The Fan Observatory Bench Optical Spectrograph (FOBOS) is a fiber-fed, bench-mounted, single-object spectrograph. The instrument is designed to observe point sources at moderate resolution to , although extended objects can also be observed if knowledge of the precise spatial sampling is not important.
This manual is intended for the observer and gives the information required to operate the instrument night to night. Although the intention is to instruct a true beginner, reading the manual is absolutely no substitute for hands-on training. NO PERSON should attempt to use the instrument for the first time without having an experienced observer present for supervision. Ideally, anyone interested in using the instrument should accompany an experienced observer for at least one night prior to observing on their own. Additional detailed technical information can be found in the ``FOBOS Technical Reference'', in Jeff Crane's dissertation, and in the FOBOS primary reference publication (PASP, 2005, 117, 526).
Please e-mail Jeff with any suggestions concerning this manual or the instrument itself.
|Jeff Crane||...||firstname.lastname@example.org||Instrument Designer|
|Steve Majewskiemail@example.com||Principal Investigator|
|David McDavidfirstname.lastname@example.org||40'' & Instrument Support|
FOBOS is operated from the 40'' Control Room on the third floor of the observatory. Parts of the instrument itself are located on all four floors of the building. FOBOS can be broken up into three main components: the Focal Plane Module, the Fiber Train, and the Bench Spectrograph itself (See Figure 1).
The Focal Plane Module (Figure 2) mounts to the base of the telescope tailpiece at the Cassegrain focus. Its functions include:
A movable fold mirror carriage just above the telescope's focal plane enables three separate optical paths:
The main length of the fiber train consists of 7 (redundant) ``science fibers'' that transmit light from the telescope to the Bench Spectrograph. In the telescope's focal plane, each science fiber is mounted in a ferrule (Figure 3) and surrounded closely by 6 short guide fibers, which can be used for fine-tuning the target alignment. Although there are several fibers that run from the telescope's focal plane to the bench spectrograph, only one fiber may be used at a time to collect light from a target. The fibers are fixed and cannot be independently positioned. At present, only five of the seven science fibers are functional, and only two are optimized for observation of science targets; the remaining three are intended for collection of diffuse background (sky) emission for subtraction during data processing.
At the telescope level, power and communication cables are tethered to the fiber train as it drapes from the Focal Plane Module to the fork of the telescope (Figure 4). When the Focal Plane Module is attached to the telescope's tailpiece, the fiber train should always be attached to the left arm of the telescope fork using the eyebolts in the fork and snap hooks attached to the fiber train. When the Focal Plane Module is not in use, it should be parked on its lift system to the left of the fork and the fiber train should be detached from the four eyebolts.
It is VERY important to make sure that while slewing the telescope, the fiber train does not catch on any foreign objects, including the tailpiece or hardware attached to the telescope. Constructing the fiber train took several hundred hours of work, and great care should be taken to make sure it does not become damaged.
The Bench Spectrograph (Figure 5) sits on a vibrationally isolated optical table in a stable, environmentally controlled enclosure.
The Fiber Train attaches to the Science Fiber Mount (Figure 6), where the science fibers are arranged in a linear, vertical array. By default, the fiber ends themselves define the ``entrance slit''. However, immediately in front of the ferrules is positioned a slot for an optional entrance slit mask. Following the slit position, there are two slots for optional interference filters and an opal glass used for making ``milky flats''. The instrument's shutter is attached to the front of the Science Fiber Mount.
The Collimator Mount holds a 100mm diameter, 350mm focal length achromatic doublet lens and an iris diaphragm. The focus of the collimator and iris diameter should not normally need to be adjusted by the observer during an observing run.
The Grating Mount contains a rotation stage that can hold one reflection grating from the inventory at a time. The primary setup 12.1 calls for a 100100mm grating with 1200 lines/mm blazed for 6000Å.
The Dewar Mount rides on a linear rail that pivots directly under the diffraction grating's reflective surface. A 135mm f/2 SLR lens on the front of the Dewar Mount focuses diffracted light onto the CCD. The SLR lens front focus should always be set to . The rear focus, once set by technical staff, should not be adjusted. The azimuthal rotation of the CCD with respect to the optical axis of the SLR lens may be adjusted using the micrometer on the top of the mount. This may be necessary to align the spectra along rows of the CCD. The azimuthal rotation of the CCD with respect to the optical table may be adjusted using a micrometer on the rear of the Dewar Mount. This may be necessary to account for tilt in the focal plane, but should not be adjusted after the start of a run.
