Laser4DIY - Laser Structuring for Prototyping PCB - Beta Test Phase Build Log

Laser4DIY is an Open Hardware project by FabLab München. It’s a laser structing device for prototyping PCBs i.e. it’s a laser “engraver” that directly removes copper from blank copper clad boards using laser ablation (instead of milling or etching). For more details see

The project is currently doing a (closed) beta testing phase with funding from Prototype Fund Hardware. We have 2 groups trying to replicate the device from the published documentation, one directly at the FabLab München (Munich, Germany), the other at the Das Labor in Bochum, Germany. At the same time we are working on the documentation, adding missing parts and improving the existing ones.

This thread will be used to give updates on the build progress of both teams. Feedback (technical issues, open questions, missing information) will be given by opening Github issues in the Laser4DIY repositories

All the best

Andreas Kahler
Chairman, FabLab München e.V.


FabLab München Team, Update 24 Oct
Attendees: Ralf, Jürgen, (Andreas)

  • First meeting, starting the build
  • First module: XY table, most parts were provided (laser cut aluminum, fasteners, motors, electronics)
  • 3d printed PLA parts (2 parts), laser cut MDF parts (2 parts)
  • separated laser cut parts, cleaned them
  • noticed issues with aluminum parts: some holes were not laser cut, will need manual drilling

Hello from Team Bochum,
unfortunately, we were a little lazy with out documentation, so here a small recap of our last sessions:

22.11.2022: First unboxing session with unboxing and assembly of the both stages

While assembling we ran into our first issue: The no-backlash-nut had a different dimension than the regular nut. Everything has been solved and was reported on Github:

02.12.2022: Assembly of XY Stage

Sorted out strange hardware issue: When the boards were assembled the GRBL Firmware stopped working. It turned out the motor shield pressed the reset button on the Nucleo-Board. Both are axis are working now.


Hi there! I’m André, also from Team Bochum.
Due to colds goring around here and some other issues, we could not meet for some time.
But yesterday, we made some progress. Let me tell you what we did.


We started building the supply unit with the power supplies and the cooling. Construction went smooth until we accidentally stripped one of the threaded couplings from the radiator. It seems, that the metal is really really thin around there and if you thread the hose-coupling in to deep, you rip it apart.

But with some plumping-supplies we could get around the situation.
So we suggest to add a hint to the documentation, that these fittings have so be screwed in very gently.

1 Like

(Next post due to limitations for new users.)

We wired in the emergency stop and all the connectors on the front panel. After that we followed the wiring-diagram and wired up all the power supplies and the fuses.

We put everything in the 19" rack.
That is the current status, we hope to make more progress soon. We will keep you updated!


1 Like

I added a note to the documentation. Thanks!

I was presenting Laser4DIY at the Forum Open Hardware last week in Berlin. Documentation, videos and photos from the event can be found here (German only): Publikation und Ergebnisse des FOH23 | Forum Open Hardware

The supply unit from team bochum is ready. We testet it for half an hour an erverything was dry afterwards.

1 Like

Laser4DIY is on the Hackaday podcast! Elliot and Dan are discussing it starting at min 29.


PS: small correction on what’s discussed on the podcast: our pump diode driver is running in constant current mode. We do not pulse the pump diode (yet). But yes, the magic of the q switch works with a cw pump diode too, and we indeed generate pulses ~2.5ns with almost 50kW peak

1 Like