by lasersafe1 » Fri Jan 13, 2012 1:34 am
Hi Bart,
I did quite a bit of research on Stewart platforms last year. As you have already mentioned, they are indeed capable of superior rigidity over standard gantry machines. Yes, EMC2 is already capable of running them too. But you failed to mention the most important part. In a linear system where a Z axis is mounted on top of a moving X axis which is mounted on a moving Y axis (like the Gantry systems), the backlash and other errors are cumulative. In the Stewart platform, they are independent and once measured, they can easily be removed by software. One of my tasks at work will be to build a highly accurate Stewart platform that is miniaturized to move a Pockels cell with high precsion with X Y Z pitch roll and yaw indepently controlled. It will be about 4 inches tall. I plan on using picomotors with encoder feedback. Rather than worry about control systems, I will use a simple single board computer running EMC2 into which I can serially command positions of all 6 axis. Similar units are sold for $68,000, but I know I can build one for under $5K with sub-micron precision.