The detector is a 20482048 SITe CCD with 24 m square pixels operated by an SDSU CCD Laboratory (Bob Leach) Generation II controller. Note that because of non-symmetric vignetting on the red side of the chip, the full 2048 columns are not actually illuminated. The actual usable portion of the spectra will cover something like 1850-1900 pixels, and the quoted wavelength coverage will drop by an equivalent amount.
FOBOS was designed to collect moderate resolution spectra of candidate K giants for the Grid Giant Star Survey. Spectra collected for this project in the Southern hemisphere cover the region 4700-6700 Å with 1 Å/pixel dispersion. To match these spectra, one diffraction grating was chosen for work in first order with no interference filter necessary.
As the instrument's usability is demonstrated, additional diffraction gratings, interference filters, and optional slit masks may be added to the inventory to allow a variety of different configurations capable of covering the full optical wavelength range. As the inventory changes, this section will be expanded to more fully describe the various configurations available to observers.
Please take care to keep the spectrograph room clean. Do not eat, drink, or smoke in the room. Every time you enter, clean the soles of your shoes by planting your feet firmly on the sticky mat in the entryway. If the mat does not feel sticky, step on a different area. When no sticky surfaces remain, peel off and dispose the top layer of the mat.
Standard safe handling procedures should be followed when working with liquid Nitrogen. In the spectrograph room, roll the 25-liter dewar to the end of the optical table nearest the storage area. Insert the ``stinger'' into the CCD dewar's fill tube. Tighten the threaded connector with the spill vent pointing toward the wall and away from the spectrograph optics. Fill the CCD dewar until N begins to spill out of the connector's side vent. This will probably take about 10 minutes if the dewar is already cool. Allow the hose to thaw until flexible before removing the stinger -- otherwise you're likely to break the fill hose in two!
The Bench Spectrograph's optical table is mounted on a pneumatic vibration isolator. This provides some dampening of vibrations in the floor that would otherwise translate to vibrations in the optics on the table.
To enable the system, first make sure that the isolator's air tube is attached to the air compressor's air hose dangling from the ceiling of the spectrograph room. Insert the plastic air tube firmly into the brass and red plastic fitting on the end of the thick black air compressor hose. On the ground floor of the observatory next to the the telescope's support column, find the air compressor. Make sure that the water drain valve on the bottom of the air tank is closed. Also make sure the regulator is closed (turned fully clockwise). Turn the OFF/AUTO lever to AUTO and let the tank's pressure build. The internal pressure should build to about 125 psi before the compressor will turn off. Now open the regulator to pressurize the hose leading to the spectrograph room to about 60 psi.
Return to the spectrograph room and make sure that the optical table has been elevated about 3/8''. If the lift distance varies greatly from 3/8'' the pressurizing or adjustment screws on the leveling arms of the isolator may need to be adjusted. Listen for air leaks in the supply line connection. If you hear one, you may need to tighten the connection.
Turn off the air conditioner and close and clamp the A/C door (Figure 8). Turn off the air cleaner and the dehumidifier. Turn on the power switch to the CCD (gray box next to the dewar on the optical table) if it is not already on. Grab a flashlight. Turn off the lights. Carefully remove the cover of the SLR lens on the front of the CCD dewar. Be careful not to rotate the SLR lens itself; the front focus of the SLR lens should always be set to . Very carefully remove the grating cover and set it aside. Be very, very careful not to touch the diffraction grating. The grating cost several thousand dollars and cannot be cleaned! If you touch it, it will have your greasy fingerprints on it for the remainder of its lifetime. When you exit the room, turn off the light and pull both doors firmly closed behind you.
