Taking exposures / The CCD controllers
VERY IMPORTANT:There have been big changes to the CCD controllers. See this page before reading any further.
The DBS cameras are currently run by Astromed controllers connected to a pair of workstations running UNIX, located on the left of the Telescope Control System. The second one on the left of the TCS is called "mouldy". From there the Red arm of the DBS, and the Imager, is controlled. The DBS Blue is run from "moist", the third one on the left of the TCS. This is a temporary arrangement and you may need to see the staff to make sure the computer is running the correct instrument, or to learn how to swap between DBS Red and Imager. The DBS Blue data is usually stored on the workstation "mallorn" in the Peripherals room. You can log on it from the workstation "moist", located next to "mouldy". Running "Teldisk" on CICADA at the beginning of your run will automatically select the right output path for your images.
The Astromed's command structure is rather obscure, so the CCDs are controlled via a collection of macros which have been predefined and are loaded automatically on the CICADA interface. To start, type "cicada &" in any window of the appropriate workstation and select the DBS Blue on "moist" and the DBS Red on "mouldy" (if unsure, check with Staff to learn the names of the CCDs you should select). Set the data frame number using CICADA. Numbers of up to four digits may be used.
In order that the correct header information be read from telescope control computer, the exposure must be started on the red arm of the spectrograph before starting the blue exposure. IMPORTANT: the DBS-Imager setup PC does not interact with CICADA, therefore whatever information appears there (slit width, filters, lamps used) must be entered in the header MANUALLY using CICADA, or recorded in your personal observing logbook/logfile.
At the end of the night, CCDs should be left on "CCD temp + flush". Select the appropriate macro from the CICADA pulldown menu.
The controllers are initialised using the "Initialise CCD Hardware" button on the CICADA menu. This should be done at the beginning of the run and every time the bias level counts are abnormal.
DBS Hardware control
Computer control of DBS hardware such as slit width, decker and
arc lamps is from the PC labelled
"Imager-DBS"
at the very left hand end
of the main console. Normally the control software will already be
running. If the DOS prompt is showing, run dbs
from the \dbs directory. i.e.,
cd \dbs
dbs
Then select "control -> status" from pull-down menus.
The mouse driven menus should be largely self-explanatory, but some details are given below.
The decker
To move the decker, click on "Decker" in the window labelled "M2" to
open the control dialogue box. It cannot be set to specific positions, but
rather is manually driven into (using button labelled ">>") and out from
(using "<<") the slit. With the dome lights or one of the arc lamps
(the Quartz lamps are too bright) on, the decker is clearly visible in
the auto-guider camera. Use this to approximately position the decker or
to see when you have driven the decker clear of the slit without having
to take an exposure.
Setting the slit width
To change slit width, click on "Slit", located above "Decker" in the "M2"
window to open the control dialogue box. Either specify the required value
(measured in arc-seconds) in the dialogue box and click "Apply" or
drive the slit manually using the buttons labelled ">>"
and "<<".
You can check the slit visually on the
autoguider camera
using scattered light in the dome with the arc mirror out. The orientation
is reversed - e.g., North up, East to right. When slit is in position angle 90,
the left end of the slit points East.
-
Whole system: A filter wheel located before the slit carries five 48mm neutral density filters.
They are Melles Griot metal-on-glass filters
and allow approximately 50%, 10%, 1%, 0·1% and 0·01% transmission at
5000Å. Each filter, or a clear aperture may be selected using the
"Main Filter" dialogue box from the "M1" window. Plots of
transmission curves
are available for some filters.
Autoguider: Conventional B, V and R filters as well as a set of neutral density (ND) filters are available for the autoguider camera. They may be selected from a pop-up menu attached to the "TV Filter" button in the "M1" window - normally to the left of the screen.
Arc Lamps: A neutral-density-wedge filter covers the arc lamps to allow them to be dimmed. Click on "ArcFilter" in the "M1" window to open the control dialogue box. The filter is adjusted clicking the on-screen ">" button to darken it and "<" to lighten. For no filtration, drive it fully to the left until the display reads "RevLimit".
Selecting arc lamps
NeAr, CuAr, FeAr and CuHe arc lamps and two quartz flat-fields are available. Only one
lamp may be selected at a time and it is specified in the "Arcs" dialogue in the "M1"
window. A mirror folds the light from the lamps into the spectrograph and is controlled
from the "ArcMirror" dialogue, also in the "M1" window. The lamps are switched on
automatically when the arc mirror is "IN". If the arc lamp is too bright, neutral
density filters are available. See
Changing filters above.
