PROCEDURE: 1h

TITLE: Basic Monitor (HRC)

CATEGORY: Monitor

PURPOSE: Measure basic performance parameters of CCDs.  These are - dark
count rate, read out noise, bias level, gain, and cosmetic defect
position and fraction.  

FREQUENCY: Run after every major change, e.g. temp, gain
setting.

DETECTOR: HRC

PREREQUISITES: None

HARDWARE REQUIREMENTS: none

SOFTWARE REQUIREMENTS: standard stats analysis tool
                       ACSGAIN.PRO
                       stsdas.hst_calib.ctools.msbadpix (or IDL equiv.)

COMMAND MODE: SMS (done)

PROCEDURE/EXPOSURES:  
        Tungsten lamp (#3) on
        "standard" filter selected :  SDSS r (F625W)
          1.   bias                        0 s
          2.   dark                     1375 s
          3.   flat                      1.0 s  
          4.   flat                      1.0 s  
          5.   flat                      2.0 s  
          6.   flat                      2.0 s  
          7.   bias                        0 s
          8.   dark                     1375 s

VARIANTS:       a- gain = 1     JTVH1A_1.SMS
                b- gain = 2     JTVH1B_1.SMS
                c- gain = 4     JTVH1C_1.SMS
                d- gain = 8     JTVH1D_1.SMS
                e- gain = 1, two more biasses - one after exp#1, exp#7
                             (for Zlatko)

TOTAL EXPOSURES: 8

TOTAL TIME: setup + 8*overhead  + 2706s = 1.19 hr
        (setup = 10min, overhead = 2min)

ANALYSIS:

(1) Run statistical analysis tool on all images (calculates mean & rms
in boxes).  For the biases this gives a first estimate of readnoise and
bias level.  For dark this gives estimate of dark-rate.

(2) Subtract mean overscan or bias level (or master bias frame made by
combining lots of biasses, if one is handy) from all frames..

(3) Use all 8 images as input to noise model program.  This will
calculate readnoise and gain from image pairs taken at the same exposure
levels by modelling the light transfer curve.  

(4) Combine multiple exposures to remove CRs.  Gives one of each bias,
dark, 1s flat, 2s flat.

(5) Find bad pixels as those that deviate by (say) 5sigma in either
    a. combined bias (mostly hot pixels)
    b. combined bias subtracted darks (mostly hot pixels)
    c. ratio of combined (bias subtracted) flats 1s/2s 
       (other pixels that don't FF).

(5) Compare derived parameters with previously determined values.
Compare bias and darks before and after flats to check for flouresence,
delayed charge.  Plot results as a function of date, T_CCD, T_env.  Note
trends, ring alarm bells for serious discrepencies.