Work Process
Design
As with most of my projects, I started by trying out different designs for the machine. In the following pictures you can see the CAD models for some of the ideas I had.
The final design was a combination of these designs. The machine moves in the XY plane by allowing V-wheels to slide along aluminum profiles, and the movement is done using stepper motors via a drive belt.
Construction
The machine consists largely of laser-cut wooden panels. These parts were laser-cut from 6 mm thick plywood and MDF.
The parts were assembled into a frame. With the frame I was able to check that the tolerances between the V-wheels and aluminum profile were good, and that the belt drive from the stepper motors was working.
At this stage I was able to start doing some simple tests on the Y-axis to make sure everything was working as expected. I generated machine code for a simple circuit board and checked that the motors were performing the corresponding movement.
After the Y-axis worked, it was time to start designing the X and Z-axis, and the following image shows a CAD model I made.
Once the design was complete, I started laser cutting the parts. I built and tested the Z-axis separately from the rest of the machine, and once I got it working, I installed it.
The next step was to install in the limit switches.
Once the Z-axis was mounted on the X-axis, it was time to mount everything on the frame. The pictures below show the first time the machine was fully assembled with all three axes.
Initial tests
At this stage, the machine was ready to run some simple test programs. In the following video, you can see the machine executing machine code generated using a CAM program.
Before a spindle motor was mounted to the Z-axis, the machine was tested with an ink pen. The pictures below show the drawings that were made.
Assembly of motor and FR4 laminate
The next natural step was to mount the milling motor so that I could start milling real circuit boards. The motor was attached using a laser-cut clamp. The clamp was designed to be narrow enough so that it could bend, and once the motor was in place, the clamp was tightened with a screw.
Here you can see the very first tests I did. It took quite a few tries before I found the feed rate and depth that were best for my machine. After a lot of testing I managed to find settings that worked.
Now that I had tested that my machine could mill circuit boards, it was time to build a more permanent way to mount the circuit boards to the work surface. I wanted a wooden wasteboard under the circuit board that could be quickly replaced, otherwise the drill would mill into the work surface when milling.
Circuit boards
Here are pictures of the first milling of an actual circuit. It's clear that I still had some problems with the settings for the machine, but it was definitely progress.
After some adjustment, the results started to look quite okay. However, I noticed that the trace width varied quite a bit, even though I had designed them to be the same width. My guess is that the belts from the stepper motors are causing backlash because they are somewhat elastic. The solution for me was to run the machine really slowly so that the mechanical resistance to the drill is as small as possible during milling.
The first real project where the machine came into use was when I built my automatic irrigation system. Both the circuit board for the microcontroller and the keypad were milled on this machine.
