RAMPS Rebuild Keeps Robox 3D Printer Out Of Junk Bin

A 3D printer is a wonderful invention, but it needs maintenance like every machine that runs for long hours. [Rob Ward] had a well-used Robox 3D printer that was in need of some repairs, but getting the necessary replacement parts shipped to Australia was cost-prohibitive. Rather than see a beloved printer be scrapped as e-waste, he decided to rebuild it using components that he could more easily source. Unfortunately the proprietary software and design of the Robox made this a bit difficult, so it was decided a brain transplant was the best path forward.

Step one was to deduce how the motors worked. A spare RAMPS 1.4 board and Arduino Mega2560 made short work of the limit switches and XYZ motors. This was largely accomplished by splicing into the PCBs themselves. The Bowden filament driver motor had a filament detector and an optical travel sensor that required a bit of extra tuning, but now the challenging task was next: extruding.

The printer’s new custom hotend.

With a cheap CR10 hot end from an online auction house, [Rob] began modifying the filament feed to feed in a different direction than the Robox was designed for (the filament comes in at a 90-degree angle on the stock Robox). A fan was needed to cool the filament feed line. Initial results were mixed with lots of blockages and clogs in the filament. A better hot end and a machined aluminum bracket for a smoother path made more reliable prints.

The original bed heater was an excellent heater but it was a 240 VAC heater. Reluctant to having high voltages running through his hacked system, he switched them out for 12 VDC adhesive pads. A MOSFET and MOSFET buffer allowed the bed to reach a temperature workable for PLA. [Rob] upgraded to a GT2560 running Marlin 2.x.x.

With a reliable machine, [Rob] stepped back to admire his work. However, the conversion to the feed being perpendicular to the bed surface had reduced his overall build height. With some modeling in OpenSCAD and some clever use of a standard silicone sock, he had a solution that fed the wire into the back of the hot end, allowing to reclaim some of the build height.

It was a long twelves months of work but the write-up is a joy to read. He’s included STL and SCAD files for the replacement parts on the printer. If you’re interested in seeing more machines rebuilt, why not take a look at this knitting machine gifted with a new brain.