One of the reasons I purchased the ShapeOko was to route printed circuit boards. This was my first ShapeOko PCB routing test.
How did it go? Well, disappointingly, not so well. Here is what I’ve done.
I started by downloading the gcode_02.ulp script written by Josef Plasil. I pulled it from the EagleCad website. There is no documentation on this ULP that I could find, but it is straight forward and doesn’t have a lot of options. It appears to do a great job.
With my new ULP in hand (or on disk), I did a quick test layout. Very quick. The layout had no components, just the word “TEST” written in the center. I ran the ULP eager to do some cutting!
I unchecked the drill and cut options. I only wanted to route out the text. Using the standard options, I clicked OK to start generating GCode.
Next I loaded the GCode file into my controller software. Zero’d my spindle at the bottom left corner of my blank PCB. I set the Z axis so it was just touching the top surface of the PCB. Finally I powered on the spindle, took a deep breath, and hit the go button. “It’s working!” I thought, so I relaxed and watched the magic happen. It all looked so promising up until this point. Now I started noticing how things were not aligned so well in the Y axis, and I could see why right away. The left side of the Y axis was dragging behind the right side which was directly driven by the motor. Arrgh.. My relaxation was over.
Googling around I found that other people have solved this problem by driving the Y axis from both sides. I guess I could do that, but I had hoped to use my 4th motor drive output for an A axis (rotating head for a future pick and place machine). Another guy on the forums mentioned extending the shaft of the motor to extend across the table to the other side. It means relocating the Z axis a bit, but I liked the sound of still using a single motor. I don’t think I’ll be doing this soon, but a mechanical tweak may be in this machines future.
Here is the final route.. It looks somewhat like the drawing above.. Right??