Keep an eye on the bias frames. Often the first sign that a CCD is being to act strangely is horizontal banding on the zero frames. Check to see that there are no faint gradients in intensity or sudden jumps ("banding") perpendicular to the dispersion direction. If you see this or anything else that bothers you, call TelOps.

If you are working far into the blue with a high dispersion grating, you will not get enough light for these flat fields. In this case, you will have to use the projector lamp (lamp 4 or 5) with the periscope at the comparison position. This setup is less precise a calibration than the dome flats because the projector lamp does not produce a good parallel beam. You must observe this lamp with the ``C/X/A. Automatic Comparison Sequence'' which is the procedure where each fiber pair is brought into the field center sequentially. This ``standard sequence'' observation is preferred to the comparison position, since every fiber (object and sky) is brought exactly into the field center to be exposed. Note that this sequence only observes "Quartz Lamp 4" (the fainter one). You will have to adjust the Quartz lamp intensity so that each fiber is exposed for a few seconds. Note that the command "C/X/A. Automatic Comparison Sequence'' does lots of things - it brings the periscope to the comparison position, the fibers to the standard position, the mirror to its position, and controls the light. You must have the correct value for the Quartz 4 mirror position stored in the mirror lookup table (in "D/C. Change mirror positions") for this task to work.

Take at least 5 frames to get a decent average. The domes should be exposed to 20000e- or so, if possible. If you are working in the blue, try to take as long an exposure as possible, taking care not to saturate in the red. I often leave the domes running at the end of the night.

If you want to use these for velocities, don't forget to take a wavelength calibration lamp afterwards.

Curiously, getting a high quality calibration lamp exposure series is the hardest of all the calibrations to make with Argus. One cannot simply use the internal calibration lamps in the Argus head because the effective f-ratio of the light from the calibration lamps for the object fibers and sky fibers is slightly different (we do not have an integrating sphere arrangement). For the calibration frames taken with the internal Argus lamps, there is an artificial shift of a little less than 0.1 pixel introduced into the sky fibers.

Instead, one really needs an emission line spectrum taken off of the dome white spot to get the ideal wavelength calibration. Right now, we do not have good lamps to do this. If you are working blue at low resolution, you can use the dome HgOB lamps (the ``work lights'' operated by the Night Assistant), but the Hg lamps do not have many lines. In the red (5800A and redward), there is a neon lamp that is attached to the secondary ring (ask TelOps), or you can use the Meinel bands of OH. The dome neon lamp is a very faint lamp and a couple of 30 min to 1 hour exposures are needed to get a good spectrum. Get 3-5 dome wavelength calibration spectra.

If you cannot get a wavelength calibration lamp spectrum from the dome white spot, you can use the ``C/X/A. Automatic Comparison sequence'' while observing the internal Argus calibration lamps. This procedure brings first the object fiber into the center, flashes the comparison light, followed by the sky fiber, for all the fibers. This is much better than just observing the calibration lamp with the fibers in the calibration position because both the object and sky fibers are moved to the same position. Alternatively, you can use the twilight sky absorption spectrum to do the wavelength calibration. At lower dispersions however, the twilight skies are not easy to use as a wavelength calibration source.

The Argus echelle mode requires the thorium-argon lamp. This lamp is quite weak and in order to get enough intensity, TelOps will remove the diffuser screen in the periscope. The comparison light will now only appear as a small beam exactly in the center of the Argus field. The only way to get the comparison light into the fibers is to bring the fibers into the center one at a time with the "C/X/A. Automatic comparison sequence". You should use the thorium-argon lamp to monitor spectral drifts in the object fibers during the night, but since it will take a long time to do all 24 fibers, I suggest only doing 4 or so fibers with the ``automatic comparison sequence.'' See the Random Hints section IX below.

Although the milk glass does a good job of diffusing the image, you will still be able to see very strong stellar features, such as Ca II H+K as a slight dip in the dispersion direction. If you see any discontinuities in intensity, the milk glass is probably reflecting light into the beam. Ask TelOps to tilt the milk glass slightly.

Generally, I found that I had enough time during the day to get the milk flats, biases, dome flats, and twilights. At the end of the night, I left the dome comparison lamps running while I went to bed.

Summary of calibrations:

Calibrations once a day:

1. Dome flats (5-10) and/or "automatic comparison sequence" of the Quartz Lamp 4 (3-5)
2. Zero frames (25)
3. Twilight frames (5)
4. Dome neon/Hg and/or "automatic comparison sequence" lamps for wavelength calibrations

Calibrations during the night:

1. Comparison lamps at the comparison position, or comparison lamps of a limited number of fibers with the "automatic comparison sequence" in echelle mode (once every 1-1.5 hours)

Calibrations once a run:

1. Milk flats (5-10)

VI: SETUP AT THE BEGINNING OF A TYPICAL NIGHT.

