There are many inbuilt commands for the MCCD program which controls the APT. If you aren't sure what arguments a command takes, type the command on the PC and it will prompt you; to avoid having to complete a command, give it a crazy answer (e.g., ``time garbage'', will abort the ``time'' command); CNTRL-C will crash out of the program, so you don't usually want to type it. To get more information on the MCCD commands use the ``help'' command on the PC. Further descriptions of the commands are available using ``man command-name'' where ``command-name'' is the name of the command that you want to obtain information on. Not all commands have on-line documentation at this time.
Only a small subset of the available commands are normally used for observations, for example:
CALIBRATING THE TELESCOPE calra Calibrate the RA telescope axis. caldec Calibrate the Dec axes. calaxes Calibrates both axes simultaneously. caltel Accept next RA and Dec as the current values. MOVING THE TELESCOPE coord hh mm ss dd mm ss Set RA and dec for next object, but doesn't slew. slew Slew to next RA and Dec, and waits till tracking. slewnowait Slews, but doesn't wait for the telescope to track. wait tel Waits until the telescope is tracking. wait exp Waits until the exposure is finished. centre x y Centre the telescope on given pixel coordinates. centre peak Centre the telescope on the brightest star. offset radec x y Shift the telescope by x arcsec in RA and y in Dec. offset xy x y Shift the telescope by x pixels in RA and y in Dec. move x1 y1 x2 y2 Moves the star at (x1, y1) to (x2, y2). move peak x y Moves the brightest pixel (from `stats') to (x, y). INFORMATION FOR FITS HEADERS observer Sets the observer's name. object Allows you to enter the name of the object being observed. filter Sets the filter. imagetype Sets the image type (e.g., dark, flat, object). epoch Sets the epoch of the coordinates. equinox Sets the equinox of the coordinates. timezone Sets the timezone (+10 in winter, +11 in summer). CONTROLLING THE CCD CAMERA clear Clears (erases) the CCD. exposure Begin exposure (shutter open). Doesn't clear the CCD. abort Aborts an exposure in progress. dark Begin a dark-frame exposure (shutter closed). readout Readout an exposure and send to the workstation. stats Readout an exposure and calculate statistics. readstats Readout an exposure, do stats and send to workstation. time s Set exposure time in seconds. scale peak c Sets the exposure time so that the brightest object in the last stats area has `c' counts. scale average c Sets the exposure time so that the average value in the last stats area now has `c' counts. sra x1 y1 x2 y2 Set readout area of the CCD. dra dx dy ddx ddy Makes a delta-change to the readout area of the CCD. ssa x1 y1 x2 y2 Set the statistics area of the CCD. ssr Set the statistics area to equal the readout area. FOCUSSING focus reset Resets the focus readout to zero. focus up x Moves the CCD camera up by 'x' steps. focus down x Moves the CCD camera down by 'x' steps. focus delay x Changes the inter-step delay to 'x' microseconds. focus set x Sets the focus to position 'x' on the readout (experimental). MISCELLANEOUS COMMANDS rf Allows you to execute a script of observing commands. echo string Writes string on the PC screen; useful in scripts. shm Show the defined macros. quiet 1 Makes MCCD less verbose. dlym ms Delay by ms milliseconds. dlys s Delay by s seconds. stm s Sleep-to-modulo. Delay until the time is zero mod s. beep Make a beep sound. Useful to indicate end of exposure. settime Sets the PC time. Normally not used. setappletime Sets the Apple time from the PC time. Normally not used. readapple Can read any location in the Apple. Normally not used. writeapple Can write any location in the Apple. Normally not used. runapple Runs a machine code subroutine in the Apple. Don't use! lrc Loads RA servo controller coefficient table. Don't use! help Gives a cryptic list of all the MCCD commands. q Quits the MCCD program.
auslast
will display the most recent Australian weather map on the workstation. It does this by making ``ftp'' contact with a machine in Queensland, and retrieving the latest GIF-format satellite image of Australia. The image is stored in /tmp/ausvlast.gif, and then displayed using ``xv''. To redisplay the image without ftping it again, just do ``xv /tmp/ausvlast.gif''
The image orientation depends on the arguments given to SAOimage: if you specify ``-ul -rot 1'' then the orientation follows the usual astronomical convention (North up, east to the left); without these arguments, North is to the left, and East is up. The image scale is 9.4 arcsecs per pixel. Note that there is a bug in SAOimage which cause images of certain dimensions (e.g., 300 by 200) to be scrambled if you use the ``ul -rot 1'' option.
