Recent work to re-calibrate the motor which moves the grating in CSHELL has revealed that for some grating angles, the grating position is not exactly reproducible. Although both the motor, which moves the grating, and the encoder, which reads the position, indicate that the correct number of steps have been moved, and the correct grating angle is displayed in the MOTORS window, the grating may not be in the correct position. This can be seen by taking an arc lamp exposure in which a specific arc line is centered on the array, moving to another grating position, and then moving back to the position for the arc line. For some arc lines, this procedure results in a shift in the position of the line between the two exposures that can be several (up to 10) pixels.

The slight difference in grating angle, and hence the difference in the positions of arc lines, should have negligible effect on the throughput or performance of the instrument. However, it does imply that all calibration frames (i.e., flats and lamps) taken for a particular observational set-up and grating setting must be taken before the grating is moved to another setting. In practice, this also means that calibration frames taken during the afternoon cannot be trusted to be applicable to data taken that night unless the observer is using only a single grating setting and does not move the grating between the afternoon calibrations and the evening observations. Observers with multiple grating settings will not be able to reliably apply calibrations acquired before a grating move to observations made after the grating is returned to the orginal setting.

In view of this problem, we strongly advise observers to acquire all calibration frames (flats, darks, and lamps) immediately before or after each object is observed throughout the night, and before any changes are made to the grating position.

John Rayner (rayner@irtf)
Last modified 24 June 1999