see what the Papilio flies us to.

March 2023

The Papilio Extruder!

And here are we now: The Papilio Extruder pushes filament for many weeks now, and it’s time to build your own!

December 2022

And then, after hundreds of hours a butterfly landed…

Okay no harmonic drive, what then? Something cheap and simple! So I went for the well known and easy to get BMG Gear assembly, some tinkering and I made a butterfly, such a piece of art! After so many hours, days months I finally got a working extruder in my hands.

September 2022

Such a fail

Months and hundreds of prints later I was at a point where I realized that the harmonic drive Idea looks super cool but isn’t easy to setup and not strong enough, even though this means the last moths have been a waste of time… that was incredibly demotivating. Okay I learned a ton of things and the harmonic drive idea is not dead yet but we need something more simple.

May 2022

Revision of the Revision of the Revision of the…

Hell I spent so much time on improving this little thing. While the Harmonic Drive looked super cool, it wasn’t strong enough. I spent months on improving the design over and over again… and again.

April 2022

Okay okay we need a reduction

The Nema 14 definitely needs a reduction, but what type of reduction? Something crazy, something special, something that just as cool as an extruder running with belts. I know a Harmonic Drive! But a harmonic drive with a GT2 belt as flex spline, the Belt driven Belt-Extruder… I thought so.


Proper Extruder

Damn JÓN! You made a thing! 😀
After the Creality Contest I focused on the Belt Mod and the BeltBox Extruder went a little out my mind, but in the meantime Jon made a piece of art with his Proper Extruder. This brought my attention back to the project and I decided also to go for a direct drive system.
Image source:

August 2021

Flip the belt

For the next design I decided to use the smooth side of the belt and let me tell you, this worked so much better! But how do we get both belts in sync? Still a problem to solve…

Somewhat later

A few versions later

A few version later, some improvements and there was actually a somewhat working design, we’re getting closer! But the toothed side of the belt still feels too smooth and slippery… we have to improve that!


The BeltBox Extruder started

After the great success of my Belt Mod for the Ender 3 I thought “What else can I add belts to?”. Since I had quite some extrusion issues at this time I decided to build a belt driven extruder… But how? Back then it was a very basic design without any reduction, just a few pulleys and idlers. And it used the toothed side of the belt so the teeth of both belts interlock and act like a dual drive system.