Re: [CAD_CAM_EDM_DRO] Announcement: Mach3 is now closed loop!

R Rogers wrote:

>Jon, There is pulse data out and data back to prove position and this is compared at the control then dealt with accordingly. What is not closed loop about that?
>
>

because it is not used immediately to adjust imstantaneous velocity. It
is only checked 10 times
a second, according to what I've read. A LOT of error can develop in
1/10th of a second.
Still better than no checking at all, but my system with EMC, for
instance, does the following
error calculation in real time, ie. at the servo update rate, nominally
1000 times a second.
But, ALSO, it is adjusting the velocity it sends to the step rate
generator 1000 times a second,
too. This doesn't really do much for a stepper system, but can help a
servo system stay closer
to the commanded trajectory. (If 1000 times a second isn't fast enough,
the servo rate can be
turned up easily to at least 10 KHz with a little faster computer.)

>Because it operates differently or may not parallel some conventional process or thinking or by someones defintion does not change the facts. The loop is closed back to the control.
>
>It works and it works very well.
>
>Isn't EMC a step software? Doesn't it halt the program if the following error is too large? The end result being the same, what is the difference really?
>
>

No, EMC is not inherently step software. My system, with either the
PPMC analog servo interface, or
with the PWM controller, has no steps ANYWHERE in the system, in
software or in hardware! (Other
EMC systems, using the Servo-to-Go or Vital Systems boards are the
same.) And, with any of these
systems, by default would cammand an e-stop in one millisecond of
exceeding the following error
limit. (EMC also has the ability to make the following error limit
proportional to commanded velocity,
so that greater error is permitted for rapid traverse than for cutting
moves.)

Jon