The pointing model for the INT has 17 terms, comprising the standard
errors of an equatorially-mounted telescope (offsets in HA and Dec,
collimation error in HA, misalignment of the polar axis and
non-perpendicularity of axes), together with an empirical (Fourier series
and polynomial) model of flexure in the telescope structure and large-scale
errors in the Ferranti encoder readings. If pointing
tests are made sufficiently frequently in order to update the pointing model,
the rms error in the absolute pointing of the INT should always be
better than 5 arcsec: values between 2.8 and 4.5
arcsec have been measured; a small residual oscillation at about 1 Hz
persists in Hour Angle.
The data necessary to produce pointing models for the telescope are
logged by the PE3220 running the task POINT. Data are logged by this
task when the LOG button on the control console is pressed. Note that
this button lights up while the logging task is active and the telescope should
not be moved until the light goes out. The pointing program creates four data
files on the PE3220 system disc. The first one is called POINT.DAT and must
not exist when the program is started. Either delete them or rename them if
they are to be kept. The extra three files created at the INT are called
POINT1, POINT2 and POINT3 (all .DAT file extension). They contain details
of the individual Ferranti encoder readings alone so that checks can be made
for any systematic drifts in the Baldwin encoders and also to provide data from
which it may be possible at a later stage
to model the individual Ferranti encoder errors and separate
them from global pointing effects such as a polar axis misalignment. The
main data file, POINT.DAT, contains the name of the observed source,
the Baldwin and Ferranti encoder readings of RA and Dec and the Local Siderial
Time of the observation. A grid of objects suitable for use in the pointing
test is contained in the file PTGRID.CAT (see Appendix ), which
resides permanently on the system disc.
To run a pointing test with POINT:
I(NPUT) ptgrid
POINT INT
If the computer responds with a message to the effect that file POINT.DAT already exists, and that therefore the task can not be started, rename that file with:
RENAME POINT.DAT, POLD.DAT
and similarly for the other three data files:
RENAME POINT1.DAT, P1OLD.DAT , etc.
Do not delete the files, since they may still be of interest. Further, report this action to the Software Manager.
NEXT sourcename
ENDPOINT
This will stop the program and print out the data files. Subsequently, press the on-line button in order to re-initialise the terminal.
AUTOPNT
The pointing procedure has now been completely automated and the procedure AUTOPNT should be used at Cassegrain. The acquisition of pointing stars is made with the autoguider probe; the stars are put at the SNAFU reference position. Therefore, the guide probe has to be told where the reference position is by placing a star on the reference point and acquiring it with the autoguider in `probe' mode. The procedure is now as follows:
(Alternative to steps 1 and 2: drive the probe to the centre position. The idea is just to put the probe at the reference point.)
AUTOPNT INT
The telescope will work its way through every available star from PTGRID.CAT from North to South and then back again. When it has acquired each star it will guide for a few seconds and then log that position before moving on to the next star.
At Prime focus the situation is more complicated because it is impossible to move the autoguider to the rotator axis of the telescope. The Prime Focus geometry is set up such that if a star is on the optical axis of the telescope pressing BEAMSWITCH B will move the telescope so that the star is centred on the autoguider in its parked position. Thus, when using AUTOPNT at Prime you have to press BEAMSWITCH B so that the stars fall within the field of the autoguider. Unfortunately the Telescope Drive Software has a feature that at the beginning of every slew the BEAMSWITCH automatically reverts to the NOMINAL aperture. Therefore, when running AUTOPNT at Prime it would be necessary to press BEAMSWITCH B every time the telescope begins a slew to the next star! This would be a cumbersome exercise, but a simple way around it is to change the NOMINAL aperture position before starting AUTOPNT as follows:
ADV/TEL
The aperture position is given in arcseconds.
INPSCF
@AOXNOM
@AOYNOM
Important: make a note of the current values.
AOXNOM new-aoxnom
AOYNOM new-aoynom
where: Tnew-aoxnom = current aoxnom + beamswitch B.xnew-aoynom = current aoynom + beamswitch B.y
PRINT filename
It is worth checking that the program is working by printing the POINT.DAT file after the first 5 observations have been made with:
PRINT POINT.DAT
A listing should then be output to the lineprinter standing in the Instrument Computer Room, adjacent to the Control Room. Check that it is switched on-line. If there does not appear any output, go to the TCS console and try:
TASK SPOOLER
CO(NTINUE)
If that doesn't work, type
STARTSPL
If this still doesn't work write a `Software Error Report' (SER).