Getting Started With Observing

During the day you should take some calibration data (darks and flats) and in the process of doing this you can verify that the instrument is working properly. We would suggest the following steps:

• Take 20 short dark exposures with a single low noise read. Measure the RMS to verify the system noise. The noise should be less than or equal to about 4 to 5 ADU.

• Take a long dark (200 seconds) and verify the system dark current. The dark current should be about 0.5 electron/second.

• Move the aperture to the .35mm slit. Move the grating to zeroth order (sidea or sideb and set the filter wheel to K2. Take a several second exposure and note the position and tilt of the image of the slit. You can adjust the position of the slit to pixel 128 in the dispersion direction using gjog (note: 1 jog step=2 pixels approx.). You can adjust the tilt of the slit using the ajog command.

• Configure the instrument as you will want it for observing (ap, filter, grating and wavelength). If possible verify wavelength setting using calibration lamps.

• Obtain flat fields using the dome lamps (at $\lambda < 2.5\mu$m) for each configuration of the instrument you will observe in. At longer wavelengths you can just use the thermal emission of the dome.

• Obtain darks or dome-lamp-off images with equal exposures times and number of coadds to those you used for flat fields. You will subtract these from the dome images to construct the flat-fields images.

Finally, during the day we would strongly suggest that you write or learn how to use the TCL scripts you will want to use during the night. During twilight you can finish the preparation of the instrument for observing. We would suggest the following steps:

• Move the telescope to a bright star. Dial in the approximate focus given in the table in Section 1.1. Place instrument in imaging mode by moving the grating to zeroth order and placing the aperture in the ``open_hole'' position use either the H or K+ filter. Do not ues the mirror position for focussing: the images from the mirror are good enough for examining the field and for acquisition, but not good enough for precise focussing. Place the bright star near the center of the open hole and mark this position on the CCD TV camera screen. Zero point the telescope.

• Drive the bright star image out of focus. Verify that the pupil looks round. If it is chopped off on one side, the pupil of the instrument may not be aligned with the telescope properly. You should talk to a staff astronomer if you suspect this may be the case.

• Move to a fainter star. A 8-9^th magnitude IR standard will work well. Focus by imaging in either the J,H or K filter. It is helpful to use a movie to do this. You can adjust the range of the image display by using the zs command to set the window limits.

• Configure the instrument for observing. You may want to observe the photometric standard to help with flux calibration.

Finally, to obtain data for an object you may follow the following procedure:

• Move to an F8 to G7 bright star near your object on the sky

• Center the bright star in the slit using the movie command by finding the half power points and placing the star between them, or by using the peak macro.

• Mark the position of the bright star on the CCD TV camera with the electronic cross hair. Check the signal level of star when it is at its centered position in the slit to verify the peak counts are less than 5000ADU.

• Observe the bright star by moving it along the slit using the command slitscan or a custom script. slitscan is capable of scanning along the slit for any orientation of the rotator (initialized by the command zrecenter.) These spectra will be used to remove telluric absorption features from your object spectra. Sample scripts are in /ua10/irs/scripts. (See section 2.7 for a discussion of scripts.)

• Move to your object and center it in the slit (see section on ``Object Centering'').

• Select an integration time that will not saturate your object or the sky background. If your object is faint it is best to limit your integration time so that sky variability will not cause problems. For the H and K bands this is between 300 and 500 seconds. For the L and M bands it is about 60 seconds.

• For an object bright enough to see in a single integration, observe it by scanning it along the slit using the appropriate script, unless it is a galaxy that fills the slit.

• If your object is too faint to see in a single integration it is best to use a number of ABBA beam switching cycles. Depending on the nature of the object, offsets may be perpendicular to or parallel to the slit direction. The commands abbaperp.tcl and abbapar.tcl accomplish this, again for any setting of the rotator. Once enough data has been taken to detect the object, it is best to dither the object a few pixels along the slit before taking more ABBA cycles to help with bad pixels rejection.

• Return to the bright star to monitor telluric absorption and to verify source centering every 1 to 1.5 hours.