User's Guide to the Hamilton Spectrograph

Table of Contents

Quick Reference
CCD Characteristics
Grating Tilt & Dewar Height
Filter Wheel & Shutter
Calibration Sources
Photon Integrator
Image Rotator
Iodine Cell & Slit Room Controller
Guide Camera & Filter Wheel
Data Taking System
Hamilton Motor Controller
Hamilton Focus
More Info:
Spectral Format
Navigating the Spectrum
Table of Orders
Setup Procedures
Observing Hints

Data Archive
Mt. Hamilton Homepage


Hamilton calibrations are of four types: flat-fields, wavelength calibrations, narrow quartzes, and darks. Though the first three of these are part of most Hammilton programs, there is no standard set of calibrations used by all observers, and individual users will have to decide which types, and how many of each, best meet their program's requirements, approaches to data reduction, and personal preference. Some guidelines are given below, along with instructions on how to make each type of calibration.

Flat Fields

The flat-field source is a continuum quartz lamp mounted above the entrance slit. It is turned on by selecting "PolarQuartz" from hammotor_guis's "Calibration Lamp" section. Flats are made using an aperture at least twice as tall as that used for observing (typically 5 or 6 arcseconds). This produces flat-field orders wider than the stellar orders, preventing edge effects when performing the flat-field division (you may hear Hamilton flats referred to as "wide flats").

Because of the Hamilton's very large wavelength coverage, combined with the lamp's energy distribution and the detector's response curve, full illumination of all the orders in a single exposure is impossible, except in those unusual cases where wavelegth coverage has been sharply reduced by windowing the CCD. Filters are used to correct the problem of even illumination as much as possible. The standard formula for Hamilton flats covering wide wavelength range combines bg12 and bg13 filters. The exposures through each filter may be taken on separate frames and later combined by adding pairs of frames. Alternatively, the filters may be combined in a single exposure by pausing it part way, changing filters, and completing the exposure. In either case, the ratio of exposure times for bg12 to bg13 is about 10 to 1. Absolute exposure times vary with the age of the lamp, the size of the aperture, and the CCD in use. Keep peak exposure levels within the device's recommended linearity limit.

Hamilton flat fields

Even with the use of filters, the number of counts in the blue is lower than in the red. In order to get sufficient signal at the blue end it is necessary to take multiple flats, which can be coadded later to produce a high S/N master flat. How many flats is enough will depend on program and reduction requirements. Most observers take at least five, some take substantially more. Flats should be taken for each night of observing; some observers choose to take a series of them both before and after the night's work. Taking flats by hand can be tedious and time consuming, but can be automated using the IDL procedure autoflat, described below. (N.B. autoflat can only take images through one filter at a time for later addition; flats combining a bg12 and bg13 exposure in a single image must still be made by hand, by manually pausing part way through the exposure to change filters.)

Using autoflat or autoflat_alternate to take flat fields:

The IDL procedures autoflat and autoflat_alternate take bg12 and bg13 flats unattended, freeing the observer from the tedium of babysitting repetitive calibration exposures (especially welcome if taking a long series of morning flats!). When the program is running unattended, it is a good precaution to leave notes on the slit room and camera room doors asking techinicans not to enter until the calibrations have been completed. It is also advisable to email stating that flats are in progress and what the anticipated finish time will be.

autoflat first takes all the specified exposures through the bg12 filter, then takes the same number through the bg13. autoflat_alternate works exactly like autoflat, except that it alternates between filters, taking a single exposure with bg12 followed by a single exposure through bg13, repeating the process until the specified number of exposures is completed. autoflat_alternate takes a bit longer, but has the advantage that should the system fail for some reason during the taking of flats, you'll at least have some matched pairs of flats rather than all or most through the first filter.

The procedures must be run from the IDL installation on any of the observatory's user computers (e.g. catpc, gouda, shard). Open an xterm and type idl then use autoflat or autoflat_alternate at described below.

At the idl prompt, type autoflat or autoflat_alternate, followed optionally by its keywords. If no keywords are specified, the program will use the default number of exposures (20) and default times for each filter (35 and 3 seconds for bg12 and bg13, respectively). autoflat and autoflat_alternate have the same default numbers and the same keywords. The following options apply to both procedures (substitute autoflat_alternate when running that procedure):

autoflat use defaults
autoflat,number=n specify number of exposures for each filter, use default exposure times
autoflat,/time specity exposure times for each filter, use defualt number of exposures
autoflat,number=n,/time specify number of exposures and exposure times