Troubleshooting guide

projector have absorption bands near 2.3 microns which will show up in the flatfield spectra. Therefore, an
alternative approach to obtaining flats is to observe the inside of the dome using the flatfield lights mounted at
the end of the telescope.
Usually only one set of wavelength calibrations is needed per target; alternatively the OH sky lines may be used
for wavelength calibration, if they are not saturated.
Flat fields should be taken contemporaneously with the target observations: there may be flexure between the
slit and the detector which is dependent on telescope hour angle (HA) and zenith distance (ZD). Additionally, if
one tries to take them at a later time it may not be possible to precisely insert the slit and grism in exactly the
same way as they were when set up on the target. More than one flat field data set may be required if the
observation spans a large enough range of hour angles (
1 hour, possibly).
We recommend the following:
If the target is a bright standard star, then the observation time probably is short and a single set of
calibrations, taken after the target, should suffice.
If the target is faint, and the time on source is long, take more than one set of flats at several different
hour angles that lie within the same range of hour angles as covered by the science target, but take
them all after finishing the target's observations. For example, if the science target is taken over the
hour angle range from -2:30 and ending at +1:00, then take a set of flats at the final hour angle, then
move the telescope back to an hour angle similar to the initial hour angle, and take another set of flats
(the telescope does not need to be tracking). Or, if you only have time for one set of flats, consider
moving the telescope to the midpoint of the track in hour angle.
The following procedure is used for taking quartz flats with the rotator/guider calibration projector after taking
science target spectra. The alignment of the slit, filter and grism should not have been changed from that used
for the science target. It should take 5 – 10 minutes to complete.
1. Have the telescope operator do the following:
Pause guiding. To do so, just hit the OFF button for the guider.
Turn down the guide camera gain.
Select the Comps mirror position.
Select the Dim Quartz lamp at 100% power.
2. Set filename and exposure time. Use config.exposure.pl.
Integration times of 10 - 30 seconds should suffice for JH-band and HK-band spectroscopy.
Take a test exposure (singleimage.pl). The counts should be <25,000 ADU for wavelengths
< 2.4 microns.
With 30 second itimes, it is possible to do 8 reads, if you can accept the little bit of extra overhead.
3. Take 5 flats with the lamps on. Use more.single.images.pl 5.
4. Have the telescope operator turn down the quartz lamp, or turn it off.
Leave the comp mirror in place.
With flats at two different illumination levels, they can be differenced to remove the dark current.
5. Take 5 flats with the lamps off. Use more.single.images.pl 5.
6. Have the telescope operator turn on the HeNeAr lamp.
7. Take 5 wavelength calibration images. Use more.single.images.pl 5.
Integration times of ~60 seconds should work for JH-band spectroscopy.
FLAMINGOS@4-m, Ver. 2.34, 2006 Apr 04 Page 29 of 44