Follow the standard start-up procedure for powering on the telescope and tailpiece. Turn on the Dell autoguider PC, making sure that the keyboard/monitor switch is turned to the correct position. Once the computer comes up, flip the monitor/keyboard switch and turn on the TCS computer. Start TCS. Turn on the CCTV monitors for the autoguider STV, dome camera, and FOBOS STVs. Turn on the FOBOS PC. Power on the Sun workstation named crux. Log in as user bench. Contact one of the people listed on the front of this manual or a previous FOBOS observer to get the account password. Start the VOODOO CCD control software (see documentation for the GenII CCD camera), DS9, and IRAF in an xgterm. If it's not already on, turn on the FMO EMCS (Environment Monitoring and Control System) PC and start the EMS software.
After the dome room has been prepared (Section 3.5.), establish a connection with the autoguider STV on the Autoguider PC, and with the two FOBOS STVs on the FOBOS PC. The FOBOS coarse acquisition STV communicates with the PC through the COM1 port while the Guide Fiber STV communicates through the COM2 port.
Around sunset, prop the doors to the catwalk open. Open the dome slit and detach the power cord from the dome. Turn on the telescope tube fans if they are not already on. Check the fiber train to make sure that it is hanging properly. It should be attached to the left arm of the fork in several places, almost all the way to the telescope's declination axis. Slew the telescope to the north and remove the cover. Remove the cover of the 8'' autoguider telescope and send the telescope to zenith again. Turn on the power to the telescope autoguider STV control box on. Make sure that the power to the spectrograph's two STV cameras and the electronics control box is on. The power strip attached to the right (West) side of the primary mirror cell should be on as should the power switches on the STV control boxes. Note that the ``CCD'' power switch at the bottom of the electronics rack in the control room must be turned on before power can be supplied to the Focal Plane Module and STV cameras. Make sure the spectrograph's electronics control is set to ``remote''. Open the in-tailpiece shutter using the switch near the back left side of the primary mirror cell. Turn off the air conditioner power on the left wall of the dome room. Set the humidistat-controlled heat lamps to the off position. Turn off the lights and shut the door to the dome room on your way out.
Don't touch the surface of diffraction grating, SLR lens glass, or collimator glass.
Don't crush, yank, or tightly bend the PVC pipe containing the fiber optics. The fleas of a thousand rabid camels will be set upon your carcass if you break the fiber optics.
For any scientific target, it's a good idea to split up your observations into three separate exposures to be combined later. This simplifies cosmic ray removal.
To adequately calibrate your spectra, you will want to take the following additional data:
During normal observing, the user will need to control the Focal Plane Module fold mirror carriage, wavelength calibration lamps, quartz lamp, coarse and fine acquisition cameras, autoguider camera and telescope, 40'' telescope, and SITe CCD.
The fold mirror carriage and calibration lamps may be controlled either by using the electronics control panel attached to the Focal Plane Module or by using the remote paddle in the control room (Figure 9). A Local/Remote switch on the control panel determines which location has control. In both places may be found ON/OFF switches for each of the four calibration lamps (QTH, Ne, Ar, Xe) and three push buttons that send the mirror carriage to its three available positions.
The three carriage positions are labeled Calib, Observe, and Acquire. In the Calib position, a fold mirror enables the calibration lamp system. In the Observe position, the focal plane ferrules are exposed to light from the telescope. If a target is aligned on the end of a ferrule, light will find its way to the Guide Fiber STV camera and Bench Spectrograph. In the Acquire position, a fold mirror enables the primary acquisition (coarse) system and the Acquisition STV will show a 6'4.4' telescope field of view. When the carriage is at any one position, the red LED above that button will glow. If the FP Module is powered on but none of these LEDs is lit, the carriage may be stuck between its three normal stops. In this case, a manual switch on the control panel can be used to drive the carriage until one of the LEDs lights up. If the carriage is run past the extreme positions toward its hard limits, limit switches will disable the motor. In this case, a limit override switch on the control panel must be depressed and the carriage driven manually back away from the limit. Be very careful not to drive the carriage to its hard limits! If you do, you may destroy the motor or drive nut.
The FOBOS STVs may be controlled by either the control boxes attached to the Focal Plane Module or by the STV REMOTE software on the FOBOS PC. Similarly, the autoguider STV may be operated by the control box attached to the telescope tailpiece or by the STV REMOTE software installed on the Autoguider PC. Other autoguider controls include a fine focus adjustment and East/West slew controlled by hand paddles in the control room. See the autoguider documentation for instructions about running that system.
