The AAT control computer system (CCS) uses an Interdata (now Perkin Elmer) Model 70 computer with digital and analogue interfaces to the telescope drive and to other on-line systems including the dome, windscreen, telescope focus drive, and the acquisition and guidance system. Until recently, only this computer could directly control the telescope, but some functions such as offsetting and mosaicing are now being incorporated into the instrument control software run on the VAX computer.
The CCS has several interfaces to the outside world. The main one, called the DIGMUX, for DIGital MUltipleXor, is a general purpose I/O system. Most of the buttons and lamps on the control panel are sensed or controlled via the DIGMUX. Other specialised interfaces are used to talk to various bits of the telescope, for example the autoguider and A&G unit each have special links. There is also a serial link to the VAX computers.
All timing for telescope control operations is derived from the observatory timing system, and is mean sidereal. The programs are synchronised by interrupts, the main ones being at 20Hz, 10Hz, and 1Hz. The timing system uses an HP105A 5MHz quartz crystal oscillator as its basic reference, and universal time (UTC) is checked periodically against standard time broadcasts. Local apparent sidereal time (ST) is derived from the HP oscillator.
The telescope has two sets of encoders on each axis; an absolute and an incremental. The absolute encoder is geared down from the final drive gear to give a resolution slightly better than 1". This encoder is used for setting the telescope after a slew, and for generating the right ascension and declination readouts. The incremental encoder is also geared down from the final drive gear, but with one bit equal to 0.05". This encoder is used for tracking and precision offsetting of the telescope.
Precise movement of the telescope is accomplished by the rate generator, which converts a velocity demand (from either the CCS or manual switches) into an analogue signal which varies with the magnitude of that demand. The analogue signal in turn is sent to the drive system to move the telescope. Feedback from telescope motion is from the incremental encoders back into the rate generator. The maximum output from the rate generator will drive the telescope at 200"/s.
The CCS controls the telescope in two modes; a digital tracking and offset mode and an analogue slew mode. In digital mode, the CCS loads the rate generator, via the DIGMUX, with a velocity demand. For slewing, the CCS sends the hour angle and declination servos an analogue demand, monitors the current telescope position during the slew and adjusts the demand according to the distance remaining from the desired coordinate. The maximum slewing rate of the telescope is 45°/minute, while the maximum rate for dome rotation is 72°/minute. In practice, slewing involves both the analogue and digital systems, with the analogue slew bringing the telescope somewhere close (within about a minute of arc) to the desired position. The telescope then settles for a moment, after which the CCS begins digital tracking and issues a digital offset to bring the telescope to the exact desired coordinate.
The timing system, telescope drives, and computer systems, together with the instruments, derive their AC power from uninterruptable power supplies which ensure that observing can continue even if the mains power fails.
The telescope control system is operated by a `cheerful and well informed' Night Assistant whose responsibility is to make the telescope perform as the observer wishes, so that astronomers need only concern themselves with collecting data. A brief outline of the control software is given here, but AAO TM 7 Telescope Control System Operations Guide contains a full description of the control system for those who want to know more.
Fully-corrected pointing and tracking are an integral part of the AAT control system. The telescope pointing is corrected by attempting to model `real' effects, such as encoder zero point errors, azimuth and elevation errors of the polar axis, mechanical misalignments and flexures, rather than using an empirical, whole-sky polynomial. Atmospheric refraction is also taken into account.
The rms pointing errors are different for each top end, but are consistent for each configuration. Typical errors are about 2" at prime focus and Cassegrain f/8, while f/15 is a little better at around 1.7" rms. The pointing at coudé is poorer, with errors around 4-5". The night assistant makes a quick check of the pointing at the start of each night.
Tracking corrections are derived directly from the pointing model in a way which completely avoids cumulative errors. Even after hours of tracking, errors in position are due only to the discrepancy between the pointing model and physical reality. For observations lasting less than 5-10 minutes, guiding corrections are small (no more than 0.2-0.5"). For longer exposures, the tracking errors should be no worse that 2" to 3" per hour. The tracking is generally smooth, though tests with the autoguider have shown some low frequency irregularities of up to 0.5" amplitude.
Differential rates can be applied to the normal corrected telescope tracking by entering data through the display sequence. The rates are 0 to ±999.9s/hour in hour angle and 0 to ±9999.9"/hour in declination.
These differential rates are not required for correcting normal sidereal tracking. They are used only on the rare occasions when a controlled drift in the telescope apparent place is needed, for example when tracking a comet or planet.
The control system accepts coordinates as either mean or apparent. Mean place is a heliocentric position and refers to the mean (precessing) equinox and equator of a nominated epoch, usually a standard one such as 1950.0. Apparent place is a geocentric position referred to the true equinox and equator of the epoch of observation. In almost all cases, mean place is the correct form. Apparent place is only used for objects within the solar system whose ephemerides are published in apparent coordinates.
All AAT coordinates are pre-1976 IAU recommendations, i.e. they are in the FK4 system, not FK5.
Because the AAT points so well, it is important that observers have accurate coordinates for their target objects. With positions accurate to 1", correct identification is straightforward in all but the most crowded fields and observing can be underway within seconds of the slew finishing.
