10
« on: March 02, 2011, 02:14:20 PM »
Hi Jen,
To test my repeat accuracy I ran the x axis 50 times back and forth over a 20" run and it came back after the 1000 inches less than .001 off of the zero position. That doesn't really test the accuracy at the exact moment of an acceleration/deceleration and I'm not sure how I would do that. With the motors locked I can stretch the belt about .005 in either direction before I hit a really solid resistance point. My belts aren't kevlar so that may be an improvement I can make should I someday find the current accuracy insufficient. My belts also have very little mass to them and I would never consider them to be contributing to a problem with inertial vibrations. I imagine though that if this is the situation in your case you could put one or several idler pulleys along the backside of the belt to kill the harmonic tendency you're now seeing. Also even though I'm new to this and can't speak with much authority I have noticed that some G-code files are more ridden with inertial changes than others. There seems to be a bit of an art to generating files that are efficient at producing the highest quality part. G-code generating programs are at best only working with a 'typical' geometric input in mind. Another point I might mention is that I'm using a Smooth Stepper to generate the pulse stream (currently set at 64k) and this pulse stream is most definitely smoother than the default parallel port output. Whether the quality of the pulse stream is contributing to your difficulties would be another item to consider. If the laser unit doesn't run G-code then you won't be using Mach and I'm going to assume that it has it's own pulser which may be the source of the smoother result you're seeing in that machine. You might want to find out what the spec is for the laser's pulsing engine just to be informed about it.
Thanks and I hope some of this is helpful. By the way I can't imagine using a 1/64" bit.
Chris