The 40'' telescope is controlled by the DFM Telescope Control System (TCS) and by hand paddles in the dome and control rooms. The SITe GenII CCD is controlled by the Sun workstation crux. See the manuals pertaining to those systems for more information.
In an xgterm on the crux workstation, change to the $HOME/iraf/ directory and start IRAF with the cl command. Within IRAF, change to the /data/bench/ directory and create a new directory named for today's date. Start the ds9 software for image display. Note that because crux is set up for 24-bit color, ximtool and SAOimage will not work.
Refer to ``The GenII CCD Camera System (Spectroscopy)'' manual for detailed instructions about running the CCD control software, Voodoo. For spectrograph work, Amplifier C is preferred with Gain 1.0 and Slow Integration speed (setup file /crux/bench/fobos/C1S.setup). For faint targets, Gain 2.0 may be useful (setup file /crux/bench/fobos/C2S.setup). Set the CCD for subarray readout with dimensions 2048200 centered at [1024, 1024]. Set the Bias Position at 2080 and the Bias Width at 20. In the FITS setup menu, load the file /crux/bench/fobos/FOBOS-fits.par and check the TCS Link box. These steps will ensure that your image headers have the keywords required for spectral reductions. Make sure the Open Shutter, Beep, and Save to Disk boxes are checked in the main Voodoo window. Set the output directory to /data/bench/date_today/ and initialize the file name to something like ccd1001.fits. Note that IRAF wants the image filename extensions to be ``fits'' and not ``fit''.
The STV cameras used in the spectrograph and with the autoguider may all be controlled using the STV Remote software. Run the software on the autoguider PC and connect to the guider STV through port COM1. On the FOBOS PC, run two instances of the software. Connect to the Acquisition STV through COM1 and the Guide Fiber STV through COM2. Refer to the 40'' manual for further instructions about running the autoguider. For the spectrograph STVs, it is not important to set the correct Date/Time, telescope focal length, etc.
You may begin imaging with the STV cameras immediately after establishing successful links. Click Image and then click Parameter repeatedly to see the adjustable parameters. Set each parameter by clicking the Value button. Choose the ``Normal'' zoom mode and the 2 gain setting. Finally, click Image again to start the continuous video stream to the monitor.
You will want to take a set of calibration images during or before each night of observing. These should ideally be done after the spectrograph room has been prepared for observing and the air has settled (i.e. when the spectrograph room is in a state most similar to nighttime observing).
A set of milky flats should probably be taken shortly after the room is prepared in the evening. An opal glass filer must be placed in Filter Slot 2 (Figure 6) and the top of the Science Fiber Mount replaced. Move the mirror carriage to the Calib position and turn on the QTH lamp. Take a series of exposures with a few thousand counts each. Remove the opal glass filter using the threaded brass tool next to the Science Fiber Mount and replace the top. Allow some time for the room to settle before taking additional calibration or targeted spectra.
Ideally, the instrument should be so stable that a single QTH lamp spectrum and a single wavelength calibration lamp exposure taken at the beginning of the night would suffice to calibrate the entire night's data. However, until instrument stability can be established, it is advised the QTH and arc lamp exposures be taken several times during the night. To do so, make sure the mirror carriage is in the Calib position. To take a QTH exposure, turn on the QTH lamp and set the exposure time to something like 1 second. The wavelength calibration spectra are slightly trickier. In the primary instrument setup, the Neon lamp is considerably brighter than both the Argon and Xenon lamps. Set the exposure time to 60 seconds or more. Turn on all three arc lamps. Make sure the intercom to the spectrograph room is on and the volume is turned up. Start the exposure and listen for the shutter to open. As soon as the shutter opens, turn off the Neon lamp. Leave the other two on for the duration of the exposure.
Following the normal start-up procedure for the 40'' telescope, align the telescope on a known, bright star using the finder scope. Move the fold mirror carriage to the Acquire position and begin imaging with the Acquisition STV. Select input 1 for the FOBOS STV monitor. The bright star should be in view. Move the telescope using the hand paddle until the star is centered in the STV field of view. Now enter the star's coordinates in the TCS's telescope position initialization function.
