Hey guys, I can proudly claim that my OX is finally built. I used uncut 1m extrusions so the working area is roughly 800mm by 730mm. It was an awesome experience! Now, I calibrated the X/Y axis measuring the real movement and I settled to a value of 59.85 "travel per revolution" (using a TinyG board). What would be the scientific way to validate that number?
After a good night of sleep I realized that the formula should simply be: travel/revolution == perimeter of the pulley The description (http://openbuildspartstore.com/gt3-aluminum-timing-pulley-20/) says that the diameter is 18.3mm so the perimeter is 57.49114. It is close to the value I found but it feels not close enough. I will test it and report. In case it will not work, is it possible that the real diameter is different?
More than likely the radius you are looking for is to the centerline of the belt. As in, one revolution moves how much belt? The inner face of the belt will be compressed at the pulley and the outer edge is stretched so the only reliable, constant dimension would be at the centerline. But all that is theoretical. The best route would be to mount a pencil in the machine and tell it to draw two lines 500mm apart in both the X and Y directions and then scale the results as needed.
I think you are right. Adding half of the belt width (0.75mm) to the diameter of the pulley gets me to a perimeter of 59.84734 which is awfully close to my manual measurement. I think that I found my magic number, thank you very much sir.
Holly guacamole, Batman. I never thought of calibrating a machine this way. I guess you could do it but I doubt you'll ever get the precision you are looking for. I use Mach3. In it there is a place for setting up your machine. You choose an axis and tell it to move a certain distance. You then measure the distance and put it in. The program calculates the compensation. You repeat it until the distance commanded and the distance measured are the same.
