Introduction

This checkout procedure describes the minimum tests required to validate that the CCD system is working properly before it is released to observers.

It is written for a generic AzCam system.

The specific system performance parameters which should be noted and compared to results obtained during checkout can be found at http://www.itl.arizona.edu/systems/systems.htm.

Initial setup

It is assumed here that the instrument has been mounted.

Check for dewar isolation from telescope. This should be a high impedance (15 MegaOhms or greater). If a low impedance is found, fix the problem before turning on CCD power.

Turn on CCD system power.

An instrument server (such as optoserver), an image sever (such as dataserver), and a telescope server may need to be running (such as telcom server) on their appropriate computers.

Start up the AzCamServer computer (such as bokccd).  Start AzCamServer and then AzCamTool.

The following exposures should be taken using non-test images with a root name ccdtest (test image unchecked on AzCamTool front panel).  Press Filename and enter this name and select an existing folder to store the images.  It is assumed the images are being sent to the observing account (such as bokobs) running IRAF on another machine (such as stjukshon).

Image readback

The data path and zero level of the CCD can be checked by taking a zero exposure. Do not turn on any lamps and make sure the dome is dark. You will have to take two exposures after power up to make sure the CCD is cleared of residual charge and all software is loaded. If there are instrument and/or telescope communication problems, you may have to uncheck Enable Instrument or Enable Telescope in Preferences to disable communication to those systems.

Zero image check

The zero exposure should have no structure or gradient and a read noise of under 10 electrons rms. The read noise will slowly decrease as the system warms up and stabilizes.  The noise can be found by typing imstat ccdtest.*[100:150,100:150]. The standard deviation (read noise) should be less than ten (10) divided by the CCD gain.  The gradient can be displayed by typing implot ccdtest.* and looking at a line plot (l) and a column plot (c). The line should be fairly flat (not more than 3 to 5 DN in a column plot). You will need to average some rows together to beat the noise. 50 is a good number (i.e. :l 400 450 in the plot window).

Digital data check

The CCD should also be checked for missing and/or stuck bits with imhist ccdtest.* on a zero frame. You should see a fairly smooth gaussian curve. You may have to zoom in on the area of interest. This is done with imhist ccdtest.* z1=xx z2=yy. The numbers xx and yy bracket the values of interest, and should be no more than 100 apart. Be sure that nbins= 65535 (in epar imhist). This insures that each data value goes in a separate bin.

You should make sure the CCD temperature is OK at this time.  The temperature should not vary by more than one degree once the stabilized temperature has been reached. It may take up to one hour after power-up for the final temperature to be achieved.

Light response

The CCD response to light and shutter operation can be found by taking two flat field images of different exposure times. The length of the exposures will depend on the instrument used, but make sure at least 1000 DN of light illuminates the first image. After finding the first exposure time, take another with twice that exposure time. The mean level measured with the imstat ccdtest.*[100:150,100:150] should be twice as large in the second image (after subtracting the zero level from each) as in the first. If this is true then the shutter is working properly, as is the CCD's response to light. If you are on the spectrograph, you will get spectra and this can be a continuum (no spectral lines) or, if you turn on a spectral source (HeAr for example), you will get lines.

Dark current and charge transfer

Take a 5 minute dark exposure to make sure there is no noticeable dark current or other time related problems. The image should look like the zero level exposure taken earlier. Do the same tests as on the zeroes. The CCDs have no measurable dark current in this short time, although you should be careful to make sure no light falls on the CCD from the dome or flashlights. You may notice some cosmic ray events in this exposure - they should be approximately point sources a few pixels in diameter. Make sure they do not all have a charge transfer tail pointing in the same direction.

Cleaning up the account

Delete all images just taken with "imdelete ccdtest.*".

Leave the system power on in order allow the electronics to stabilize if it is safe to do so.

Logging setup information

Send an email message to the appropriate CCD system log stating your name, the instrument mounted, the CCD read noise and zero level, and any comments you think are worthwhile.