At the beginning of the night, a few telescope calibrations must be made. The night assistant will first do a zero-point of the telescope. You must bring some fiber and the periscope to the center of the Argus field. Illuminate the fiber and the Night Assistant will mark the position with a cursor on the TV in the periscope field. He will then bring a bright ephemeris star onto the cursor and zero the telescope coordinates. Afterwards he will probably check the offset from the field center to the main field and mark the position in the main field.

You must focus the TV camera. This is most easily done when the sky is still bright. The edge of the periscope has been carefully set to occur exactly in the focal plane of the telescope. You can focus the TV by entering ``Keypad mode'' on the Argus computer and typing ``T''. Set the step size to 10. If the periscope frame is grossly out of focus, you must bring the TV focus up to about 700 with larger steps of 50 or so. I find that the focus is repeatable to about 15 units. Once the TV is focused, the Night Assistant can focus the telescope in the main field with the handpaddle. Note that at NO time do you ever focus on the images in the periscope, which are not quite in focus.

Note the logic of focusing: the rim of the periscope is machined to be in the focal field of the fibers. You therefore focus the RCA TV on the periscope rim first, then focus the telescope in the large field. Objects should also be in focus in the periscope, but this is not guaranteed. The TV stays pretty much in focus on the periscope rim all night but you may want to check this. The main field telescope focus changes around the sky and drifts by -125 units per +1 deg C.

If you want to focus the TV during the night, do the following.

• Have the Night Assistant turn down the RCA TV
• Close the main field shutter. This is done in Keypad mode with the "Grey -" key on the Argus keyboard.
• Have the Night Assistant bring up the TV to where you can clearly see the periscope rim
• Focus in steps of 10 units
• After you are done, don't forget to turn off the field light and open the field shutter.

Here is one other trick on focusing the TV. I usually focus by noting where the rim goes out of focus on both sides of the TV focus, and then averaging the two values. It takes about 70 steps to go from one side of focus to the other.

VII: OBSERVING PROCEDURES

Give the coordinates of the field center to the Night Assistant who will begin the slew. You may not want the slew to start until after the chip is read out, but I have never noticed anything evil if the slew is started early. It is said that slew can sometimes add noise to the CCD frame, but I have never seen it. If you decide to save time by reading the CCD during a slew, keep an eye open for extra CCD noise, banding, etc. If you see banding, don't blame me.

Enter the object and assignment files into the Argus computer from floppy, diskette, or hard disk. Move the fibers to the object positions. This can be done during a slew.

If you have an object at the field center (which is the recommended procedure although not always feasible) send the periscope to the center. If you have assigned a fiber to the central object, move the telescope so that the fiber image and object align. At this point, the telescope should be set correctly and the Night Assistant can start guiding.

On the guide/acquisition TV (RCA ISIT) you will see:

                    ------------------------------------
                    |                                  |
                    |                                  |
                    |                                  |
                    | 37"          Main field          |
                    |------        24' 30" S           |
                    |peri- |       of optical axis     |
                    |scope |37"                      3'| East
                    |field |     +                     |
                    |------                            |
                    |                                  |
                    |                                  |
                    |                                  |
                    |                                  |
                    |               4'                 |
                    ------------------------------------
                                  North

The main field is 24' 30" south of the optical axis; that is, with the object centered at the ``+'' in the main field, you must move the telescope 24' 30"S to place the object in the center of the periscope located on the optical axis. Note that if you have an object on the periscope that is NOT on the optical axis, doing the 24' 30"N offset will NOT put the object in the main field!

If no fiber is assigned to the central object, just move the central object roughly to the periscope center. In this case or in the case that you have no central object, now move the periscope to a fiber assigned to one of the brighter objects closest to the center. Move the telescope so that the object and fiber align. The Night Assistant should now start guiding the telescope on a star in the large field.

With the telescope now guiding, you should now check all 24 fibers to see if they are centered on the fibers. In our tests on astrometric positions (NOT HST GUIDE STAR POSITIONS!), we have found a fiber pointing to 0.8 arcsecond rms, or roughly four steps. A few fibers sometimes are off by up to 10 steps or so, due to a presently undiagnosed mechanical problem. Use a step size of 1 or 2 to bring the fibers onto the objects.

To center the fibers onto the objects, turn on the fiber illumination lamp with the digital potentiometer. With the periscope on a fiber, bring the fiber illumination up to a level where you can just barely see it. If you bring the intensity up too high, the TV image will bloom and the image center of mass will appear to move along the scan line.