Important: For the first few times the telescope is slewed to a new position, every time a large () slew is performed, and especially when large slews toward or away from the Northwest are performed, walk up to the dome and make sure the fiber train does not get caught on the telescope, tailpiece, or any other foreign object.
The Focal Plane Module must be focused. This can be accomplished by moving the telescope's tailpiece focus until the bright star used for coordinate initialization is in focus. The distance from the fold mirror in the carriage to the focal plane ferrules is the same as the distance from the mirror to the object plane of the coarse acquisition system. Therefore, focusing the image on the STV camera has the effect of bringing the ferrules into the telescope's focal plane. Note that there are aberrations in the image produced by the very simple optical system used for coarse acquisition. Stars will appear point-like in some areas of the image, but may have a ringlike appearance with a bright, off-center core in other areas. When focusing the instrument, attention should be paid only to the bright core of the star, or better -- the star should be positioned in the less aberrated area of the camera toward the lower third of the monitor. On 13 Oct 2003, with an exterior temperature of F, the spectrograph focus was at .
The star can be focused by eye, but perhaps a more accurate method is to use the ``Optical Quality'' mode built in to the STV controller. Move the telescope to position the star on the screen in an area where the optical aberrations appear minimal, but not too close to the edge. Press the Monitor button. Press the Parameter button until you see ``OPTICAL QLTY''. Push the Value button. The STV should detect the star and begin monitoring its profile. Adjust the telescope's focus while noting the change in the FWHM reported by the STV. Set the focus so that the FWHM is minimized. Note that the actual number reported is meaningless unless you enter the telescope parameters in the STV's Setup menu (not required, but may be interesting).
Once the telescope position has been initialized, coarse target acquisition may commence. Slew the telescope to the target's coordinates. Make sure the STV monitor switch is set to input 1. The target should be in view. If it is relatively faint, you may need to bump up the STV exposure times to see it.
Two science fibers are available for targeted acquisition. These are fibers 1 and 4 in Figure 12. For each of these ferrules, the science fiber is intact and the guide fibers are properly aligned around them. Fibers 2, 3, and 5 have functional science fibers, but the guide fiber arrangements are flawed, so the prescribed fine alignment procedure will not work. However, these can be used to collect sky spectra for background removal during processing. Fibers 6 and 7 are completely nonfunctional. The positions of Fibers 1 and 4 have been marked on the TV monitor with a grease pencil, with Fiber 1 being nearest the center. Using the hand paddle, move the telescope to position the target over the desired focal plane ferrule position. Fiber 1 appears to have the best throughput, so it is preferred.
When the star has been approximately aligned with the focal plane ferrule, move the fold mirror carriage to the Observe position. Switch the STV monitor to input 2. You should see light coming through some/all of the guide fibers corresponding to the science fiber chosen (See Figure 13). If the target is faint, the integration time on this STV may need to be increased in order to see the signal.
Using the hand paddle, move the telescope slowly until the light coming through all of the guide fibers equalizes. This indicates that the source is centered, and therefore over the science fiber. This procedure will only work for point-like sources; extended objects can only be coarsely aligned. When the target's fine alignment has been established, the autoguider may be engaged if necessary.
An auxiliary autoguider is available for use during long exposures. The autoguider is an STV attached to a piggy-backed 8'' Meade telescope on the side of the 40'' tube. See the 40'' manual for instructions on operating that system. Note that the autoguider must be recailbrated every time the telescope is moved significantly in declination. Even while using the autoguider, you should periodically check the guide fiber output to make sure the guiding is working well. If the alignment appears to worsen, turn off the guiding, correct the telescope alignment, and then re-engage the guider.
At the time that these efficiency data were collected, the telescope's mirrors had not been aluminized for more than four years. Due to persistent problems with humidity and condensation at the site combined with the advanced age of the mirror coatings, we expect that the telescope's efficiency has adversely affected the total system efficiency as plotted in Figure 14. Thus, the system performance will likely improve following the next mirror realuminization. With that caveat, estimates of peak signal to noise ratio (S/N) per pixel as a function of exposure time and target magnitude are presented in Figure 15.
