MDM Trouble Report for 2008 Nov 2

Telescope: 1.3m
Observer(s): Halpern and Schroeder (Columbia)
Instrument: CCDS

Problem(s) Encountered:
In the afternoon, it was noticed that McGraw had crashed. 
After rebooting the telescope had to be rezeroed. 
At the beginning of the evening, it was noticed that the dome 
was not centering on the telescope position. 
Rather, the dome believed that whatever position it was 
currently positioned was center. Executing dome put allowed us 
to move it to different positions, but it was clear that the 
dome was not tracking the telescope. Executing home dome also 
worked properly. Rebooting TCS did the trick, but it later in 
the night we noticed the dome was orientated +2.5 degrees 
beyond where it reported itself to be in DomeAz.

The ACS system continues to crash. There are two basic kinds of 
crashes. The rarer of the two crashes is the one where the 
connection to the camera is lost due to a loss of signal. 
This does not usually cause the TIKFAM/CCDS computer to freeze 
but rather the program rightly insists that it cannot find the 
camera. The ethernet to USB connection green light in these 
cases has gone dark while the yellow light remains lit. 
Unplugging and replugging the ethernet cable turns the green 
light back on, but power-cycling the camera is still required.

The other type of crash is more serious because it involves the
CCDSOPS program freezing. Since there is no option on Microsoft 
Windows to kill -9 a job (going into Task Manager and trying 
to end the program does not work in this case), the only 
recourse is to restart the computer from the start menu. 
Unfortunately, this takes a ridiculous amount of time because 
the computer spends five minutes waiting for an end program 
response from CCDSOPS before finally deciding to override it 
and shut itself down. Upon restart it is necessary to power-
cycle the camera in order for the newly restarted CCDOPS to 
recognize the camera. This is quite a time-consuming problem, 
although it only happened twice tonight.

Some of the functionality in the TCS is a bit frustrating. 
For example, when using the GUI's virtual paddle, you cannot 
hit buttons in rapid succession and you can only slew up to 
60 arcseconds at a time making certain tasks a bit more time 
consuming than they would be if the old  real  paddle still 
had set function capability. Also, the system does not handle 
multiple commands well. For example you cannot home the dome 
and then immediately close the mirror covers. Instead you have 
to wait for the dome to reach home and then execute the mirror 
open command. It would be nice if the zero command accepted 
different equinoxes.

Tomorrow we will try to take flats against the screen in the 
dome since the MIS flat lamp does not give enough counts for 
our wavelength range of interest.

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

Submitted on 2008 Nov 3 [6:50:09]