Support the OpenGov idea to create a "Platform for number crunchers across (US Federal) government" HERE. A small team is building a small pilot and I'm happy to report that R appears on many of the posts. If you like the idea (or even if you don't), sign up and vote.

Rweblinks, by Abhijit: http://files.meetup.com/1503964/Rweblinks.html.

Just joined R bloggers (http://www.r-bloggers.com/). I'm using R to write G-code for the milling maching. This evening (10/25/2010), I'll give a talk on this at the DC-area UseR meetup. Here's a link to the presentation: http://files.meetup.com/1503964/Using%20R%20to%20Cut%20Triangles.pdf

Giving talk at 8/25/2010 DC-area meetup. Here's a link to the presentation: http://bit.ly/cWK4vD

2/10/2010 - After hours and hours of programming and trial runs, I've cut my first triangle and it looks great.

The plot below and to the right shows the first triangle. The small dots are vertices of all 100 triangles. The plot shows the original coordinates (rightmost of the triangles). Next, it is moved so the rightmost vertex is at (0,0). Then it is rotated counterclockwise so that the first cut is horizontal (from 0,0 to about -80,0). Finally, allowance is made for the cutting tool's radius. The green lines mark where the tool's center will travel when cutting the three edges.

To view the intermediate steps, click on the link below.

CNC

1/30/2010 - Milling, at last!

After hours and hours of coding, I had a program that would carve a brief message into a block of wood. There are still some programming issues.

I now understand that the software has two ways of interpreting an arc command. For example, "G2 X3 Y3 I0 J1 could be interpreted as "move clockwise to location (3,3) (x=3 y=3) with the center of the circle at coordinates (0,1)" OR it could mean "move clockwise to location (3,3) with a center that at (3-0,3-1)=(3,2)." The center, of course, can make a big difference!

Right now, I'm using the mill to cut a 25mm diameter hole in 3/8" PVC. This is part of a bracket that is needed to better-hold the carving tool. With only one bracket, the tool is held too loosely and moves a bit when it is cutting.

Here is a movie of the mill and the software doing their job. This is taking a while, because I don't want to put much stress on the single bracket. The tool is moving at a rate of 10mm/min. One trip around the 25mm circle takes about 25*pi/10 = 8 minutes.

January 17, 2010

Today, I learned to solder. There were quite a few wires to solder. Each axis has two emergency stop switches, and each of these had at least two wires. I also put in a panic button. If things go bad, all I need to do is push the panic button to cut off the mill's DC power.

With all the stops wired and in place, I ran each axis until it triggered its stop. It worked! The motors each stopped running. I could no longer move them through the software, but I could turn the screws by hand - but only a few turns until the switches clicked off and the controller locked the motors in place. Turns out these positions (smidgens away from triggering the emergency stop switches) were excellent choices for the axis zero settings.

The other thing I learned today was how to write the control code, called G-code. It was very cool to write a line of code, send it to the controller, and see the machine behave exactly as expected.