Quote Originally Posted by boonie View Post
Yes, it looks like the servo drives will be implementing localized closed loop control over their respective axes (for their individual steps). Since the axes are not linked I believe they can come out of sync, leading to elliptical error when cutting circles. I'd think that the way to correct this would be to collect all the F/B in the controller and compare the two axes doing the circle to each other and the circle vector... or go really slow, in tiny steps to insure that there is no overshoot/fall-behind... which is how you'd run a stepper motor system. (?)

From the controller's point of view, isn't the only difference between a stepper and servo system using this style of servo drive that the servo drive can notify the controller (through the err/res channel) when the servo misses a "step" (or several...) and is no longer on course, allowing for the controller to kill the program?

----question:

Am I thinking this through correctly?
Closed loop is closed loop. Trying to subdivide it into different stratas and then to specify that only one type is "TRUE" closed loop is not honesty. The step and direction servo drives are very similar in operation to other kinds of servos, the main difference is that the electrical connection between the drive and the controller system is public and open to inspection. Other servo systems rely on black boxes and technical jargon to keep public inspection and comment from affecting opinions. They are also much more difficult to tune and trouble shoot.

There is a misconception that servo control loops are somehow inaccurate. They are not, they are required to fault at a preset error value. The resolution of a servo system can and should be designed to be at least an order of magnitude more precise than the target machine. For example if a controller faults when it is out of position by 125 steps, the resolution of the system is designed so that 125 steps is less than the expected error of the machine. A Bridgeport in good condition can sometimes hold a .002 inch tolerance. If you set .0005 inch as your max deviation from commanded position, any servo fault will be considerably more accurate than the machine can accomplish. At that point your oval circles can likely be attributed to the usual suspects, mechanical backlash and end play.

Fred Smith - IMService
http://www.cadcamcadcam.com