CAMAC Progress Report

                      Week Ending: March 29, 2002
                               Week: 61


Tom Folkers:
     1. With the help of Michael T. and Michael D., we determined that 
        'class' can read data files in both the big-endian and little-
        endian format. The significance of this is that now we can 
        record all data in native Intel byte order. No swapping of 
        byte order between processes are necessary.
     2. Preparing for trip to Mt. Graham.
     3. Three days on site for engineering tests.

Tom Sargent:
     1. I improved the performance of the servo logic considerably 
        (when following a sun avoidance path) by adding a "course 
        correction" operation on the final path interval - typically 
        about 4 or 5 seconds before termination.  This eliminated 
        overshoot and makes the slews finish very nicely. 
     2. We (TF mostly) tested the pointing model software for the 
        first time. In comparison to the pointing model on the old SMT 
        control system it was within about 5 or 10 arcsec in elevation 
        and 30 or so in azimuth.  Since the old system is typically 
        good to an arcsecond or so, these results have to be seen as 
        erroneous.  What checking we did in at the telescope didn't 
        reveal where in the computation the differences are.  We 
        should be able to pin this down in the lab next week, though. 
     3. I found at least one type of sun avoidance path that fails.  
        The path went below the horizon in one case. 
     4. There were some paths that had the elevation axis run away.  
        It fails to sense the end of a path subinterval. The elevation 
        acceleration in effect at that time just persists, so 
        elevation continues accelerating. I found the reason for this 
        and believe I have fixed the problem.
     5. Tested the beam switching menu, and found that it's not 
        working. It appears to do everything but actually move the 
        telescope.  I'll check it next week.
     6. Fixed a bug in the Krupp software logic which caused the 
        program not to resume driving the telescope after it was 
        stopped for some error condition, such as driving the 
        telescope into the limit. 
     7. I still need to finish the hide-from-sun logic.  The front end 
        part that detects a sun incursion, determines how serious it 
        is (just grazing the avoidance zone or not), logs the event, 
        etc. works, but the part that drives the telescope to a safe 
        spot isn't done yet.

Pete Hubbard:
     1. Trip to SMT for encoder testing and the source of noise. The 
        power supplies have there sense lines tied in parallel with 
        each other not twisted. The Rod +5 V supply measures 5 mv 
        difference between the output ground and sense ground. This 
        drop is a likely place to see building currents induce noise.
     2. Helped with finding noise source in the 2mm receiver setup. 
        Some loose cables and the strongest noise is 240 Hz. This 
        noise and other harmonics decrease when the refrigerator is 
        turned off, a ground loop is suspected.
     3. Changed the computer room temperatures probe signal connector 
        pin location further from building drive signals. Also 
        attached the shield from the building drive signal cable to 
        chassis ground. This reduced the noise pickup in temperature 
        readings.

Lori Warner:
     1. Studying up on the use of SCCS.
     2. Still waiting on VxWorks test station...

Dave Ashby:
     1. No Report.