Previous: Control System
Up: Control System
Next: The Detector Memory System
Previous Page: Control System
Next Page: The Detector Memory System
The main elements of the control system are shown in Fig. 18. ISIS mechanisms are controlled by the observer typing commands at the ICL instrument control interface. The ICL interface runs on the system computer (currently a Vax 4000 called LPVF), and the display is via the Observer's Vaxstation 3100 (LPVS2). ICL commands are typically procedures containing lower-level commands to d-tasks (device tasks) which are high-level (PASCAL) programs running on the system computer (LPVF) as sub--processes of the main observer's process. There are d-tasks for each of the major mechanisms e.g. ISIS, A&G, UES, telescope, DMS, each CCD. The d-tasks in turn exchange messages over the utility network Ethernet with the Motorola microprocessors controlling individual stepper-drive motors, barcode readers and shutter modules. The messages are typically commands to move, status requests and status returns.
Most of the mechanisms are controlled via Motorola 6809 (locally known as `4MS') microprocessors; there is one for each of ISIS, A&G, FOS, LDSS, TAURUS, UES and IPCS. The TV, autoguider and Detector Memory System (DMS) each use a 68020 processor in a VME chassis (housed in the blue control cabinets). The CCD controllers (including that for the autoguider CCD) are based on 68008 microprocessors. The microprocessors are programmed in FORTH. Each of the microprocessors, and the system computer, is connected to the utility network via a Network Interface Unit (NIU). Each NIU can connect up to four microprocessors. There are currently two NIUs on the telescope and two in the control room. Data from the CCDs is transmitted via fibre-optic links to the autoguider and DMS VME systems. Data arriving at the DMS are automatically sent to the data disk on the system computer and may also be transferred to the SPARCstation (currently LPSS3). The integrating TV system is currently stand--alone and cannot be controlled across the Utility Network. If the ICL/d-task control route fails for some reason, mechanisms can often be controlled more directly from the engineering terminals in the control room, by keying in messages to be sent over the utility network. At a still-lower level, duty tecnicians can plug a terminal directly into a microprocessor on the telescope and send commands from there.
LPVF is linked via DECNET, running over a separate Ethernet, to the telescope-control computer (a microVAX-II, currently LPVB), to the VAXstation running the user interface, to other VAX 4000 computers on a DEC cluster, to the DECservers serving terminals and printers, and to SPARCstations on a SPARC cluster. Separate Ethernets at the INT and JKT are connected to the WHT ethernet via fibre-optic links.
The DMS and several of the VAXes and SPARCstations can be used for data reduction. The DMS provides a few simple analysis functions e.g. cuts, statistics. FIGARO and other reduction packages are available on the VAX clusters on both the mountaintop and at sea level. IRAF and SAOIMAGE can be run on the SPARCstation on the control desk (Section 7.3).
Telescope control is effected via commands issued to the control system running on the telescope-control computer, which is interfaced to the telescope mechanisms via CAMAC. Most telescope-control commands can be issued from the ICL interface and it is intended that in the future the distinction between operator control of the telescope and observer control of the instrument will blur.
Most of the information needed by the observer about the status of the system is included in the telescope-control information display, the instrument-control MIMIC and the white-board. The MIMIC is discussed in Section 7.6 below; it displays information contained in the `notice-boards' which are continually updated by the d-tasks as a result of the commands and status returns they receive. Ephemeral information which is not (yet) held on-line e.g. contents of filter slides, gratings loaded, CCD locations, is usually recorded on the whiteboard in the control room. There is some overlap between the information held on-line and that written on the whiteboard. Some useful instrumental information (e.g. focus offsets) is recorded in the telescope logbook.
[ TIFF ]