CAMAC Progress Report

                      Week Ending: July 12, 2002
                               Week: 76


Tom Folkers:

Tom Sargent:
     1. Learned that the crashing of the DSP simulator program seems 
        to be due to too many commands being sent to the DSP in too 
        short a time. Tried a FIFOing scheme to smooth out the flow of 
        commands, but it still failed occasionally. 
     2. Found & fixed a few more bugs in the sun avoidance path logic, 
        that is, special cases that weren't handled correctly. 
     3. We went to the SMT and left the system installed there. I 
        found a way to make servo compute a "straight line" path in 
        AZ-EL space quickly enough that it was worth doing first 
        before trying a more complex sun-avoidance path. If the 
        straight path works (doesn't get too close to the sun) it uses 
        it.
     4. Added tsAcknowledge() and validate() to servo.  These use 
        SAMBUS bits to indicate when the telescope is on source and to 
        acknowledge a command has been received.
     5. Found and fixed a bug in thew logic that made it some times 
        drive one axis direct (as though the sun was down) while the 
        other was trying to follow a sun-avoidance path.  The sun 
        avoidance logic seems to work OK, though.
     6. Discovered that the coarse encoder positions still have way 
        too much noise on them. They can flop around by as much as a 
        few degrees on consecutive reads. We need to filter these 
        more.
     7. I got smtxdpy2 set up in my office and was able to control the 
        telescope with it remotely.
     8. I wrote, but didn't test the 'hatches' command.

Pete Hubbard:

David Forbes: