I have been working on cleaning up my gcode test scripts and thought I would go ahead and share them here. Among other things I cleaned it up so that we can accurately time how long each cut takes. This is written in C++ and compiled with VC 2017. You can download the code here.
I'm interested in working out if running slower is better than running fast and taking multiple passes. To that end I have written a routine that will cut a series of lines at different speeds and different number of passes, all while logging out the time . The idea being that we can run the test in our chosen material, and then pick the speed and pass that cuts through the material the fastest.
In order for this to work well we need to make sure that our acceleration settings are tuned up properly. If they are set too low then the machine will never make it to the requested speed unless you are cutting very long lines. SanSmart has a nice writeup on all of this available here, you can read the two other parts of there writeup here, and here if you are interested in the whole guide.
I'm hoping to get a chance to run this tomorrow, we will see then if it is a worthwhile test. Hopefully it runs faster and more efficient than the normal grid of squares test that varies power and speed. In this case were running at 80% power because it has the best potential for cutting without over driving the laser (extending the life of the laser) so we don't need to vary power as well. Although it may be interesting to vary power on my balsa wood piece that caught on fire, that is another experiment for another day.