On the ground floor of the observatory, turn the air compressor's OFF/AUTO switch to the OFF position. Turn the regulator counterclockwise to set the outlet pressure to zero. Detach the isolator's air tube from the compressor's air hose in the spectrograph room. To do this, press the the end of the red, plastic connector in while you detach the hoses from one another. Open the regulator on the compressor to depressurize the air tank. Close the regulator when the pressure reaches about 20 psi. Open the drain valve under the air tank to drain accumulated water. When the water has drained, close the drain valve.
Carefully replace the cover on the grating and then the cap on the SLR lens. Fill the dewar using the same procedure outlined in section 3.1.. Open the A/C door and turn the A/C to Medium Cool. Turn on the air cleaner. Turn on the dehumidifier to the halfway point. When the dewar is full, turn off the lights and pull both doors firmly closed behind you.
Replace the telescope cover and send the telescope to zenith (turn off telescope tracking on the electronics rack in the control room first). Turn on the A/C power switch. Set the heat lamp control to ``humidistat'' and set the humidistat to 50%. Power off the spectrograph electronics control box, two STVs, and the autoguider STV. Close the in-tailpiece shutter. Close the exterior doors and dome slit. Turn off the lights and close the door when you leave.
Back up your data. If you are at the end of a run, delete your files from the hard drive. After backing up your data, log off of crux. Shut down the autoguider and FOBOS PCs. Power off the TCS PC. Turn off the TV monitors. Send the Focal Plane Module fold mirror carriage to the Calib position. Power down the telescope and tailpiece in the standard way. Leave the FMOEMCS PC and software running. Fill out the observing log. Turn the lights off and close the door when you leave.
Problems? Here are a few suggestions to remedy problems that have occurred so far...
An IRAF package called FOBOS has been written to assist with data reductions. This is installed on crux at FMO and also locally in the Astronomy Department. The available routines include:
To make full use of FOBOSHEAD and FOBOSLOGS, you'll want to keep a text comments file while observing. This file should have the following format: the image prefix begins each line, followed by a colon as a field separator, then the object name, followed by a colon and any comments for the file on the same line. When running FOBOSLOGS, this file will be interpreted by LATEX, so any special LATEX characters must be ``escaped'' by a backslash (). A few lines from an example file follow:
ccd1035 : HD4388 : RVstd K3III V=7.34 v\_r=-28.3 km/s ccd1036 : HD4388 : ccd1037 : HD4388 : ccd1038 : QTH : ccd1039 : NeArXe : Neon off after < 1 second
Refer to the FOBOS package help files for more detailed information. Peter Frinchaboy has also written a helpful ``Cookbook'' for FOBOS data reductions. In brief, a typical reduction for a run might look like this:
Figure 16 shows a comparison lamp spectrum taken with the primary setup in the manner described in Section 4.4.. A subset of the identifiable lines is labeled. Most of the Neon lines are considerably stronger than the bulk of the Argon and Xenon lines. However, all will be useful for wavelength calibration provided that the strong lines do not saturate and the weak lines have good singal-to-noise.
Note that there are no line identification lists that come with the default IRAF NOAO installation that are appropriate for this instrument. The commonly used idhenear.dat line list will not work well because it contains no Xenon lines, but does contain Helium lines. A special line list, called nearxe_ggss.dat has been prepared for the primary FOBOS setup. This was created by using Argon and Neon lines taken from the IRAF henearhres.dat line list and Xenon lines taken from the National Institute of Standards and Technology website. Each of these lists was then used to identify lines in spectra of individual calibration lamps, and those lines that did not appear in the spectra were deleted. nearxe_ggss.dat is what remained. Some lines in this list seem to work better than others. In particular, you may choose to delete lines that are blends or closely spaced. This line list has been installed in the IRAF linelist libraries on crux at Fan Mountain, and locally in the Astronomy Department. To use the list with one of the IRAF identify procedures, enter linelists$nearxe_ggss.dat for the coordli parameter.
Neon is very good for the red part of the primary setup. Argon has fairly good coverage throughout. Xenon is mainly useful in the bluer region.
Figure 17 shows a blue Argon + Xenon comparison lamp spectrum (60s exposure at low gain) taken with the ``blue grating'' (1200@157 in Table 1). The line identifications, listed in the file linelists$arxeb.dat, may be useful for wavelength calibration of blue spectra.