Cambridge AIT "Observing" Proposal

Introduction.
	The VIRCAM AIT presents an opportunity to acquire data useful
beyond the scope of the assembly, integration and testing requirements.
By obtaining a large number of frames with the science templates, in
conjunction with a running TCS simulator, a body of extremely realistic
data will be obtained for exercising all the pipeline modules. Further
specific observations using spot-projection facilities will be 
invaluable in determining, or placing limits on, cross talk and persistance
characteristics of the detectors in their full focal-plane configuration
and with near-point sources.
	Flats will not be so flat and point sources not so sharp as
flats and images on the real telescope; however, the calibration products
from these data runs would make a first, `quick and dirty', set of 
calibration data with which to bootstrap the data-reduction pipeline following
first light.

1. Reset test:
   Simply use img_cal_reset to take around 30 frames. Repeat this experiment
   several hours later to check on the stability of the reset with time.

2. Dark and reset anomaly test:
   Take 5x 1s, 2s, 5s, 10s and 20s darks for NDIT = 1,2,3. Look at the
   structure and variation of reset anomaly as a function of exposure
   parameters. Also look for second order artefacts like curtaining as 
   found in the WFCAM chips. Look for hot spots that grow with time.
   Repeat this experiment several hours later to check on the stability of 
   the reset anomaly with time.

3. Darkcurrent test:
   use img_cal_darkcurrent to take frames of 1, 2, 4...256s

4. Domeflat test:
   use img_acq_domescreen (to get TCS simulator in correct state),
   use img_cal_domeflat first to adjust exposure time/lighting level to
   mid-range of ADUs, then take around 30 flats.

5. Noise and gain:
   Use img_cal_noisgain several times with appropriate lighting/exposure
   levels.

6. Linearity and bad pixel test:
   Adjust lighting level so ~20s give not-quite-saturated images,
   use img_cal_linearity to take 20-flat sequence. (current template
   does not take corresponding darks - so use img_cal_darkcurrent to
   take matching dark frames). Repeat this experiment several hours later to
   check on the stability of the linearity solution and the bad pixels with
   time.

7. Persistence Test:
   Adjust spot level to give well-exposed, unsaturated spot. Take
   one full regular readout using img_obs_exp, insert dark filter, then follow
   with uncorrelated readouts as fast as possible for about one minute. Repeat
   the experiment with a series of reset-read readouts to look at the effect
   of resets on the decay of the persistence image.

7. Cross talk test:
   Adjust spot to roughly centre on channel 1 of a particular chip.
   Take test frames to adjust exposure to obtain 2-3 times over-saturated
   image. Use img_obs_exp template to readout frame, repeat for all 16
   channels on that chip.  Possibly repeat on all chips. (When moving to a 
   new channel, it may be necessary to shift to a different part of the 
   channel then for the previous exposure if persistence is in fact a 
   serious problem).
   [Can't use img_cal_crosstalk because it needs a real telescope].

8. Twilight Flats:
   Exercise system by using img_acq_twilight/img_cal_twilight to take
   3 "twilights".
 
9. Imaging simulation:
   Again adjust illumination/exposure time to give "good sky", use
   img_obs_acq to "acquire" an "object" and take a pawprint using
   img_obs_paw and a tile using img_obs_tile.  Possibly repeat.