One trick that I use to increase my efficiency in tweaking the fibers is the use of two step sizes. For most fibers, I just need to step the fiber once or twice with step sizes of 1-2. In order to judge the offset, I first step the fiber in one direction (say, down) by a large step - typically 15 units. With the fiber well off the star now, it is easy to tweak the fiber left-right to get the fiber directly below the star. Step the fiber back up to the star with a big step. Now step the fiber to the side with a big step, and adjust the up-down position with the small steps. When you are done, make sure you step the fiber back to the star before moving on!

The procedure I use to center extremely faint objects is:

• - allow the TV to leak for a few seconds with no fiber illumination.
• - quickly turn up the pot until the fiber is barely visible then turn it off
• - remembering the relative offsets, dead reckon the now invisible fiber onto the object.
• - quickly turn up the fiber illumination to verify that the fiber is on the object.

I have found that after a few hours of moving the fibers, I can dead reckon the fibers onto the object in a single offset. When I get efficient, I can tweak up all 24 fibers in about 10 minutes on a faint field. The "=" key especially useful in speeding things up because it moves to the next fiber and also moves the periscope.

Once you have tweaked the fiber positions, you must turn off the fiber illumination lamp, bring the periscope home (out of the beam), and take a throw-away bias frame (using preview) to clean the chip. Unfortunately, the fiber illumination system presently floods the chip and you must take a bias frame before you observe to wipe the chip.

You may want to write a crib sheet to remember the sequence of events in observing object fields and comparison fields. Here is my crib sheet:

The 4-Step Calibration sequence:
1. Fibers to calibration position (can be done while chip is reading out).
2. Periscope to calibration position (can be done while chip is reading out).
3. Mirror to position 1 (or 2,3).
4. Lamp 1 (or 2,3) on. Observe!

The 3-Step Observing sequence (after fibers positions have been tweaked up on the objects):

1. Turn off fiber illumination lamp.
2. Do a zero second exposure with preview.
3. Periscope to home. Observe!

VIII: POSITIONER COLLISIONS:

Let's hope this never happens to you, but it may.

Argus detects a collision via electrical contact between the arms. For almost all collisions, Argus will immediately try to pull back the arms involved in the collision RADIALLY until the electrical contact is broken. It uses repeated small steps of 1 unit. If the positioners have collided in the field of view, an outward radial motion should always resolve the collision. If this radial motion does not resolve the collision, Argus automatically tries the azimuthal motion for the hell of it. If it still cannot undo the electrical contact, it will timeout and tell you it failed.

At least 90% of the collisions are successfully resolved this way. You are now faced with trying to understand what happened. In most cases the problem has occurred because two fibers are close together, and one of fibers has lost its position slightly. This type of collision often happens when you go to the comparison position, because the fibers are placed very close together in a circle. Another common collision problem occurs when two fibers are very close, and during the "tweaking up" process, you happen to move a fiber using the keypad into a collision.

You can sometimes recognize the fiber that is causing the collision by merely looking the actual fiber positions on the TV and the computed fiber positions in the Argus keypad mode. If a fiber is seriously lost, the actual and computed fields will quickly tell you who's been a naughty positioner.

Unfortunately, there are a few cases where this simple procedure will not work. The worst case is when the positioners collide BACKWARDS; that is, the positioners extend behind the circle that you can view with the field TV, and their backsides can collide outside this field, especially if the fibers are retracted towards the edge of the field and are at some extreme azimuth. In this case, the outward radial motion is exactly the WRONG thing to do.

If you are in an unresolvable collision, you should call TelOps. They will try to resolve the collision by manually moving the fibers by doing the following:

1. - Go to Diagnostics and enter the Keypad mode. This is a special keypad mode called ``pulse mode'' where the arrow keys move the fiber positioners RADIALLY and AZIMUTHALLY, not N-S and E-W. You can also enter pulse mode from any keypad mode by typing ~ .
2. - Note the sense of the arrows in pulse mode. The up arrow is radially in, and the right arrow is clockwise. An easy way of remembering this is that the arrows move positioner 13 (the one at the bottom) in exactly the the directions indicated by the arrows.
3. - Determine whether the collision is in the field or a backwards collision.
4. - Move the positioners radially or azimuthally. Note that when in collision mode, Argus ignores all step sizes and only moves the positioners by steps of 1 unit.

If all this fails, you have a serious collision. TelOps and your Night Assistant will have to bring the 4m to the PF platform, open the PF cage, and press a little red emergency-override button while someone moves the positioners in units of more than 1 step. Try 5 units at this point. The red button allows you to step the positioners by more than one step (and also allows you to damage the positioners if you are not careful).

One final type of ``collision'' that occurs is when the periscope gets completely lost and gets pinned azimuthally (to the right). In this case, go to pulse mode, move the periscope azimuthally to the center and then radially back to home. Do a Periscope Home Checkout to reset things.

