Building an OpenSource 2D milling machine

Recently we have build an OpenMarker milling machine for our Corporate Lab.
One could buy all parts independently, print the pieces etc… but we bought it directly from the designer. Our is A3 sized. Lots of fun ahead!


All pieces were set. We are ready to start!

As for Google Carboard, the instructions are not really clear for those like us who are used to extremely detailed and poka-yoke style procedures. [My best quote: « to build the left-hand part, do like the right one, but opposite »] During all the building process we have written our lessons learned and will share it later on on the WIKI of the project (in French).

Most of the time, things go well and the different steps are easy to figure out and overpass.

Mechanical parts

img_20160915_135738Unboxing all pieces of the OpenMaker  Milling machine was the first fun step of the process.







Setting up the workplace.



Our workplace is wide enough to work at a good pace, 2 people at the same time.


We were very happy to see that after only 2 hours we had already build 2 main parts. Lots of work remaining but so happy to see the first results.

In parallel we worked on the software side: firmware, GUI and G-code generator.

  1. Firmware: We downloaded the original firmware and started to work on it, first to understand it, then to upgrade it if possible and if we have good ideas…),
  2. GUI: We downloaded the GUI for the Openmaker machine, we will see later what we will can do with it,
  3. G-code generator: For our 3D-printer we use CURA. However it is impossible to use it with our miling machine so we will use Pycam.

Building the milling machine is very easy even if you could need some tools you don’t have.


After 4 hours of work, mechanical part was almost done. Only the Z-axis delayed us due to some interaction among screws and axis.

Our workplace after work is contained to a samll part of the Lab in order to deploy it quickly when we will be back while not disturbing other users



The instructions are sometimes approximative so we had to redo some steps. After all mounting the Z axis has been the toughest part of the mechanical set-up. Brocken 3D printed parts, uneven screws, bad mounting… At the end it seems we managed to it no so bad

Electronical parts

Our milling machine is controled by and ARDUINO UNO card. It is mounted on the side of the machine and coànnected to a CNC shield. It will help us to control the 4 motors.

Later we will 3D print a protective box to avoid being polluted by cutted parts.img_20160921_094300-2img_20160921_0943072-2








Endstops are some of the most important and tricky-to-set lements of the I&C of the machine



Connections are kind of easy but still requieres high attention as always with electronic parts.
At the end, our machine looked good.


Last tunings

Once the machine is ready, you have to power it on. As everyone knows, powering on a machine for the first time might be tricky. For us it was as a white smelly smoke suddently went out of the power supply. Quick reaction, power off… The power supply was just defective… (hopefully we had a second one). So take care when you first switch it on.

Small adjustments esp. to set the zeros and achieve a smooth and regular movements of the tool were fast and easy. We were then ready for our first hole!img_20160926_140946It tooks us a lot of time to have the milling machine properly machining this wood part. We had mainly issues with what can be seen as an overload of the Arduino, loosing commands and interpreting in the wrong order Gxx orders.

As a conclusion

It took us almost 6 hours to build the mechanical part of the machine and 2 extra hours to set it up (setting, soft integration, safety checks, first holes…).
Our first feedback is that with a little extra settings and tunings we will be able to have a working milling machine. Precision is astonishingly good and quality of the cut is really fine. Cutting material is a balance between your displacement speed and the rotation of your tool, this has to be carefully set but it seems to produce high quality cuttings.

Our problems related to the I&C of the machine are still pending but we are working on it with the manufacturer (W7 in the new outlaw 😉 ). It empeaches us to make big objects.

I will post another article about what are going to be the next steps:

  • First real pieces,
  • Hacks of the machine (how can we improve the design of the machine)
  • Extra development (we have some Python Master here so we will see what they can do to enhance, esp the UX and UI of the software)

Thanks for reading!

Laisser un commentaire

Votre adresse de messagerie ne sera pas publiée. Les champs obligatoires sont indiqués avec *