After the problems with the prism slide in July 2001 the slide itself was cleaned and relubricated. Torque measurements at different positions along the slide were recorded using the "cshow" command:
cshow -s esi -timestamp prismtor prismraw
The resulting data stream contains near-simultaneous values of the prism position (PRISMRAW) and torque (PRISMTOR). Similar measurements were made (both before and after preventive maintenance) on the other two slides (low-D prism and imaging flat).
Below are shown the torque values as a function of slide position for two configurations: a reduced slide speed (in the dashed red line) and the post-preventive maintenance normal speed.
Note that the normal operating torque on 18 Oct. 2001 approaches -3, the nominal limit. The +3 limit is not approached at the rotator angle tested. Other rotator angles would show different characteristics, and this angle is probably close to the worst case. Unfortunately it is also the default rotator angle for ESI.
It would be interesting to see similar data on the prism torque measured in the lab before shipment from ESI. This information may be locked away in the waiaha telemetry database, but Keck has no knowledge of how to access that database.
The absolute upper limit for the Galil controllers is +/- 5, and it would seem prudent to increase the software limits to +/- 3.5 to avoid the messy situation encountered in July 2001, in which the software limits caused a scramble to decrease the maximum velocity and caused loss of efficiency on sky.
The torque on the imaging flat slide does not come close to the limits (+/- 3). The Low-D flat shows similar good performance across its range.