After a collision, you are left to figure out exactly what happened. You can retire the fiber or try to recover its position so that you can continue using it. If you suspect the fiber will never be useful, you should retire it. To do this, select the bad fiber with "D/X/F/S. Select Active Fiber" . Move the bad fiber home with "D/X/F/H. Home Position Motion." Then edit the fiber table to disable this fiber in "D/X/F/E. Enable Fibers." Make sure you note the bad fiber in the nightly report so someone can try to fix it the next day.

Note that if you retire a fiber to its home position, it still can be involved in a collision - a backward collision. In the home position, the fiber positioner's tail sticks out backwards quite a distance. If you go to a new field and try to move to the new fiber positions, the program will check for collisions and it may detect a collision between your retired fiber and an active fiber. Although this situation is quite unlikely, you may have to redo your assignment.

If you want to try to recover the fiber, you should have TelOps or your night assistant do the following procedure. We list it so you can figure out what they are doing.

1. Rezero the home position of the bad fiber(s). Do a "home checkout" with "H/V. Verify home position". This will drive the bad fiber up against its hard stops. You will hear the loud clicking of the positioner against the stop. Do this twice.
2. Bring all the fibers to the standard position with "C/X/S. Standard position."
3. Bring fiber 7 or 8 into the center of the field. We recommend fiber 7 at this time. You should use the fiber that was used in the afternoon checkout session with the WATEC camera.
4. Move the periscope so the WATEC camera is looking at the field center with the keypad mode command "@".
5. Back-light the fiber and adjust the RCA tv to see the fiber in the center of the field.
6. Mark the position of the fiber with a grease pen on the tv.
7. In the keypad mode, step out the reference fiber, and bring the bad fiber into the center. Once it is centered, hit "A: adjust fiber center" Do this for all the bad fibers.

This should take at most 15 minutes.

One very important thing to note is that if a fiber is really lost, don't assume that the fiber that comes up in the field is the object fiber. It is possible to make a mistake and center the positioner on the sky fiber! You can always check which fiber it is by moving the fiber radially (in pulse mode) and remembering that the object fiber is at the largest radius.

Suppose lots of fibers are screwed up, and it is 5:30 am and you need that last field in Horologium or your s.o.b. collaborator is going to kill you. What are you going to do? An quick emergency procedure to get Argus back into a shape where it can be used again is the following:

• Re-zero all the fibers (or just the wacko ones if there are only a few bad fibers) with a "home checkout" using "H/V. Verify home position". Do this twice.
• It may happen that there will still be some collisions as Argus tries to go to the home position. If this is the case, turn down the RCA TV, turn on the field lamp, see which fibers are colliding, and using "pulse" mode in the keypad, try to retire the fibers manually. Once there is no collision, use a step size of 300 units (which is the maximum that Argus allows xxx?) to get Argus to a position where the fibers are well away from each other. Then do the two "home checkouts."

The "home checkout" should get the Argus fibers back to a state where they can position to 6 or so steps - close enough to get the fibers on the stars by eye. Note that the "home checkout" re-zeros the starting positions of the fibers, but does not erase the central positions that were stored in the afternoon during the WATEC procedure.

IX: OTHER RANDOM HINTS, MAYBE EVEN HELPFUL ONES

To display you image on the SUN, I suggest set stdimage= imt31 which gives a stdimage about the same relative size as the Loral 3K. The stdimage devices are listed in /usr/iraf/dev/imtoolrc . To see the whole image, plot with
display obj000 1 zs- zr- z1=700 z2=1000 fill+ .
You can use imexamine to look at the data or implot . I prefer implot , with step=100 to step through the columns quickly, to get a good idea of the relative flux in the fibers as a function of wavelength. The m key in imexamine is also a good quick way to get the maximum counts in a small region.

If you are experimenting with Argus and need to take lots of test exposures to judge, for instance, the best mirror tilt for the comparison exposures, consider setting the binning factor to something like 5x2 up to 5x5 to speed up the read time. Use "setdet force+" in the "IRAF Acquisition" window. Don't forget to switch back after you are done!

Don't waste exposure time. If you can't figure out what the name of your object is, start the exposure anyway, and use "pictitle" to edit the title before the exposure is done. Or use "hedit xxx.imh title objectname up+" if the image has already been taken.

Most observers like to have the Argus simulator running on another PC in the console room. If the PC is missing, request it from TelOps. Make sure the PC is linked to the SUN via the PC/NFS, and that you have run "gonfs" to establish the link. "gonfs" also allows you to use the PC as a SUN terminal. Use "telnet" to log on to the SUN if you want to use it as a terminal. If you want to use it as an Argus simulator, change directory to C:\ARGSIMUL and run "SARGUS C:".

