MDM Observatory 2.4m Telescope Observer's Report for 2002 Jul 09

Observer(s): Rick Pogge
Institution(s): OSU

Instrument: MDM8K

No Observing (see comments below)
Conditions were Overcast most of the night.
No equipment or software problems were encountered.

Seeing and Weather:
   Clouds all night.

Observing Summary:

Monsoons began today with the first rainfall we've seen up
here in a while.    Only a few holes overhead, and no lightning
danger right now, though we were shutdown for a few hours
this afternoon.

I got a basic TCS link incorporated into the MDM8K data-taking
system today.  Detcom has spotty internal documentation, so a
lot of guess-work was required.  Tests with the scope parked
show we can now get full TCS pointing info (RA, Dec, LST, HA,
airmass), telescope info at the 2.4m (focus and rotator angle),
and times (DATE-OBS, TIME-OBS, and JD) into the MDM8K FITS
headers.  Will work on cleaning things up, documenting it,
then migrating it to etna for use at the 1.3m.  At present
it will not be possible to command either focus or offset 
motions from the detcom console - it is harder to break into
the command block than I thought, so I implemented the tcs
stuff by hijacking the CFHT tcs hooks.  In either case, it
will require the MDM8K users at both telescopes to run a
*different* version of xtcs (xtcs4 instead of xtcs3) until
I can get around to changing both ccdcom and prospero to
talk to the new xtcs interface.  Will write up the procedures
for the observers as part of updating the online and
mountaintop handbooks later this week.

To see a sample TCS header block (extracted from one of the
test images), go to

 http://chichon.kpno.noao.edu/8ktcshead.txt

Please recall that the TCS only returns 5 numbers: RA, Dec, 
Equinox, Focus, and Rotangle, so all other data must be
computed based on the time read off the system clock.  The
system clock is synched to network time services.  Anything
else that can be calculated after the fact should, reducing
the burden on the data-taking system. 

Latency in the system is about 1-2 seconds, but I have tricked
things by reading the clock a second time *after* the TCS read
returns info.  Since the next step in the exposure execution 
flowchart is opening the shutter, these are the
UT time (and corresponding JD) that get written into the 
header.  If timing matters at the couple of second level for the 
science, the relevant header data are TIME-OBS or JD.  Note
that the times previously reported in MDM8K headers was
HSTTIME and DATE, which are when the header was created, not
when the exposure was started.  The new interface eliminates
this source of confusion.

No activity planned for tomorrow as we have business off-site
on Mt. Graham.


------------------------------

Submitted on 2002 Jul 9 [21:21:32]