Observing Notes

[last modified July 26, 2007. rdb]

• If you are a new user, please see the Known Problems Page.

• The telescope focus is best made by tunning the active optics system on a bright star with the Calibration Wavefront sensor (CWFS).  Once focused, the following offsets may be used where the offset is added to the CWFS position in the TCS:
• f/7 K: +60
• f/7 H: +60
• f/7 J: +60
• f/3 K: +50
• f/3 H:+50
• never believe these numbers
  
• These offsets assume the internal camfocus has been adjusted for f/3 and f/7.
• At f/7, do a focus run of +/-50 about the nominal value (e.g., 10-110).
  
• The focus changes by about -100 units/C.

• With the Rotator at 0 degrees, OSIRIS is up looking. Facing the telescope, CW rotations are "+" and CCW "-".
  
• In order to avoid spilling LN2, try to keep the Naysmith Rotator angle between +80 and -90 degrees.
• There is a Rotator Tool which will show the rotator angle vs hour angle for a given PA on the sky.
• PA 0 is North up and E left on an IRAF display. PA 270 is North left and East down on the IRAF display
  
• Select the PA so that the rotator will remain within +80 and -90.
  
• Always update the current PA using the script "zcenter.pro." This will allow the other data taking scripts to work correctly.
• Tracking an object beyond these limits is possible, just more LN2 will be spilled.
  

• Use XD for low res spectroscopy unless a large slit length is required. XD provides covereage between 1.15 and 2.3 um. Because the spectrum is dispersed over only ~600 pixels, there can be order overlap in normal f/3 mode. The nature of this overlap is that the blocking filter will pass light from the shortwavelength part of the next highest order and this light will fall to the long wavelength end of the spectrum in the principal order. For example, in a normal f/3 H band spectrum which is 4th order, the wavelengths from 5th order near 1.5 um will appear in the spectrum past 1.8 um (that is at small pixel numbers, or near the bottom of the image). This is most easily seen by considering the XD spectrum. The only difference is that in f/3 mode, the spectrum is not cross dispersed and the blocking filter is centered about a single wavelength band.
• To take an "ABBA" sequence along the slit, use "slitscan" with 2 cycles. This will give an "ABAB." To take a classic ABBA where "B" is off source, use "abba".
  
• Always set the PA using "zcenter" before taking data (default PA=0 in Prospero). There is a ~2 degree offset between the TCS value of the PA and the best angle to be used in zcenter.pro inorder to keep stars in the slit while offseting along the slit. The current value is -2.2 degrees and this offset is added automatically in the zcenter script. Thus, one should always enter the TCS value of the PA in zcenter.pro. The value of the offset can be changed by editing the script zcenter.pro.
• In f/7, there is vignetting in the camera, so you will see dispersed light outside the nominal range of the spectrum. Pixels 250-900 define the approximate good range.