Skip to main content
Submitted by lumsk on Tue, 11/21/2000 - 21:36

I have a DMC1820 controlling an XY table. There is a load encoder of 1um/count connected as main encoder. The motor encoder is connected as aux. encoder which have a resolution of 0.33um/count. I make use of the daul loop feature of the controller by using the DV command and start tuning the system. As described in the manaul, I start with a small KP value and increase KD until the system becomes unstable. I then decrease KD to a stable value (200) and then increase KP. My problem is when I increase KP (up to 14), the system soon becomes unstable, far before the required accuracy can be reached. The positon error is around 50-150 um. I have estimated the backlash between the motor and load, which is about +/- 5um.

Q1) Is the backlash too large for using the daul loop?
Q2) I have just go through your web-tutorial on the daul-loop topic, which introduce the improved dual loop. It seems just address my problem. Is the improved daul loop implemented on my controller broad (it is purchased on DEC/99) ? If not, how can I upgrade the broad?

Comments 1

Galil_Todds on 12/04/2000 - 11:38

There is no single number which represents 'too large' an amount of backlash for the controller. The system characteristic which may affect the stability more than the backlash is the load matching. For stability in a system, we recommend the inertial mismatch between the load and the motor be less than 10:1. Even with very large backlash in a system, if this inertial matching is good we can stabilize the system. Conversely, if the inertial differential is great even a small amount of backlash will be difficult to stabilize. So that is the first calculation I would recommend making in order to find how well your system can perform.

The advanced dual loop can offer better performance over what the standard dual loop offers, but this is with a well balanced system.