Subject: Updates to DOP for the 3.5m


Submitted: Mon, 2 Oct 1995 14:14:12 -0500 (CDT)

Message number: 8 (previous: 7, next: 9 up: Index)

Fellow DSC users,

During our last APO run, Heidi and I worked on the DOP product
and fixed some irksome problems.  The primary change is that I eliminated
the calls that send commands to the MC (dscheader, resync, time) each time
a frame is completed and that have
been leading to halts of the TCC.  These commands were fatal to our
drift scan observations.  To replace it, the TC35 process now
monitors the MC status output continuously.  Frame headers are still
fixed as before, but now the MC values that are copied in are the latest
that the MC sends out on its status line.  This should be OK for most purposes.
Note, however, that time quantities (UT, ST, etc) are not broadcast by the
MC, so these will not be written to the header.  I have reinserted the
UTCLOSE parameter that the DA write in the header since this is an accurate
value of UT at the frame end [actually, I think it is TAI, which is 30 seconds

Another change that was made is that frames on the unix side which are being
worked on by the header fixer are now copied from the physical region to
a normal region first.  This avoids a problem whereby old frames were being
corrupted by new frames before they could be written to disk (a problem
that occurs when drift scanning fast).

So where is all this code located?  I have not touched the dop directory
on the DSC account.  There are still a couple of files that have not been
checked into the FNAL CVS code repository (and can't until some unknown
problem is resolved).  This copy still contains the code that call dscheader.
The new code is in ~skent/dop (I started with dsc/dop and modified from there).

Both sets of code are now also installed in the products area on fnstar and
have been declared to UPS.  So the politically correct way to access them
(once you log in to the dsc account) is:
   setup dop rodgers  -  old version
   setup dop devel    - new version
This defines DOP_DIR, so you can proceed with dscRun, etc., as described
in the DOP primer.

I have also simplified the startup procedure.  From the DSC home directory,
after you have setup dop and typed "dscRun dsc" (which starts the DA
program running), type "source dsc.tcl".  This will
run all the startup commands AND
will also prompt you for the name of a data directory, logfile name, and
group name.  The directory, logfile, and group are created if they do not
exist.  The names are stashed away in a file (~dsc/.dscDefaultDirs),
so if one needs to exit and restart the DA program, the old names can be
restored fast (just type <CR> at the prompts).

dsc.tcl also defines 4 new commands.  Their usage is:
   goStare <ExpTim> <ObjName>
   goDark <ExpTim>
   goDrift <nframe> <ObjName>
   darkDrift <nframe>
These will update the usrObjName status entry with the object name and
will also write the object name to the header with keyword OBJECT.
If desired, I can add them to the dop product proper.

Brian Yanny has worked on the MC code that creates the RAA and DECA positions
(these are the approximate RA and DEC of the telescope that update continuously
when the telescope is drift scanning).  Previously, positions were several
arcmin in error; now they give the correct RA and DEC at the instrument
center in FK5 coordinates.  Remember to issue to the MC the command "approxra 1"
in order to get this info broadcaset.  The most recent broadcast value
goes into the headers.

Steve Kent
APO APO APO APO APO  Apache Point Observatory 3.5m  APO APO APO
APO  This is message 8 in the apo35-dsc archive. You can find
APO  the archive in /u/strauss/apo/mailer/apo35-dsc on
APO  To join/leave the list, send mail to
APO  To post a message, mail it to