MDM Trouble Report for 2013 Nov 08

Telescope: 2.4m
Observer(s): Frank (OSU)
Instrument: OSMOS

Problem(s) Encountered:
The fast targeting procedure using  osctrtask  as described in the
gave me some trouble. 
Here is a detailed description of the steps involved in one example-- and the 
odd outcome that the offset either put the target only halfway
or twice the distance to the slit, when entering the suggested dx 
dy in xmis :

Towards the end of the night, TON34 became a nicely bright 
target that I hoped to be able to catch despite the very poor seeing.
Hence :
   1. I moved to its field (RA 10:19:56  DEC +27:44:02 ),
      and decided on a fine guide star via the help of 
      JSky... - sitting at 10:20:29.73 ; +27:54:07  at 11.8 mag
      outside the forbidden zone of OSMOS.

      The guide probe moved to : 3074, 4263, 
      and this corresponded in MaxIM to a good central location
      of x=193 , y=224.

   2. So, tracking started in a very stable manner.
   3. I took a 60s direct imagine for acquisition, and another
      60s exposure with the 1.2'' CENTER slit.

   4. Invoking epar osctrtask as described in the manual,
      said code went through its regular calculations after being
      given the correct input images.
      The resulting estimates for the slit position (y=2019), and the 
      location of the target (x = 1995, y=2147)  look very sensible.

   5.  Then I follow osctrtask's advice and apply
       a DX=-450  and DY=+212    in the appropriate
       panel at the xmis2 window.

       Having first had the idea that the trouble arose by accidentally
        double hitting carriage return  (as the warning also 
       clearly states in the manual), I have made absolutely
       sure that each correction was only applied once,
       by verifying the change seen in the X and Y.

   6. Using the handpaddle, I then moved the telescope such that 
      the displaced guide-star again ended up in the afore-noted
      box, after which autoguide was resumed.

However -- in this case (as one can easily see in the files named 
frank_08_11_2013.0052.fits  ,...0053.fits (input for the code) and
the final shifted version seen as a direct image)) the
target is not shifted all the way into the slit -- a short calculation 
reveals that it only gets about halfway there (y_end = 2087 instead
of 2019, coming from y_initial=2148).

In two or three other cases, the correction lead to an overshoot
by a factor two.

I have been unable to find a good explanation for this  (initially
I assumed that I had made one of the more obvious mistakes
like double-hitting when entering dx and dy , or having the slit
erroneously set to the position  1.2'' inner , but the - somewhat tedious -
description above shows that I have tried hard to eliminate most of 
these. 

Does the readout mode matter ? I am using the full 4x4, hard to 
see why that choice should be of consequence.

It's such a great help, that not being able to rely upon it would
be a serious blow.



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

Submitted on 2013 Nov 9 [7:22:33]