we now have a nice setup to intercept and analyse serial communication between the driver and the laser cutter. But for now we stopped working on reverse engineering and decided to actually fire the machine up. We figured out cutting wood and pwnschlager did put a lot of effort into figuring out how to create a decent engraving on cheap plywood.
After we verified it worked we started using a wine configuration i created based on lutris to run the driver in linux - works like a charm. Anyway, we hesitated to do actual laser jobs because we found out that you are supposed to wear protective glasses while operating the machine and also because we weren't done reading the manuals. which we really should do before we play with such a complex machine.
After again spending a bit of time trying to send a job via the network i gave up and decided to see if it works via usb. And indeed, it does. In the meanwhile meks joined me and we kept working together on creating a reproducable setup to create/modify/send/sniff/replay laser cutter jobs. The usb presents itself as a usb serial device which is pretty convenient because we can easily sniff serial traffic.
I spent the day trying to figure out how to send jobs via network to no avail. I can see that the driver sends a single 6-bytes udp-packet to the LC but never gets an answer. When you active the discovery the software sends that exact packet to every single address in the subnet. Different cables directly connected to my laptop (in conjunction with tools like ethtool) ensured that it wasn't a fault on the physical connection.
This is about our forward- and reverse-engineering efforts revolving around the BRM 90130 laser cutter at the Metalab
chaos.social – a Fediverse instance for & by the Chaos community