Still working on repairing the foam cutter. This is getting to be as bad as the first time. I spent ~100hours writing the program and troubleshooting, plus learning how to use CAD for the airfoils and translating into something the cutter would follow to more time sorting out the original driver board. It turned out that the output byte for step/direction had to be fed to the driver in reverse order, bit by bit. That took a lot of troubleshooting to figure out.
I tried open source now, didn't really work. LinuxCNC is mainly for milling machines run through a parallel printer port. It supposedly has an addon/addin for a four axis foam cutter, but I couldn't even get the Linux up and running. Very hardware dependent- if someone hasn't written a driver for it it won't work. Plus a ton of different versions/flavors with all sorts of interdependencies. Windows is not great but Linux isn't really better, unless you are a real nerd. Plus everyone wants to use Gcode, which is fine for a mill or lathe, nobody seems to be doing EDM(similar 4 axis to a foamcutter) and it's easier form me to scale the templates(usually larger/smaller than the foam) in CAD than in yet another program I'd have to learn.
I've finally found a program that at least will read in the templates I use(after modifying them to 1in. long). The 1in. comes from the low speed airfoil project Mike Selig(now Professor) started at the University of Illinois Urbana-Champaign. I've been working over the years with several SAE college program students doing the GoKart event(downforce wings). Scaling airfoils up and down using the typical spline curves can result in significant 'wavies' in the airfoil, particularly around the leading edge. I've always used polylines. They are easy to tweak for getting the airfoil exactly the way I want, with spar slots and other things like internal cutaways. Some CAD programs will save the file in a simplified text file that produces polylines as two column text so its easy to translate into foamcutter language and be easily able to cut multiple panels in one block.
The other big problem has been hardware and the either $$$$ or poor documentation. I've had to spend hours trying to figure what will work with what.
Looks like the homestretch now though. Just two more wires to go until I can see if it's going to work!