The Prime Focus Camera Unit is now mounted and dismounted
with the probe `park' position defined by a mechanical template, which ensures
a very reproducible geometry from one run to another. It is therefore very
unlikely that the geometry parameters at Prime will have to be modified:
In other words, be sure to have a very good reason for changing any
parameters. If you find something odd with the geometry try first to attribute
it to some other cause. In particular, check that the telescope was initialized
with the correct configuration
file for the CCD chip mounted: either with GO INT, PFRCA, COMPILE
or GO INT, PFGEC, COMPILE (see section ).
The relevant geometry parameters at Prime, as shown in figure
, are:
Figure: Prime Focus geometry parameters
for the centre of the CCD chip. The GD PARK position is at the top right-hand corner of autoguider field. Further, for the position of the rotator axis, we have:
The current values of these parameters can be inquired at the TCS terminal
with the INPSCF task (see section ).
Note that figure
refers to a rotator position of 180.
The sign convention is such that distances measured in x below GD PARK
(as shown in the diagram) are positive, and distances
measured in y towards the left from GD PARK are positive. This means that
AOYNOM and TDATA(12) are normally negative. The x measures are close to zero,
and therefore unpredictable.
Because at prime focus both the detector and the autoguider probe are positioned off-axis, the default beamswitch apertures A and B are defined such as to enable viewing of the autoguider field with the Finder TV camera and the CCD detector itself:
AD(VISE)/T(ELESCOPE)
In order to test and re-determine the geometry parameters, proceed as follows:
We now have determined corrections to the APERTURE B offset coordinates.
The mean of the two sets of measured pixel coordinates gives the coordinates of the apparent rotator centre. The differences between these and the pixel coordinates of the chip centre, converted to arcseconds as in the preceeding paragraph, and converted to radians by dividing by 206264.8, result in the corrections to the current values of AOXNOM and AOYNOM. Update these parameters in the sense that, if the apparent rotation centre is above the chip centre, ADD the difference to AOXNOM, and, if the apparent rotation centre is to the right of the chip centre, ADD the difference to AOYNOM.
GDXCTR = 39.3 (550 + TDATA(11)/0.97)
GDYCTR = 39.3 (900 + TDATA(12)/0.97)
AD/T
and check that:
AP/A equals TDATA(11) , TDATA(12)
AP/B equals TDATA(13) , TDATA(14)
TV\?OFFSETS
and follow the prompts (Beware: defaults are not accepted; you must enter values, even if they are unchanged):
You should now be able to find guide stars in any orientation of the rotator. If you can't, report it to the Astronomy Group.
All default values are loaded automatically at startup. If you are sure the new
geometry is correct, inform the Software Group, who will enter the new values
as defaults in the proper `.STD' configuration file. Figure
shows the current geometry of the INT Prime Focus Unit.