The finder flexes with respect to the main telescope by 110 arcsec in Right Ascension and 75 arcsec in Declination. The flex in RA is very reproducible and can be modelled accurately, whereas the effect in Declination is less reproducible and can be modelled only to a (maximum) uncertainty of 45 arcsec.
The finder flex must be calibrated whenever guide stars are to be located using the finder display. A simple linear model of the finder flop as function of zenith distance is implemented in order to compensate for the flop in the finder mounting and allows guide stars to be located all over the sky. The zeropoint of the relation must be determined at the beginning of the night with the telescope very close to zenith; thereafter adjustment is automatic.
The finder flop zeropoint determination is done by positioning a calibration star at a reference position on the Cassegrain TV camera (normally the SNAFU position) or on the autoguider probe at Prime. Subsequently, the same test star is located on the finder display and its position as measured there is used to calculate a zeropoint (offset) between the main telescope and the finder. Because the test star is practically at zenith, the zeropoint can be derived directly, without invoking an approximate flop model.
The values of the zeropoint derived each night should be logged by the Telescope Operator. However, note the effect of the reference position used as implied by the procedures described below.