CAMAC Progress Report

                      Week Ending: June 07, 2002
                               Week: 71


Tom Folkers:

Tom Sargent:
     1. Worked on sun avoidcance with one axis not controlled by the 
        computer. Seems to work, but could use more testing.  There 
        are a couple of "pinch points"  when one axis is disabled that 
        leaves the telescope stuck, that is, the computer can't move 
        the telescope away from the sun without first making things 
        worse by moving much closer to the sun for a short time.  We 
        may want to make this more logic elaborate in the future, but 
        for the moment it is servicable.
     2. I Made tracker refuse to accept commands to go to a new object 
        if it is too close to the sun. Previously in this case, servo 
        would just pause where it was and wait for the operator to 
        select a new object and then make servo resume tracking.  
        Tracker now sends an error message to monitor saying that it 
        ignored the command and continues tracking the previous 
        object.
     3. I altered servo to allow the operator to command it to acquire 
        an object that has not yet risen.  It drives the azimuth to 
        track the object's azimuth and just limits the elevation to no 
        less than 1.0 deg. When the object gets up to 1.0 deg 
        Elevation, servo begins tracking normally.  So, it just always 
        takes as the elevation the maximum of the computed elevation 
        or 1.0 degrees.

Pete Hubbard:

David Forbes:
     1. I finally have a set of working computers that run the 
        software I need. I have not had too much experience being a 
        Linux system administrator, so this has gone more slowly than 
        I would have liked. 
     2. The Linux workstation is named smtxdpy1, and works reasonably 
        well. It has the ability to export files to the target system 
        smtob and the Windows computer putz.
     3. The Windows PC is named putz. It has Xilinx ISE WebPACK 4.2 
        installed and running, and can get to files on smtxdpy1 just 
        fine. It is a 133 MHz machine which is a bit slow by modern 
        standards, but it can compile the Xilinx FPGA firmware in 
        about 15 minutes which is acceptable for now.
     4. The target system, the Broadax touchscreen computer, is 
        currently named smtob. This name needs to change to smtaos. It 
        is running Linux 2.4.2 as a RedHat 7.1 install on its little 
        hard disk. I will probably need to install the RTLinux patch 
        to it soon. Help will be appreciated. 
     5. I have been able to compile Dave's ECP mode parallel device 
        driver5/aos.c. It fails by reading FFs instead of proper ECP 
        register data. It does  talk to the parport device well enough 
        to negotiate the port mode, but crashes the kernel when asked 
        to transfer data. I have not tried to debug this yet, and may 
        not need to.
     6. I cannot compile the older EPP Linux driver code, since it 
        expects RTLinux. I need to install RTLinux on the touchscreen 
        computer to make this work, or rearrange the makefile and/or 
        code to not require it.
     7. Dave's latest ECP firmware is not fully baked. The main 
        control logic file control.vhd does not compile, and is 
        clearly in a state of major flux. It looks like the source 
        files to reconstruct the ECP code that's currently programmed 
        into the FPGA are not available.  The EPP version from 
        14Dec2001 compiles OK. It may not work with the current 
        hardware, however, since he added a bunch of wires to convert 
        to ECP. I am going through the old and new backplane 
        schematics to document the changes. So far it looks like there 
        may only be a couple wires that moved.