The problem of wavelength calibration with the echelle is rather critical. The Th-Ar lamp is weak and only produces a small patch of light in the center of the field. You cannot use the comparison position to take the Th-Ar lamps. The fiber must be brought into the center of the field using the standard exposure sequence to get enough light. If you use the "C/X/A. Automatic Comparison Sequence" on 24 fibers, this will eat up a lot of observing time. Instead, you may want to only bring in, say, 4 fibers (1,8.16,24) and use these to calculate an average shift for the whole frame. Pat Seitzer, in a program to measure accurate radial velocities, found this technique quite adequate for radial velocity work. To only bring in a small specified number of fibers, edit the fiber table using "C/X/E. Enable/Disable table". Note that you are only enabling/disabling the fibers for the standard and comparison sequence. The fibers are still active for Argus motion for program fields.

If you wish to observe a standard star in all the fibers, or in a certain number of them, you can bring one fiber in at a time to the center, pull it out, bring in another, etc - all in a single CCD exposure. To do this, you must go to the menu "C/X. Standard Exposure Sequence" (this is rather confusing because you are NOT doing a comparison lamp here). Edit the fiber table for the fibers you wish to bring into the center using "C/X/E. Enable/Disable table", bring the fibers to the standard position with "C/X/S. Standard position". Bring a fiber into the center, center the star on the fiber, and pull out the periscope. Take an exposure with "C/X/T. Take exposure." Before the Argus sequence starts, the Argus program "C/X/T" will tell you to start a really long exposure on the CCD computer. Note that the accuracy of Argus is such that the small positioning errors will cause fiber-to-fiber variations in the detected flux.

At the time of the writing of this manual, the atmospheric dispersion corrector (ADC) is not turned on. The ADC is a part of the new prime focus corrector. The PF corrector is the one you are using, and the ADC is left in a neutral position. We hope to have the ADC functioning as soon as possible. Ask us about it when you arrive.

There may be times when for some reason, the Argus program loses communication with the Sun or some other peripheral. In this case, you will get the dreaded DOS message "Abort, Retry, Fail, or Ignore?" that appears in native text. DO NOT ABORT! This will kick you out of the Argus program, and the positioners may get permanently lost. TYPE "R" OR "F". Only use "A" if all else fails. Sometimes "F" followed by "ESC" works.

The dewars should hold LN2 for the whole night. Prior to the beginning of the night, ask the Night Assistant to fill the dewar and check that all the doors into the bench spectrograph room are closed. With that done, there is NO reason to go into that room for the whole night. The less traffic into the bench spectrograph, the smaller the drifts you will see.

The present TV camera in the 4m Argus is an RCA ISIT, which can be damaged by bright light. Be careful not to turn on the field lamps or dome lights while the gain is up on the RCA. We have also installed the CCD Acquisition camera for those trying to see very faint, but this camera has a very limited field of view, and it cannot be used if you mix bright and faint stars. Ask us if you need it.

If you observe in bright moon, you can put a red filter in the TV to improve the contrast. While this cuts the moonlight a lot, it also makes it difficult to go faint, because the TV camera is not very red sensitive. Remember that the TV should see roughly the same color light as you are detecting in the spectrograph. If not, at high zenith distances, the TV image may be offset from the desired image due to atmospheric refraction.

Carefully inspect the first dome or twilight flats for unwanted ghost images. The order separating filters sit in the middle of the collimated beam and can reflect light directly onto the grating. TelOps has been instructed to very slightly tilt the filter to get rid of the reflections.

If you leave the calibration lamps on during the exposure, your frame will not be affected, but it is a good idea to turn off the lamps after use, to save lamp life and avoid any possible light leakage into the prime focus field.

Sometimes the periscope begins to get slightly lost and your star always appear at the edge of the periscope field. To fix this, bring a fiber to the center of the field. Go to menu "D/X/P. Periscope Operations." In the keypad mode, center the periscope on the fiber (NOT vice versa) using steps of 10 or so. Once the periscope is centered, run "D/X/P/A. Adjust center reference" to reset the center position of the periscope.

If you do not use all the fibers because you do not have enough objects to assign, you should assign the unused fibers to random sky positions in the field. This will at least allow you to produce sky spectra of higher S/N that can be later used in the sky subtraction.

There are reports that the hydrogen-doped fibers can lose their blue transmissivity in time. We will attempt to monitor the fiber throughput each engineering run. If you suspect poor throughput, let us know.

The program "D/E. Enable/Disable Fibers" has an extra feature which will allow you to set all the fibers to E or D by typing -1 or -2 for the fiber name.

The comparison circle has a radius of 400 units or about 75 arcseconds.

