The control software is written in FORTRAN (with a minimum of assembler) and is, in fact, applicable to both INT and JKT. The differences between these telescopes are handled by Control Files with system parameters, which are read in when the software is being initialized at the start of an observing session.
The system control is organized as Tasks, which may run simultaneously or separately. The principal task, ``TNHZ'', is, of course, the one that drives the telescope, and, therefore, is active throughout. Other tasks that control the Dome, the Cassegrain Instrument Rotator, or Focus are run only intermittently, obviously because they are not needed all the time, and even then these functions do not need to be controlled to the same accuracy as the telescope itself. Yet another task is run every two seconds or so and drives the Information Display, on which such parameters are shown as the source name and its sky position, aperture offsets, differential tracking rates, the current pointing and focus of the telescope, and the location of the dome aperture.
The main control task for the telescope drives (TNHZ) runs at Hz.
A cycle consists
basically of three steps. First, the actual encoder positions are read and
the time is obtained from the Time Service rack in the CLIP Centre.
Subsequently, the desired encoder values (for that moment) are calculated:
The program calculates the position of a virtual telescope, without pointing
defects, taking into account non-sidereal tracking rates, offsets, handset
demands, autoguider error signals, as well as precession, nutation,
aberration and atmospheric refraction. All
calculations are performed in the proper coordinate system, as specified by
the observer (mean or geocentric apparent). Finally, a parameterized
model of the telescope's pointing (Section
) is applied in order
to produce a desired position reading for the encoders, which is compared with
the true reading. If necessary, velocity corrections are
applied to the drive motors in order to zero the position errors.
For large errors the velocity correction will be proportional to the square root of the error, within the mechanical limits; if the error is small, the correction will be approximately proportional to its magnitude. When a slew to a new object is requested, the slow-motion clamps are initially released and the telescope is driven in quick motion until it is within a few hundred arcsec of the object. The clamps are then re-engaged and final acquisition is made in slow motion, using the high-resolution Baldwin encoders for smooth tracking.
The telescope RA, Dec and Zenith Distance, as well as the universal and sidereal times, are output to the dedicated Beckmann Display on the Control Desk at 0.1 s intervals.