Light from a single star is often used as a flux standard. In multiple fiber spectroscopy it is desirable to collect light from this single object through most or all of the individual fibers. It is possible to move either the telescope or the fiber probes so that the calibration star is placed in one after the other of the fiber entrance apertures. Usually these are quite short exposures.
For this reason, we have implemented a sequential move command in the Enterprise. This command, called by `` A( '', moves each probe in order to its current buffer position, waits there for a predetermined dwell time, and then retracts the probe to neutral while moving the next probe into place. The operational procedure follows:
Light from a single bright galaxy is often used as a velocity standard. In multiple fiber spectroscopy it is desirable to collect light from this single object through several of the individual fibers. Typically we might put the galaxy nucleus in the center of the field (guide star position) and put several probes on it during the course of a CCD exposure. A typical operational procedure follows (see Single Star Flux Calibration if you have short exposures):
If your exposure in each fiber is more than a few minutes long, you may want to retract the probe momentarily to check on the telescope tracking.
Probe 01 in the MX focal plane has a tiny knife-edge screwed to the tip. By defocussing the acquisition television with a lens that lets you view the pupil, this knife-edge may be used to focus the telescope.
This section describes the procedure for marking the position of the MX field center on the acquisition TV screen so that you may guide the center star on this mark. Yes, there probably is a better way, but either we haven't thought of it yet, or we can't afford it. Position should be stable for the duration of the run, but its good to check now and then.
Since the neither the MX bolt pattern nor the IIS rotation pointer is precise at the sub-arcsecond level, it is important to set the absolute instrument rotation each time the MX is mounted on the telescope. You will also want to check the rotation if you have coordinates which are not accurate with respect to the absolute reference frame of the sky.
The IIS (instrument rotator) ring should be clamped with the index pointer on the chisel mark. The adjustment and the pointer are on the west side. The C-clamp location is on the east side. Once the IIS is clamped, you can make the final rotation alignment with the fudge factors in the Enterprise software. The key parameter is the variable FANG which is the absolute field rotation in microradians. Typically, the range of adjustment from the chisel mark is 0 to 2000 microradians. In typical MX style, FANG is stored as an I*4 variable. The memory location of FANG is at 8000. To check the current setting, say `` D 8000 '' while talking to the Enterprise. The first line of the response will read 8000 4C 04 00 00 00 00 00 00 00 00 00 ..... Naturally the values are displayed as hexadecimal bytes in reverse order. The value of FANG here is 0000044C or 1100. See the sample listing below. The typical range of rotation angles is from 00000000 to 00001800.
Optimum rotation is defined as maximum signal from a cluster of alignment stars. Be careful to avoid a single bright star which dominates all the others and gives one probe undue influence on the signal. Always be sure to guide carefully on the center star while checking the rotation.
To change the rotation value from the command line use the task mxobserve.radjust and type `` radjust 0000044C '' or whatever angle you wish to change to.
To change the rotation value directly in Enterprise type `` R0000044C '' or whatever angle you wish to change to. The `` R '' command automatically reconverts the coordinates in buffer D000 into the current buffer. After you haved changed the rotation angle, `` A@ '' will move the probes to the rotated position. This allows you to quickly iterate the PMT counts because the probes only take a few seconds to move.
To change the platescale value from the command line use the task mxobserve.sadjust and type `` sadjust 000001F4 '' or whatever scale factor you wish to change to. This updates the variable DSCL at location which modifies the platescale by parts per million. The platescale rarely needs to be adjusted.
<01:1>D8000 8000 4C 04 00 00 00 00 00 00 00 00 00 00 00 00 00 00 L............... 8010 F4 01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 8020 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 8030 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
Once you've gone this far, we can tell you where other important variables are stored. Details are given in the source code files ESUBS.FOR, EPROG3.MAC and EPCOM3.MAC.
This section describes the procedure for calibrating the coordinates of each of the 32 MX probes at the beginning of an observing run. This normally occurs in the afternoon after the instrument has been mounted. Only J. Hill, the instrument specialist, or the most experienced observers should ever need to do this. Make sure the MX support struts were installed and adjusted correctly before you start.
or if using IC2400 terminal:
or if using em4010 on a PC:
or if using X-terminal:
(port = "/dev/ttya") serial port for enterprise
(tfile = "tmp$ttarget.txt") stream file
(necho = 1) echo interval
(verb = 4) verbosity
(version = "03 June 1992") Version date of this routine
(mode = "ql")
(cfile = ) Name of the center calibration file
(apndset = yes) Append to output file
(clobset = no) Clobber output file
(stream = "tmp$ttarget.txt") Name of the enterprise stream file
(portname = "/dev/ttya") Name of the serial port
(instrument = "mxspect") Instrument pset
(verbosity = 4) Output verbosity parameter
(mode = "ql")
What you need for probe alignment is:
^[[D ^[[C ^[[B ^[[A
Recently we have been using the Exceed xwindows program running on ``bokpc1'' connected via a long 10baseT network cable. The xterminal ``ncdxtb22'' can also be used on the same long network cable. Ask the instrument specialist if you are uncertain about moving the auxiliary PC or the xterminal out to the platform. The 40-foot 10baseT cable lives behind the partition at the South end of the control room.
We replaced the bad fibers in October 1997, so this procedure is not needed.
Several of the probes may have broken or damaged object fibers. The target selection programs are able to make use of these probes by assigning targets to the sky fibers. Unfortunately, the center alignment procedure does not deal well with aligning the sky fibers onto the field center. The following procedure allows a manual workaround to this shortcoming. When the center file has a fiber number other than 01, ``loadcenters'' corrects the offset. It reduces the motor 1 coordinates by 016F for each probe where you centered on fiber 05 instead of fiber 01. The scale factor is the fiber index times 1016 microns times 0.090229 steps per micron.
The two historically broken fibers in position 01 are on probes 03 and 12. As of July 96, these two fibers are corrected in the manner described above if you do ``makeitso circle=yes fix=yes''.