Click ``etc'' and then ``new'' in the SAOimage window, and then use the delete key to delete the existing image name, and then type another one in. Note that the switch ``-fits'' must be present (if not, you may still get an image, but the data values will be garbage (but sometimes only subtly so)).
The image that is currently reading out can be displayed about 1 second after the last byte of data is received (the elapsing of this time delay is indicated in the ``APT link'' window by a ``+'' appearing before the number of bytes).
coord 12 32 34.5 -23 12 22 slew
will move the telescope to RA 12 32 34.5 , Dec -23 12 22 . Note that coordinates are apparent coordinates at the current epoch. In practice, the scale of the APT is such that you can use mean coordinates from 1950 to 2000 without your object falling off the full CCD frame. The ``epoch'' and ``equinox'' commands only affect what is written in the FITS headers. Precession will be included in a future software update.
The ``slew'' command won't return control to the PC until the slew has completed and the telescope has resumed tracking. If you don't like this (e.g., if you would like to readout the CCD while slewing to the next object), use ``slewnowait'' instead--you can use ``wait tel'' before beginning an exposure to delay until tracking has resumed. The PC doesn't always do a good job of deciding when tracking has resumed.
Measure the pixel coordinates of the object on the SAOimage display, and then use ``centre'', e.g.,
centre 223 777
will place the object that is currently at coordinates (223,777) into the centre of the image. Note that the coordinates are relative to the readout area that you have chosen with ``sra''.
An alternative method of centering an object is to use,
centre peak
or, for example,
move peak 223 777
which will move the brightest pixel in the previous image for which statistics was done to the centre of the image, or to (223,777), respectively. Here is an example of how you might use this to move the brightest object in a portion of the CCD to location (450,450).
sra 400 400 500 500 ssr ! set the readout and statistics areas speed hi time 1 exp stats speed lo ! take an exposure and get statistics centre peak ! move the peak to the middle of the area
move 100 100 244 199
will move the object at position (100,100) to the position (244,199).
``move'' is not accurate for large displacements (say a few hundred pixels), or if it has been a long time (say 30 minutes) since the last slew. The solution is to do a second ``move'' after examining the new coordinates of the star.
Alternatively you can use the ``offset radec x y'' or ``offset xy x y'' commands to shift in arcseconds or pixels, respectively. For example, the above ``move'' command could have been given as:
offset xy 144 99
If the ``calaxes'' procedure described earlier is insufficient, go to a bright star, centre it on the CCD (at any desired reference position), and then type ``coord hh mm ss dd mm ss caltel'', where hh mm ss dd mm ss is the RA and Dec of the star.
Here is an example of commands that you might use to streamline this procedure:
star = coord 12 34 56 -31 12 12 ! a macro with the star's coordinates sra 1 1 800 1200 ! set a nice big readout area star slew speed hi time 1 exp readout ! slew to the star and take an image centre peak star caltel speed lo ! centre the star and calibrate on it
Note the use of the ``star'' macro to avoid having to enter the star's coordinates twice. Also, the CCD speed is increased for this observation to save time. The exposure time will have to be reset before continuing with the observing program.
With creative use of macros, the amount of typing in the above sequence can be reduced. For example, by defining
standard = star slew sra 1 1 800 1200 speed hi time 1 exp readout speed lo calibrate = star caltel
you could then type
star = coord 12 34 56 -31 12 12 standard centre peak calibrate
to get the same effect as the original sequence.
After the telescope has been calibrated, you can point to the zenith with:
coord hh mm ss -31 16 30 slew
where hh mm ss is the current sidereal time.
speed lo time 10 sra 1 1 800 1200
clear expose readout
will take a 10 second exposure, reading out somewhat more than the full CCD area so that you get a good sample of the prescan and overscan regions. If you are doing a trial exposure to check centering, then changing to ``speed hi'' will save 28 seconds. Don't forget to change back to ``speed lo'' for greatest accuracy.
10(clear expose readout beep)
will take 10 exposures, beeping after each one finishes. Note: if you type the above instructions from the UNIX shell using the ``apt'' command, you must enclose the instructions in single quotes so that the parentheses aren't interpretted by the shell; so you would type:
apt '10(clear expose readout beep)'
The CCD has a preamplifier with two gains (set by ``gain lo'' and ``gain hi''). The difference in gain between the two settings is about 2.5. ``gain lo'' is useful to record the full well-depth of the CCD--saturation occurs at about 40,000 counts; its disadvantage is that the noise is not as well sampled (which is probably unimportant for most APT observations, since sky-noise will dominate), and that the full 65,535 count range is not available. ``gain hi'' is the usual setting if you don't need to use the full well-depth; it may also be more linear.