If you are find that there are difficulties in reading your positions or assignments on the Argus PC, consider the following common mistakes:

1. Wrong epoch.
2. More than 80 chars on a line.
3. More than 1500 objects in an object file.
4. Blanks occur in the name field.
5. No proper motion is given. You must enter 0 0 if there is no known motion.
6. Position table was changed after the assignment file was created.
7. (this is a subtle one). The object appears more than once in your list. Normally this is not a problem for Argus, except if you declare this object as the center reference and also assign it. This can cause a divide by zero error whose only symptom is to kick you all the way out of the Argus program with no error message. We suggest you test your data against multiple entries by sorting on position.

All the Argus documentation can be obtained via NetScape.

APPENDIX A1. ARGUS/SUN COMMUNICATION AND ARGUS.ID FILES

Information about the Argus fiber assignments is now being included in the Arcon image headers, as is the spectrograph setup information from the instrpars pset.

Instrpars appears as follows:

assigfil=                       Name of Assignment file
coordfil=                       Name of Coordinate file
idfile  =                       Name of Id file
(argusro=                home$) root directory for Argus files

                                # PARAMETERS OF BENCH SPECTROGRAPH

(complam=                 HeAr) Comparison lamp
(grating=                KPGL1) Grating identification
(tilt   =                100.1) Grating tilt (units ????)
(echmode=                   no) Spectrograph configured in echelle mode?
(colfocu=                   99) Collimator focus position (units ????)
(camera =       blueAirSchmidt) Spectrograph camera identification
(camfocu=                   99) Spectrograph camera focus (units ???)
(filter =                 none) Order sorting filter

                                # ADDITIONAL INFORMATION

(instrna=                argus) Instrument name

The value of "argusroot" must be set to the name of directory where the Argus assignment and coordinate files are being kept. The parameter "argusroot" should be something like "home$", "home$coord", etc. depending on how you set up gonfs. When you run an Argus command like "Move" or "Automatic Comparison sequence", the Argus PC will write two files called "argus.hdr" and "argus.id" to the E:\ directory. These files contain the information about what fiber is assigned to what object, and what object/assignment files were called to generate the id file.

Provided this mechanism is working it is not necessary to set the values of assigfile, coordfile and idfile in the pset, they will be set automatically to the values found in the "argus.hdr" file. The parameters can be used to manually set the names of these files if, for whatever reason, Argus is not generating the argus.hdr file.

The parameters of the Argus bench Spectrograph as entered in the pset also get written to the image header which may be useful for documenting your setup.

Here is an example of part of the resulting image header:

    COMMENT INSTRUMENT PARAMETERS
    INSTRUME= 'argus'              / Argus fiber feed & Bench spec.
    ASSIGFIL= 'mm7_1.ass'          / Name of Assignment file
    COORDFIL= 'mm7all.obj'         / Name of Coordinate file
    IDFILE  = 'argus.id'           / Name of Id file
    COMPLAMP= 'Quartz'             / Comparison lamp
    GRATING = 'KPGL1'              / Grating identification
    TILT    = '100.1'              / Grating tilt (units ????)
    ECHMODE = 'no'                 / Spectrograph configured in echelle mode?
    COLFOCUS= '99'                 / Collimator focus position (units ????)
    CAMERA  = 'blueAirSchmidt'     / Spectrograph camera identification
    CAMFOCUS= '99'                 / Spectrograph camera focus (units ???)
    FILTER  = 'none'               / Order sorting filter
    TELFOCUS=                27000 / Telescope focus
    COMMENT FIBER ASSIGNMENT INFORMATION
    FIBER1  = '1   0 23542 16.002 0.257 1.703 0.304 7B-RR' /
    FIBER2  = '2   1 sky'          /
    FIBER3  = '3   0 8835 15.886 0.012 1.332 0.018 7B-Bl' /
    FIBER4  = '4   1 sky'          /
    FIBER5  = '5   0 130921 15.147 0.218 1.217 0.381 7A-RR' /
    FIBER6  = '6   1 sky'          /
    FIBER7  = '7   0 66826 16.937 0.141 1.411 0.202 7A-RR' /
    FIBER8  = '8   1 sky'          /
    FIBER9  = '9   0 34541 16.010 0.187 1.601 0.187 7A-RR' /

		etc

The fiber assignment info in the header can be used to generate the aperture identification table needed by the reduction task imred.argus.doargus as follows:

hselect image_name slfib*  yes | words > idfile

Please note the following feature of the "idfile". You will use the "idfile" to stuff the names of the stars into the IRAF header. (I know this makes little sense. After all, the star names are already there as the SLFIB* keyword. Unfortunately, the SLFIB keywords are not recognized by IRAF, and you have to transfer these names to a keyword which IRAF likes). To insert the correct names into the header, take the output from the

hselect

saper image_name apidtable=idfile

The "idfile" has the format "fiber-number beam-number stuff" where stuff is usually the name of the star. The "stuff" must be enclosed in apostrophes or weird things will happen. This is an undocumented feature of this format. You are now warned. An example of a correct line in the "idfile" is:

28  0 "274 12.97 9.99 -231.8 339.3 99 23"

APPENDIX A2. GONFS

In order to establish the link between the PC and the SUN on either the Argus PC or the simulator PC, you should do the following:

1. If you are in Argus, exit the program with "T". Make sure you do not reset Argus when you leave!
2. Re-boot the computer.
3. When it comes back up, type:

   gonfs [username] [directory]
   

   where [username] is your login name (something like v17), and [directory] is an optional subdirectory in /ua11/v17/.

For instance, suppose you had the files in /ua11/v17/nite1/coord/. You would enter:

gonfs v17 nite1/coord

Note there should be NO final "/". You do not have to point to the actual subdirectory. Had you typed

 gonfs v17 nite1 

you could now access the directory from Argus (and DOS) as E:\coord

Remember that you have linked the SUN with a PC, and the PC is running DOS. You must chose file names that are compatible with DOS on the SUN. We have found that you should chose names for files like abcdef.ghi where the name is less than 7 char and the extension is 3 char or less. Similarly, directories should have less than 7 char. Also, it seems that the name should be written in lower case. Although DOS looks like it uses CAPITAL letters, any capital letters will confuse DOS.

APPENDIX A3. ASTROMETRIC FIELDS

We have a number of fields with astrometric positions resident on the hard disk. These fields were measured by Arnold Klemola at Lick Observatory. To check the astrometric quality of Argus and to check the rotation:

1. Copy one or more astrometric fields from C:\argus\astromet to your disk area E:\. Edit the object files to have stars with a mag range of < ~1.5 mag. Assign stars to fibers. The assignment file must have one star must be in the center. Please do not modify the object or assignment files on disk area C:\ .
2. Move positioners to position. Put the periscope on the central object. Mark the fiber with the digital cursor. Move the telescope to fiber position. Start guiding.
3. In command mode (F3 on the Argus keyboard), type org (for original position).
4. Tweak up the fibers on the stars.
5. In command mode, type corr (for corrected positions).
6. In command mode, type gr (for graph it)
7. See if there is any rotation or organized motion. Enter the new rotation into the Argus program. There are 3 paramters: krot krad kang , which are rotation (krot) and translation(krad and kang). Type param to see the present values of these parameters. To change a value and see its effect type:
   

    
   krot -0.3 
   move 
   gr
   

   etc, changing the parameter until the rms is minimized. To permanently enter the value, run param again.

     SUMMARY OF "ASTROMETRIC STANDARD FIELDS" FOR CTIO ARGUS ALIGNMENT
     *****************************************************************

     ********************************************************************
    N  STANDARD FIELD  RA (B1950) DEC     EPOCH     STARS    DATE DONE
    1  CETUS          00 00 48  -20 51    1986.91   1-250    1993 Dec
    2  LEPUS          05 22 12  -19 09    1988.06   1-148    1991 Nov
    3  PYXIS          08 00 17  -20 11    1987.15   1-235    1993 Feb
    4  CORVUS         12 03 16  -21 04    1987.24   1-170    1993 Feb
    5  SCORPIUS       16 19 56  -19 59    1987.47   1-344    1994 Feb
    6  M75 (N6864)    20 03 08  -22 04    1987.65   1-219    1992 Jul
     ********************************************************************

Notice that the fields are equinox ~1987, so that is is possible (but not likely) that a star of particularly high proper motion (>~ 0.1"/yr) could be off by as much as 1 arcsecond.

APPENDIX A4. CCDINFO INFORMATION

The ccdinfo command in the "IRAF Acquisition" window has all the important information about your CCD setup. The CCDs are regularly tested in La Serena, and the read out maps may have changed from your last run. The ArCon can read out more than one amplifier. In the case of the Loral, there are two amplifiers ("ll" and "lr", or "lower left" and "lower right"). Presently, the "lr" amplifier is the preferred one.