Taking calibration data
Bias Frames
The repeatable bias structure is small; typically ±1 in B and ±3 in R. The bias level is of order 1000, so simple consideration of the Poisson statistics shows that you need >>10 bias frames to really measure it properly. Do not mistake random noise which often appears for genuine bias structure. Using too few biases which include random rubbish will make you data worse, not better. The CCD controller sometimes creates considerable noise in the form of coarse horizontal striping, which cannot be corrected by bias frames (it looks a little like a badly tuned TV signal.) This must be cleared by running "Initialise CCD Hardware" from the Cicada pull-down menus, but this also resets the bias levels rendering the old bias frames unreliable.
An example subsection of an R bias frame.
Arc lamps for wavelength calibration
NeAr, CuAr, FeAr and CuHe arc lamps are available. The CuHe lamp actually appears to be CuHeAr. The blue end of the spectrum, is certainly dominated by Argon lines. To switch arcs on, see Selecting arc lamps above.When choosing an appropriate exposure time, it is helpful to run `Autoscale and Update Display' (macro 15 on menu tablet) and then make a horizontal cross section plot through the arc image (macro 5) to ensure the stronger lines are not saturated.
Sample exposures with 2arcsec slit
| 300B | 600B | 1200B(I) | 1200B(II) | 158R* | 316R* | 600R | 1200R | |
| NeAr | 10 | 30 | 1 | 1 | ||||
| CuAr | 10 | 30 | 1 | 1 | ||||
| FeAr | 10 | 30 | 1 | 1 | ||||
| CuHe | 10 | 30 | 1 | 1 | ||||
| Quartz 1 | 3 | 1 | ||||||
| Quartz 2 | 3 | 1 |
* Note: When using the low resolution gratings, the arc lamps are likely to be too bright for even a 1 second exposure. You will need to use the arc lamp neutral density filter. Starting from the fully transparent position, (i.e., display reading `Rev Limit'), drive the filter in using the ">" button for about 4 - 6 seconds. This should be sufficient to allow a 1 second exposure without saturation. You will also have to increase the blue arm exposure to compensate. The QI2 lamp is slightly fainter than QI1, so may be better if you find your flats saturating at 1 second.
You may also find the DBS arc atlas useful.
Taking flat fields (Quartz lamps, Dome flats, Twilights)
Several types of flat field image are possible. Two internal quartz lamps are available in place of the normal comparison arcs. Some observers have reported finding dome flats more effective except in the UV. The new red detector has severe fringing that can be removed with a flat field frame taken at the same spectrograph orientation as for the star.Quartz lamps: Two separate, but similar lamps are available in the arc lamp assembly. They are selected through the DBS control computer. Without using filters they are rather bright and can saturate the red CCD in less than one second at low dispersions. The QI2 lamp is slightly fainter than QI1. The blue arm will require longer exposures.
Dome Flats: A ring of small lamps is installed on the telescope
top-end ring and may be used to illuminate the dome ceiling. The lamps are controlled
from the telescope console with the commands:
TEL$ SWITCH FLAT ON
and
TEL$ SWITCH FLAT OFF
The following suggested procedure requires you to first
start the telescope control computers and log on.
The fluorescent strip lights in the dome must be
switched off.
The incandescent lights are less important.
Move the telescope to zenith park and then slew it down slightly in altitude
to prevent any harsh specular reflection off the ceiling. For example:
TEL$ PARK
TEL$ SLEW * 75
Flats are best taken against the wind-screens rather than the actual dome roof, so
issue the command:
TEL$ CONFIGURE WINDSCREEN_CONTROL CLOSED
Open the mirror covers once the wind-screens have stopped moving.
The lamps' brightness is controlled by a black round dial marked "Flat field illumination control"
on the telescope console. It is located under the other light controls near the
top centre of the console. With the lamp brightness set at around three-quarters,
expect exposures of between 10 and 100 seconds, depending on dispersion. The
blue arm typically requires about five times the exposure of the red.
Twilight flats: The illumination pattern across the slit is normally different for flat fields than data frames taken against a dark night sky. Twilight fields can be very effective at correcting the flats before they applied to data frames. See, for example, "A User's Guide to Reducing Slit Spectra with IRAF" by Massey, Valdes & Barnes.