Sunday, March 28, 2010

Mill: lessons learned on milling circuit boards...

In one of my last posts I broke out some reoccurring problems I was experiencing on a consistent basis while trying to mill circuit boards. These six problems that I identified have all been solved. Here are some pictures of the finished board.

I also prepared a video last week of the whole process.

Now, on to how these problems were solved, in order of simplicity.

5) Drifting Hole

This was easily fixed by enabling the tapping feature of pcb-gcode.

2) Bad Motor Coupling

Here is an exaggerated image of the problem I was dealing with.

Because the coupling was off by only a small amount I decided to just take a file to it, which removed most of the symptoms and only took half an hour.

3) Not Round

4) Incomplete / Grounded Pad

6) Inconsistent Traces

All of these problems can be attributed to the part that connects the y axis threaded rod to the bed.

After closely examining the entire machine while running small quick movements I noticed it wobbling. I doubled the size of the part and the number of screws. Now it doesn't move.

1) PCB Thickness Tolerance

I am releasing the beta software I am using to add probing routines to the gcode produced through pcb-gcode. I'm calling it PcbSubtraction. I hope to extend it and add more documentation over time.

Saturday, March 20, 2010

tv: Open Source Circuit Boards at Home

Season 1 : Ep. 1 | 9:55


Have you ever wondered how to make double sided circuit boards at home and without all the chemicals? This episode shows you how. I'll show you how to mill away the copper you don't need, leaving a complete, drilled and scored printed circuit board prototype.

If you want more details on the mill used in this episode, check out my milling blog at:

For any unanswered questions, or any feedback you may have, feel free to add a comment to this episode.

Tuesday, March 9, 2010

Mill: preview of problems solved...

When I have more time I will be writing in depth on the solution to each of the problems identified in the last post. As a preview, here is an image of my latest go.