        (gain = 2)              Gain setting 
    (preflash = 0.)             Preflash time (seconds) 
        (xsum = 1)              pixels summed in X direction 
        (ysum = 1)              pixels summed in Y direction 
      (xstart = 1)              Start of ROI in X 
      (ystart = 1)              Start of ROI in Y 
       (xsize = 3072)           Size  of ROI in X 
       (ysize = 1024)           Size  of ROI in Y 
      (extend = "separate")     Type of extension to include overscan 
   (noverscan = 60)             Number of overscan pixels (binned) 
      (xskip1 = 20)             X pixels to skip at start of overscan (binned) 
      (xskip2 = 0)              X pixels to skip at end   of overscan (binned) 
      (xtrim1 = 0)              X pixels to trim at start of data  
      (xtrim2 = 0)              X pixels to trim at end   of data  
      (ytrim1 = 0)              Y pixels to trim at start of data  
      (ytrim2 = 0)              Y pixels to trim at end   of data  
  (amplifiers = "lr")           Readout amplifiers to be used 
     (pixsize = 15.)            Pixel size in microns
    (nxpixels = 3072)           Detector size in X 
    (nypixels = 1024)           Detector size in Y 
     (detinfo = "")             Optional image header info about detector
     (detname = "Loral3K_1")    Detector identification
        (mode = "ql")           
 

			*** Table of gain values ***

      i                       ARCON 3.5  /  Loral 3K
      n                                                             Full CCD
      d DCS  Dly  ___Read_Noise___  ____1/Gain_____  __Read_Noise__ SingleRead
      e (us)           (ADU)             (e-/ADU)         (e-)      Time (s)
      x ---- ---  ----------------  ---------------  -------------- ----------
                      ll    lr          ll    lr        ll    lr
 1:   1   5   1      2.54  1.48        4.33  7.82      11.0  11.6      88.2   
 2:   2  10   3      3.88  2.17        1.99  3.96       7.7   8.6     120.5  
 3:   3  15   2      5.68  2.97        1.39  2.59       7.x   7.7     152.4   
 4:   4  20   1      7.08  3.87        1.03  1.94       7.3   7.5     184.4 
     
              Last update by Ricardo Schmidt, 1995 Jul 12 

	*** Select gain setting from the first column

	*** The current gain setting is 2
 

APPENDIX A5. THERMAL ENVIRONMENT AT THE CTIO 4M

Jack Baldwin and John Filhaber have been carrying on a program to improve the image quality of the 4m telescope. Part of this project is to control the thermal environment of the 4m. The goal is to eliminate or reduce all heat sources (especially hot, sweaty ones like you - that is why you are on the ground floor now!). The 4m mirror is being actively cooled every day to a temperature predicted to be tonight's nighttime temperature. The cooling is under control of a PC, and we are still experimenting on what the best algorithm is to predict tonight's temperature. The observing floor and oil pads are also being cooled.

You can monitor the temperature profiles yourself with some tools provided by Jack and John, although you cannot alter the cooling program. Two basic ASCII output programs give general temperature information. The first is weat which queries the CTIO weather station. This station is on top of the old tower to the NE of the 40" telescope. This weather station is not extremely accurate - the humidity value is particularly poor. The other program is temp4m. This gives the 4m thermal environment. Here is a typical nighttime output.

cl> !temp4m
DATE: 12/30/1995 TIME: 02:20:21 OUTSIDE: 15.3 SER TRUSS: 16.6
DOME: Hi: 15.5 Lo: 15.6 Outside Cage: 16.0 Floor: 15.9
TEL: Primary N: 15.8 S: 15.6 E: 15.9 W: 5.1 Cass Cage: 16.9
OIL: Exiting Cooler: 16.3 Before Pad 12.6 After Pad 14.2
     Glycol: -1.4  Valve Voltage: 2.25 Pump On/Off: 0.00

The sensor locations are self-explanatory. At night, the Hi dome, Lo dome, SER TRUSS, Outside Cage, and Floor should be very similar (within 1 degree) if the inside of the dome is at ambient ( OUTSIDE). The mirror should be within 1 degree of the ambient also, preferably slightly on the cold side. There are 4 values of the mirror temperature - the W value is rather flakey and should be ignored. The oil temp should be close to ambient during the night also.

The OUTSIDE temperature is not correct in the afternoon because the sun hits too close to the sensor. This sensor is on a long pole that sticks out towards the 60".

You can also get plots if you are run the following programs from a xgterm (NOT an xterm):
/u4m0/seeing/temper/dometemp or
/u4m0/seeing/temper/tlog4m.

When you enter the program, type "?" to get a list of the commands. The dometemp program is particularly easy to control. Try plotting outside, dome low, and mir south (or east, or north) to get a good feel for the temperature environment.

Jack has also installed an image motion monitor on the 4m from a design by George Jacoby at KPNO. This program measures the rms image motion from video frames of a star. Your night assistant will make one or two measurements per night. The absolute value of the image quality from this monitor is on a rather dubious scale (but probably pretty close to the real value) , but the relative values are important.

We also have real seeing measurements from the seeing tower. These can be accessed from "Current Logs from Site Seeing".