By far the best method of acquiring faint sources with the AAT is by offsetting from a bright reference source. Offsets use the incremental encoders, whose resolution is 0.05", and give results of great precision and repeatability. If the reference star's position is accurately known, the final precision is limited mainly by the gears that drive the incremental encoders; and is usually much better than 1", even after offsets of a degree or more.
It is important, however, for the differential offsets to be specified in geocentric apparent coordinates. The AAT control system automatically compensates for differential refraction and flexure, but not for differential precession, aberration and nutation.
The telescope can perform some preset scanning and trailing patterns, described in detail in chapter 4 of AAO TM 7. Another powerful facility is the ASPECT scanning mode described in § 5.8, and in AAO UM 15 ASPECT: Area spectroscopy using IPCS with AAT scanning.
All scans are done by changing position rather than rate, using streams of offsets in the apparent place coordinate system. As a result, differential pointing effects such as refraction and flexure are fully corrected throughout the scan and a large raster scan, for example, will describe the same pattern in the apparent place coordinate system at whatever hour angle it is performed. At the end of a scan, the telescope takes up exactly the same position it would have reached had the scan not been performed. Furthermore, if the starting time is known then the position of the beam can be predicted at any moment during the scan (all timing is in mean sidereal time).
There are some restrictions to scanning:
During scanning, the demands sent to the drives contain discontinuities which the telescope cannot follow. These discrepancies - which can be regarded as a feature of the real telescope rather than the inertialess virtual telescope - are ignored to simplify the timing of the pattern, allowing the instantaneous beam position to be predicted knowing only the starting time of the scan. They are easily allowed for by specifying a pattern slightly larger than the one required, then discarding the `blurred' edge region of the observation. Position errors will vary with the velocity of the scan, but should generally be less than 1" in right ascension and declination by 3 seconds after the discontinuity, and by 5 seconds the edge effects have vanished.
Five modes of scanning are always available: RASTER, TRAIL, SPIRAL, CIRCLE and DOTMAT. Others can be provided on request, and it is also possible for another computer to generate a scan pattern via the interprocessor link.
Over the years, many other utilities have been developed to enhance the performance of the telescope. These range from the various scans mentioned above to a routine to tell how close to the moon the telescope is pointing. Many of these tasks are of interest only to the night assistant, but a brief outline of some of the most commonly used is given here. AAO TM 7 gives more details of these and other utilities.
If an instrument provides a DC signal bearing some monotonic relationship to the detected radiation, that signal can be used to automatically peak a source onto the detector. This is particularly useful for infrared detectors. It takes about 30 seconds to perform the required scans.
Predicts guide probe settings for the autoguider given the field and candidate guide star coordinates. A similar routine exists at all Starlink nodes under the name AATGS.
Allows real-time searches at the telescope for suitable guide stars. Near the Galactic plane stars may be found in only a few tens of seconds, but nearer the pole searching takes much longer. If you need guide stars for your observation to be a success, it is best to come prepared.
The HST Guide Star Catalogue is now searched on-line for Guide Stars, and provides a list of suitable stars within seconds.
For many types of observations, it is useful to align the instrument along the parallactic angle. The routine PA calculates this angle for the current telescope position or any other given position.
Calculates the contribution of the Earth's velocity to an observed radial velocity for the preset coordinates and a given epoch.
Enables the telescope to be moved using the general purpose buttons and keeps account of the total distance moved. It is designed to measure the precise offset between sources, for example between a supernova and a reference star so that the supernova can be found again when it is fainter. It uses the incremental encoders, whose resolution is 0.05" , to measure the distance moved. Each offset logged is the difference between the apparent places of the telescope at the time of logging and at the commencement of the measurement.
Displays on the terminal the approximate apparent places for the Sun and Moon, their zenith distances, and their angular distances from the preset object. The positions are accurate to about 10" rms and the zenith distances ignore refraction.
TVG and TVC
These two routines allow the telescope to be guided from a bright(ish) star visible in the TV field, either reflected from a slit jaw or visible through a dichroic. TVC assumes the field is rotating, aligned to the parallactic angle, and compensates accordingly. Only UCLES has a suitable beam rotator at present.
If there are more than a few objects to be observed, much time can be saved by creating a source file (data catalogue) in advance. For CCD imaging surveys where each object is observed for a few minutes or less, it is essential to create such a file before the run.
The data catalogue consists of a series of named records and each record typically contains the position of one object, though it may have other kinds of data such as the parameters for a raster scan. Once the source positions are loaded into the control computer, it is possible to move very quickly from one object to the next. Even more time will be saved if records are sequenced in the order they are to be observed, as the record pointer is automatically advanced each time a record is read from the file.
Chapter 5 of AAO TM 7 describes the correct format for the file. Details are also given in the Users' Guide for the f/1 CCD system, and will be incorporated in the next edition of the AAO CCD Users' Guide, AAO UM 17. Usually the observer creates a catalogue file on the VAX and the night assistant transfers it to the control computer at the start of the run.
An alternative to this mode of operation is the use of an "offset run file". These files can be used to control both the telescope motions, and data taking. They are described in Section 6 of the IRIS User Manual (AAO UM 30 - see the Documentation section of this manual, or the Documentation area of the AAO WWW pages).
This Page Last updated: