At the telescope: daytime

The general rule is that the astronomer does not have rights to the telescope until after 4 pm, at least on the first night. Before that time, site staff have a healthy check list to get through.

We now summarise the astronomer's check list for setting up TAURUS/TTF ahead of the first observations. The support astronomer will be there to handle the set up. The items are shown in a sensible order to be carried out.

You are ill advised to change your mind about the optical set up (top end/camera choice) or detector once at the telescope. If it is absolutely necessary, we will need to find creative ways to placate irate telescope engineers and support staff.

Etalon balancing. If this is the first night for TTF, it is important to get TTF loaded intoTAURUS and balancing as soon as possible. It can take 12 hours to settle down. Instabilities are almost entirely due to the time it takes the electronics to reach some kind of thermal equilibrium.

Calibration lamps. Check to see if the lamps you need to calibrate your filters are in the chimney. If not, we will need a site person to hunt down the lamps, and to install them in the chimney. The lamps currently available are listed here. If you are using the BTTF in the U and B bands, we must check to ensure that the chimney lamps are bright enough for the calibrations. As a last resort, we can move the chimney lamps down to the guide probe.

Camera focus. It is best to check the TAURUS camera focus, since there are always differences for the different detectors. I have learnt not to trust my previous focus for a given detector, as changes in mounting rings have been known to take place. For this, we
use the colander (matrix) mask illuminated with white light. The point sources should have FWHM of about 2.7 pix if the camera focus is right, although there will be a slow radial trend due to the TAURUS optics. This operation only takes 10 mins.

CCD flatness. If this is the first night that TAURUS has been on, it is worth checking that the CCD has been mounted parallel to the image plane. For this, we use the colander (matrix) mask once again illuminated with white light. The point source distribution should show only radial trends (see section X) with no lateral gradients.

CCD bias.  From the CCD window, use the bias command after repeat 10 in order to build up a decent set for bias subtraction. As a general rule, using the median of the bias strip is not good enough.

TAURUS wheels. Check to see if the four wheels (aperture, focal, pupil, etalon) are loaded up with the correct filters, masks and etalons. Each of these have 8 positions, except the etalon wheel which has 6 positions (although only 4 for balancing etalons). It is imperative that you ensure the right filters are loaded into the right positions.  A very useful mask to have in the aperture wheel (shared with LDSS) is the colander (holes) mask. This is used to check (a) that the CCD has been mounted orthogonal to the optical axis, (b) that the etalon is tilted correctly to position the ghosts, (c) the TAURUS camera is correctly focussed, and (d) to measure optical distortions over the field due to the TAURUS optics.

Here is an illustration of the four wheels. You can download a postscript image here for printing out and marking up.

Plate parallelism. This operation is required only once at the beginning of the night. It is found to be stable over a wide range of plate spacings once established. We need to ensure that the TTF plates are parallel to maximise the system throughput. This is achieved through charge shuffling and a sophisticated Hartmann test at the AAT. (You may have noticed the four quadrant masks in the pupil wheel of TAURUS.) At the WHT, this is a slower procedure as charge shuffling has yet to be implemented.

Wavelength calibration and bandpass selection. You need to decide (i) what spectral bandpasses you need for the night, and (ii) which blocking filters are required for the night. We will need to carry out wavelength calibrations for each of the filters for each bandpass you choose. So if you select two filters to work at two different resolutions, there are four wavelength calibrations to be done. Each of these takes about 15 mins; the new shuffle calibrations at the AAT are expected to reduce this to about 3 mins.

File translated from TEX by TTH, version 0.9.