Discussion

• This test was made on completion of an optimization effort which resulted in the best performance yet seen for this detector. Unfortunately, however, no baseline test data set was obtained before the optimization so exact details of the change of performance are not known. Table 1 gives an approximate list:

    
  Table 1: Loral 3 before and after optimization.
  

• In the course of the optimization, the operating temperature of the CCD was raised from to . This is likely to have increased the device's QE in the red, albeit only slightly. Since we have no accurate QE curve for this device and no way of obtaining one at present, caution should be applied in comparing data taken before and after the optimization.
• The interference pattern in the bias images appears to originate from intereference with the readout by the Utility board in the Leach controller. The coherence of the signal during a single readout was obtained, some time ago, by synchronising the Utility board clock with the readout. Careful grounding schemes applied at the telescope appear to have resulted in significant suppression of the signal. This will be reported on elsewhere.
• The bias level appears to be stable to with ADU within a readout. Under most circumstances, this does not present a problem.
• The presence of a glitch at the start of bias overscan and the exponential rise to the true bias level in the full frame biases seems to imply that the overscan is not obtained in the same manner as the bias from the imaging area. The timing control software does indeed include a pause at the transition but it should not result in the observed effects. Investigation will continue at a low priority level as the effect is small.
• Although the readout noise measured in this test is the lowest yet known for this CCD, the presence of the interference signal and the apparent lack of intereference at the summit implies that the true readout noise is still lower.
• An incomplete data set was collected. Quantities not measured with this test set include: dark current, CTE, analogue saturation level, trap quantity, shutter delay, shutter delay map, ADC biases, QE. For a description of how a complete data set may be collected and reduced, see the document ``In situ CCD Testing''