IDS AUTOMATIC QUALITY CONTROL V0.2 ---------------------------------- DLP (09/02/1997) IDS Configuration ----------------- The Quality Control measurements for the IDS have now been automated. To the Support Astronomer this means that very little effort and interaction is required to produce reliable QC. While for the instrument specialist a consistant set of historical measurements for the IDS will be produced. This will become increasingly important now the new Data Aquisition system is the default system as a much greater choice of CCD's are available. For these reasons it is requested that during the afternoon before a scheduled run and on every service/discretionary night the Support Astronomer attempts these measurements. The results will also indicate to the Support Astronomer if the instrument is performing at an acceptable level. In order to obtain meaningful results we have been forced to restrict the IDS configuration to the following setup which MUST be adhered to: 1) the chip corrected for tilt and rotation and the spectrograph focus set 2) R1200Y grating 3) grating exercised and set at a central wavelength of 6563A 4) comparison mirror in and neon lamp on 5) slit width 220 microns 6) chip window: (i) TeK 1K chips: x 1-1120, y 400-656 (ii) Loral 2K chips: x 1-2140, y 900-1156 (iii) EEV 1K chips: x 1-1214, y 400-656 7) standard readout speed 8) 1 second integration 9) any camera/collimator/chip combination ANY SETUP DIFFERING MARKEDLY FROM THE ABOVE WILL FORCE THE SOFTWARE TO REJECT THE QC MEASUREMENTS. Lets face it, none of these requirements are particulary stringent so there is no excuse for getting it wrong. How do you run it ----------------- With the IDS setup as above take a 1 sec integration. Because the new DAS is running in a solaris environment and as such cannot directly communicate with the filesystem on the user cluster, it is necessary to ftp the fits file to your scratch space on the user cluster. With the new integrated ICS + DAS observing system (Sparc based) simply type (on your xwindows terminal) after transfering the file: ~quality/qc/src/ids/ids_qc With the Perkin-Elmer ICS (adam based) and DAS use ~quality/qc/src/ids/old_ids_qc The difference due to the variation in header keywords between the two systems. After starting up a couple of STARLINK packages running, the script will prompt you for the fits file name without the extension, so if the file is called r8787.fit enter `r8787' and the script will be happy. What happens next depends on if the camera/chip combination has been used before. If there have been previous runs with this combination you will see a series of graphs on the terminal: 1) the first graph is a plot of the 1d extraction of your arc alongside a `master arc' for a similar IDS configuration. On the master arc panel, two arc lines will be marked: these are the lines the software requires for the QC measurement. The bottom graph shows the extracted 1d spectrum of the image you have just taken. The line identification routine will endevour to identify and mark the same two lines as used in the master arc. However, the algorithm is pretty crude and if for some reason the central wavelength is off it may well identify the wrong lines. In the xwindow that you started the script running in, it will ask you to confirm if these are the right lines. With reference to the master arc spectrum, if either one of these lines is misidentified you should reply with a `n' and a cursor will appear. Click (left mouse button) on the line at lowest pixel number first and then click on the second line. 2) the second set of graphs is a representation of the IDS Quality Control. Four sets of graphs are shown showing: a) historical line flux (the measurements for the average of the two lines) - IDS through put. The different coloured measurements correspond to the different collimators (ie red - Ag red, green - Al wide and blue Al uv) b) historical line position - accuracy in wavelength setting. c) historical line FWHM and also for all unblended lines - focus setup. d) historical pixel difference in line centres for the two lines - a measure of the sprectral stretching. 3) hitting return in your xsession window will then produce a third set of graphs which are the CCD quality control. The two graphs show: a) CCD read out noise b) CCD bias level In all cases the current observation is indicated by an enlarged star. If this camera/chip combination has not been run before several outcomes are possible: 1) If this is the first run with a new chip of known generic variety (known to the Quality Control account that is), eg first run of Tek1 while another Tek chip is known, then the program will still go through the line identification part as descibed above, and inform the user that there is no historical measurements for this chip (ie no graphs). 2) If this is the first run through with an unknown chip of unknown generic type, eg first measurement with any Tek chip the program will crash. IN THIS CASE YOU SHOULD KEEP the 2d FITS FILE and inform dlp. This problem is likely to only occur in the next few months while the history files are being created (I will endevour to be at all QC runs when a new kind of generic chip is introduced to the instrument so this problem can be avoided). This is the end of the IDS (+ CCD) Quality Control. The whole process from starting the integration to finishing the QC script should take less than 5 minutes and will be much faster when the user cluster moves to solaris (the whole process can then be automated). What does the IDS QC software do -------------------------------- When the script is started it interacts with various files in the QC account writing history records for both the camera/chip combination and also a seperate CCD history file. These are stored in ~quality/qc/dat/ids and will be called something similar to tput_235.TEK1. The CCD QC file is stored in the same directory and is called something like ccd_qc.TEK1. You are welcome to inspect these files at any time but please be careful as they are completely unprotected - SO DONT DELETE THEM!