Observing Operations | Reviews | Survey Management

Sloan Digital Sky Survey
Review of Observing Systems and Survey Operations

Overview of the Motion Control Processor
and Telescope Control Computer
Robert Lupton
April 10, 2000

Functionality

The Telescope Control Computer (TCC) is a DEC Alpha running VMS. The TCC programs were written by Russell Owen (University of Washington) and are essentially identical to those used to control the APO 3.5m telescope.

The TCC performs the following functions:

  • Converts positions specified in e.g. (ra, dec) into (az, alt, rotation)
  • Controls the 2.5m's motion by means of sets of PVT (Position, Velocity, Time) triplets, specifying the desired position and velocity of the telescope at the specified time. These PVTs are sent to the MCP.
  • Controls the 2.5m's optics, allowing for deflections and responding to requests for piston motion of the secondary
The Motion Control Processor (MCP) is a MV162 processor running vxWorks sharing a VME backplane with a number of other boards:
  • An MEI board to control the telescope's axes (az, alt, rotator) and read the fiducials. This MEI board is responsible for the actual PID control of the axes
  • A number of Industry Pack boards to interface with the programmable logic processor that controls e.g. the flat field lamps and the brakes
  • A second MV162 (The Telescope Monitor Processor, TPM) responsible for logging the telescope's status
  • More Industry Pack boards to read e.g. the position of the mirror activators
The MCP is responsible for:
  • Translating motion commands from the TCC's PVTs to commands that the MEI can understand
  • Accepting and implementing requests from the TCC or engineering menu to e.g. set the brakes, or reset the amplifiers
  • Accepting and acting upon commands from the user-level programs (IOP and SOP) to e.g. turn on flat field lamps
  • Interpreting the true positions of the axes, as revealed by crossing fiducial marks, and making appropriate allowances for errors
  • Reading the state of the interlocks system; the axis positions voltages, currents, and other state; and the state of other components such as the flat field screen. This information is made available to the TPM via shared memory in the VME crate, and to other interested parties via UDP packets.
  • Maintaining a log of what it's doing
Status

Of these systems,

  • The TCC is fully functional
  • The TPM logging is functional, although there are still some minor issues to be addressed
  • The MCP is reasonably stable, although there is some missing functionality and some outstanding problems. For example,
    • It is intended that the MCP be able to control instrument changes but this has not yet been implemented.
    • On-the-fly corrections to the encoder positions are implemented, but disabled pending further testing
    • Offsets of -30' in alt take the telescope out of closed loop
    • There are some mysterious failures and hangs
When the MCP works, as it almost always does, its performance is very good. It is not clear whether the intermittent axis-motion problems will be solvable without a total rewrite of the code that handles PVT commands from the TCC.

I believe that all of these systems are maintainable, although the long-term owner of the MCP code has not been identified.  Documentation for the TPM and (especially) the MCP is seriously lacking.



Review of Observing Systems and Survey Operations
Apache Point Observatory
April 25-27, 2000

 
Submit comments and questions to the SDSS Helpdesk. Last updated 04/01/14