The APT CCD flat-fields very nicely. It is possible to use the one set of flat-fields for a week or more. These statements apply when the camera is left on continuously--there will presumably be differences in the response as the CCD cools down, and as the CCD temperature varies (due, for example, to changes in the ambient temperature affecting the Peltier cooler).
The wide field of the APT presents difficulties when trying to obtain twilight flat fields: it is hard to find a patch of sky that is uniform in brightness over degrees. The zenith is probably a good choice.
The RA tracking should be turned off (by switching off the RA servo amplifier) while taking flat fields, so that stars are not always in the same location on the image, and can be removed by median filtering.
The ``stats'' and ``scale'' commands are useful in calculating the exposure time for flat fields. ``stats'' reads the current image from the CCD and calculates the minimum, maximum, and average pixel values. ``scale average c'' (where ``c'' is a number from 1 to 65535) sets the exposure time for the next image in order to obtain an average of ``c'' counts (as estimated on the basis of the most recent ``stats'' or ``readstats'' command).
So, the following command sequence can be used during the evening twilight to decide when the exposure time is within the right range to make a reasonable flat field (times shorter than 3 seconds are affected by the finite opening/closing time of the shutter (5 milliseconds each way), times longer than 40 seconds are affected by stars):
sra 400 400 500 500 time 0.1 100(clear expose stats scale average 20000 dlys 60)
This will take one frame every minute or so, and will try to set the exposure time of the next frame to the value that will give 20000 counts per pixel. For some time after sunset the sky will be too bright, and the exposure times will be less than one second. At morning twilight it would be best to replace the number ``0.1'' in the above script with ``10'', the times will start by being too long, and will gradually decrease to under 40 seconds or so. There is a window of 11 minutes during which twilight flat fields can be obtained. It is possible to get 11 images during that time.
After the exposure time has reached the correct range, the following command sequence can be used to automatically obtain up to 20 full-frame flat field images (the first line sets various keywords for the FITS header).
imagetyp flat object twilight filter V go = speed hi sra 400 400 450 450 clear expose stats scale average 20000 go2 go2 = speed lo sra 1 1 800 1200 clear expose readout 20(go)
Note the use of a small readout area and hi-speed readout to gather statistics quickly. The loop should be terminated (by typing a carriage-return on the PC) after the exposure time has moved out of the range 3 to 40 seconds.
The dark current may be non-linear with exposure time, although this could be due to not correcting properly for the finite readout time. It is of the order of 1 electron per second per pixel, and is usually negligible with respect to sky noise.
A focus change of 25 makes only a barely noticeable difference.
The focus display is reset to zero whenever you start MCCD, or when you use the ``focus reset'' command.
The software commands ``focus up'' and ``focus down'' move the CCD camera up (outside focus) and down (inside focus) respectively. The numbers on the focus display decrease when the CCD camera is moved up, and increase when the camera is moved down. This is the opposite to what you might expect.
The field is not completely flat. There is a small difference (perhaps 25 to 50 steps) between the best focus at the centre and the best focus at the corners. For best average focus you should choose an area of the CCD between the centre and the lower right corner as viewed with SAOimage (without the ``-ul -rot 1'' option; upper left if you use this option).
Reminder: B filter focus is 60 steps up from V filter, i.e., if V is focussed at 0000, then B is focussed at 9940.
If you type ``focus reset'', then the focus display will be reset to zero, and subsequent focus commands will update an internal software counter which is written to the FITS headers. Currently, the software counter is in the opposite sense to the focus display (i.e., when it goes up, the display goes down).
The focus motor has 200 steps per revolution, and is connected to the focus micrometer by a 2:1 reduction toothed belt, giving 400 steps per revolution of the micrometer, which moves 0.5mm per turn. Therefore one step of the focus motor equals 1.25m movement of the CCD.
Note that due to a lack of torque in the focus motor, the belt driving the micrometer has had to be slackened off, which has introduced some backlash. So, for precision work it is best to approach a focus number from the same direction. It is also possible that the stepper motor will lose steps due to the torque problem.
Not automatic yet. The ``filter'' command allows you to specify the filter being used, so that the FITS headers have this information.
Some data have been obtained for determining the linearity of the CCD. The CCD is non-linear at up to the 5% level, although it appears to be accurately correctable. Michael Ashley has an algorithm available that uses IRAF's ``imexpr'' command to correct for the non-linearity when using the low preamp gain.
Saturation occurs at a data value of about 40,000 when using ``gain lo''. With ``gain hi'' the ADC limit of 65,535 is reached before saturation occurs.