85 data were collected on Sep 16/17. Only a very few are not point-like, and none of these seems obviously discrepant. They are shown below displayed in two ways - below-left using the VAX/FIT7 package, and below-right using the 'call a9' routine of TPOINT :

The strong trend of daz residuals -vs- elevation (left plot, upper-center) does not show so obviously in the TPOINT plot, where the most obvious systematic is that of dS -vs- azimuth (upper left panel). Raw rms residuals in (daz,del) are (3.2",2.5"). The data were reduced with TPOINT. Using just the usual 7 TPOINT parameters reduced the raw residuals to approx 2.3" in each coordinate, but did not impact the dS-vs-azimuth systematic :

This systematic was removed by introducing a harmonic term HESA [ elevation is a function of sine(azimuth) ]. - and the residuals fell to approx. 1.9" in each coordinate.

There is currently no facility to apply the HESA term at the telescope, so the best 7-parameter model was installed at ~15:00 HST on 17 Sep.

Collimations for the heterodyne FEs were updated at this time also.