Navigating the Spectrum
Vertical Positioning |
Horizontal Positioning |
Using Day Sky |
Using Th-Ar |
Previous Thorium-Argon Spectra
If you took a successful day-sky or Thorium-Argon test exposure, you can use it to judge the current
position of the spectrum on the CCD, and decide if and how much you
need to adjust it. If you haven't already made a test exposure, now would be a good time to do so
(bearing in mind that for purposes of positioning the spectrum, a
day-sky exposure is easier to find your way around in, but if you need
positioning precision of a few pixels, Th-Ar is the better bet.)
The exposure(s) you use for positioning need to be roughly in focus,
especially if you're aiming for precise positioning. A preliminary
focus by eye is will do; rigorous focusing
should not be attempted until the spectrum is at least close to the
position at which you plan to observe. So, if your spectrum looks
fuzzy, change "Dewar Focus" in hammotor_gui
by at least 10 steps (50
microns) and take a new exposure. Iterate until
You only need to see a few lines to get a
good eyeball focus, so you can save readout time by making the CCD
The 2048x2048 CCD's in use with the Hamilton capture virtually the
entire useful spectrum produced by the instrument in a single exposure
(the 4096x4096 CCD in dewar 4 does capture the entire useful
spectrum). However, there is no single setting used by all
observers. Particular program requirements may call for particular
positioning. E.g., programs that entail many short exposures may wish
to trade bandpass for shortened CCD readout times by positioning the
spectrum so that only the bandpass of greatest interest is read out,
or a program that is especially concerned with a few particular
spectral lines, may wish to adjust the positioning to ensure that none
of these features falls on a CCD defect (our chips, are in fact,
cosmetically excellent, but some low-level defects may be present).
The position of the spectrum is not set in advance by the mountain
staff, and must always be checked at the beginning of a run, even if
you are directly following another observer who used the same
detector. Even if you are confident that the setup you are inheriting
from the previous observer is suited to your observations -- e.g., you
are observing on the same program -- you should check the position.
The position may be checked using either daylight (weather permitting)
or the Thorium-Argon lamp. The former is in some respects easier, but
the latter is better if you are aiming for a very precise positioning
-- e.g., to exactly reproduce a setting from an earlier run -- because
narrow emission lines can be more accurately pinpointed than the
broader solar features.
"Dewar Height" stage
moves the spectrum up or down on the CCD. The height value does not
have any absolute physical significance (the numbers represent
stepper-motor units), but the number of pixels moved is related to it
by a constant, equal to 8.97 motor units per pixel for dewars 6 and 8
(7.18 motor steps/pixel for dewar 4). A larger height value moves the
spectrum up on the CCD. E.g., to move the spectrum up on the detector
by about 100 rows (pixels), hammotor's
height value should be
changed by about +897 units for dewar 6 or 8. (Note the constant is
not perfect, and that another iteration or two may be needed for fine
"Grating Tilt" function
moves the spectrum left or right on the CCD. As with "Dewar Height,"
grating values are related to pixels by a constant; in this case equal
to about 1.15 units per pixel for dewars 6 and 8 (0.92 units/pixel for
dewar 4). Larger grating values moves the spectrum to higher-numbered
columns. Again, a couple of iterations may be necessary to obtain a
Whether using day sky or Th-Ar, start by identifying a line or two, then,
by counting orders above and below the known feature(s), and using the
Table of Orders
, you should be able to determine the
current wavelength coverage.
For most purposes, placement of the red orders with respect to the first row
(at top in our illustration) is the primary consideration, because even
if the redmost orders in the Hamilton's useful range are placed on the
first rows of the CCD, the bluemost orders will be included.
Checking Position Using the Day Sky
The solar spectrum has clear, easily recognizable features that make
it ideal for determining the wavelength coverage with the currrent
spectrograph settings. (It's also really beautiful to look at!)
Weather permitting, the Coudé Auxiliary Telescope (CAT) is used
to provide day sky illumination, but the CAT need only be partly
opened for this purpose. Opening the CAT for day-sky illumination is
described in the CAT
. Remember that weather conditions must be within observing limits
to open the CAT, even for the purpose of making a day-sky exposure.
On a sunny day, an exposure of 3-5 seconds, with a typical observing
aperture (e.g., 640-microns) will produce a good exposure. In cloudy
weather or late in the day, longer exposures may be necessary. Your
day-sky exposure should look similar to the one illustrated above.
Follow the directions for vertical and horizontal positioning, above,
to make adjustments.
Checking Position Using the Thorium-Argon Lamp
The performance and output of calibration lamps varies. Over the life-cycle of
a particular lamp, intensities (e.g. of spectral lines) may decrease or
increase with time until one, or more, of the constituents of the lamp
ultimately decays. Decay can occur over a noticeable time interval or may be
In the event of the decay of an existing calibration lamp, attempts shall be
made to install a like-for-like replacement from stock spares, sufficient
to generate spectra which can provide overall calibrations consistent
with the preceding lamp. However, design, suppliers and costs of replacement
lamps evolve. There is no guarantee that replacement calibration lamps shall be
identical to their predecessors. Replacement lamps may differ from their
predecessors in important ways (e.g. manufacture, geometry, relative
percentages of chemical composition), all of which affect the characteristics
of resultant spectra. Therefore, during the life-cycle of a particular
calibration lamp and following the replacement of a calibration lamp,
intensities (e.g. of spectral lines) may differ markedly.
Designers of automated reduction procedures are advised to generalize their
routines to accommodate changes in intensitie (e.g. spectral lines;
peak wavelength), in preference to fixing analyses to one contemporaneous
set of reference spectral features.
The Th-Ar spectrum is very rich, and can present the novice Hamilton
user with a bewildering array of bright dots. Bon courage!
You'll soon find patterns and lines that will become familiar
signposts. Click on the image at left for help to get started finding
your way around the Th-Ar spectrum. Wavelengths of individual lines
can be found in Lick Observatory Technical Report #73, The Hamilton
Spectrograph TH-AR Atlas by M. D. Shetrone (it's a green,
paperbound book, copies of which should be floating around in the 3-m
and CAT control rooms).
The UCB astronomy department has also published an on-line, Hamilton-specific,
NOAO has a non-Hamilton-Specific
Follow the directions for vertical and horizontal positioning, above, to
Hamilton's orders are neither straight nor parallel to one another, and will
always show some rotation with respect to the rows of the CCD.
As can be seen in the day sky image
the CCD is set so that the the reddest spectral orders are nearly parallel to
rows on the array. Occasionally the CCD's rotation may get out of adjustment.
If the spectrum looks overly rotated to you, consult a technician or
Do not attempt to adjust the rotation without staff assistance.
Previous Thorium-Argon Spectra
During the last quarter or 2011, following decay of a Thorium-Argon
calibration lamp, a replacement lamp of alternate design, manufacture
and composition was installed.
As of the first quarter 2012 (beginning 1 February 2012) the dewar 4
CCD is the only detector available for use with the Hamilton
Thus, the accompanying illustration of a pre-2011-10-12 Thorium-Argon
spectrum is included for historic interest only.
Last modified: Wed Feb 22 16:00:00 PDT 2012