Vern, Fil, Nutsinee

WP#5085

From the rotation stage random walk study we found that CO2Y rotation stage is off by ~30 deg when it's not behaving properly. Today we did some more test to narrow down the source of the problem and attempted to fix the problem. Turns out swapping out the rotation stage isn't as straight forward as I first thought so we went for less complicated approaches.

First we went out to the Ethercat chassis and look at the CO2Y rotation stage signal with an oscillascope while running the random walk script. The signal looked fine (Vern has the data saved in a floppy disk).

Then we swapped the sensing and motor cables such that CO2X software controls CO2Y rotation stage and CO2Y software controls CO2X rotation stage. This time CO2Y rotation stage behaved nicely and CO2X jumped. Attachment 1 shows CO2Y requested angle vs. measured angle and the differences. The data took after cable swap is plotted in cyan. Attachment 2 shows the same thing for CO2X. The jumpy red plot is the CO2X rotation stage data took after the cable swap (controlled by CO2Y software). Attachment 3 is CO2 power plotted against the requested angles. We concluded that the problem must be related to the software.

We planned to replace two things in the chassis: the CO2Y interface card (D1300131) and the dual motor controller (EL7342). After swapping the cables back in place, we replaced the CO2Y interface card (only). We also found lose connections in many places inside the chassis while doing the work so next Tuesday that chassis gonna have to come down so Fil can take a closer look. Anyway, swapping the interface card seems to have made the problem worse. Both rotation stages are now very jumpy, they have trouble going to negative requsted angles, and the measured angle goes beyond +-90 degrees. We were able to put the both CO2 power back to where they were before we started our work this morning but I wouldn't commission them at this point. The work is to be continued.