Chapter 8. The CCD Cameras

8.1 The CCD Chip

Each of the LASCO telescopes uses a front-side illuminated, 1024x1024 pixel CCD manufactured by Tektronix for recording images. The imaging area is a 21.5 mm square, and each pixel is a square, 0.021 mm per side. The device is operated in the Multi-Phase-Pinned (MPP) mode. While the MPP mode reduces the full well, it keeps thermally generated noise (dark current) to a minimum. The noise immunity of the MPP implant also helps to avoid the effects of noise generated by energetic particle radiation. The quantum efficiency of the CCD is about 0.3-0.5 in the 500 to 700 nm spectral region.

The flight candidate CCDs have very few defects, less than 10 hot or dark pixels over the entire array, and no column defects. Since the full well capacity is between 150,000 and 250,000 electrons, and the read noise of the output amplifier is approximately 5 electrons, a dynamic range exceeding 30,000 is possible. Both vertical and horizontal charge transfer efficiencies (CTEs) are better than 0.999999 for signal levels greater than 0.1 of full well. At low signal levels (0.01 of full well), the CTE drops to about 0.999995.

The CCD is mounted on a custom multi-layer ceramic package that satisfies several requirements. The optical focal plane must be located very accurately in all dimensions. Dowel pin holes were provided in the copper-tungsten frame to accurately position the CCD. Measurements of the focal plane were made in a mechanical/optical jig, allowing a custom mounting plate to be made for the CCD to remove any manufacturing tolerances. The multi-layer ceramic package is able to heat the CCD, either to control the operating temperature of the CCD, or to increase its temperature above the ambient temperature to drive off any condensed vapors that might have collected on the surface. A temperature sensor is integrated onto the package, and another on the CCD die itself.

The package has provisions for a non-flight cover over the CCD to be attached, that protects the surface of the CCD during handling. The cover shorts all of the pins to a large ground plane, to protect against damage caused by electrostatic discharge, a common failure mechanism for CCDs. This grounding cover remained in place until after the CCD was installed into the camera circuitry, after which discharge damage is minimal.

8.2 The CCD Cameras

The CCD camera was designed to be very conservative in the critical spacecraft resources of power and mass, using about 5 watts power, and weighing about 3 kg. The CCD is operated at -80 C. It is susceptible to condensation, and needs to be kept very clean. Thus, the camera electronics are housed in a compartment separated from the CCD, and vented to the box exterior. The CCDs are cooled by passively radiating heat to deep space. A temperature of -80 C was chosen to reduce the effects of permanent proton radiation damage to the bulk silicon, which causes a drastic drop in the ability of the CCD to transfer charge. The CTE can then drop to 0.999, which would virtually destroy the image quality. The effect of this CTE difference is that for a point in the center of the array, the photoelectron charge packet would be reduced by 64% for a CTE of 0.999, compared to only 0.1% for the nondamaged CCD. Providing a mechanical shield only reduces the damage effects by a factor of about 3-6 for a reasonable shield thickness, which is insufficient, but cooling to below -70 C can avoid almost completely the CTE damage produced by radiation effects.

The camera accepts commands to initiate various setup parameters, to initiate the clearing cycle, and then to initiate readout. The readout rate is 50,000 pixels/sec, or about 22 sec for a full image readout. The camera provides all of the clock signals and voltages required by the CCD and the analog signal processing chain. The camera can also control the CCD. Unwanted lines can be dumped at the rate of 0.060 msec per line. Thus any line on the CCD can be accessed within about 60 msec. Additional capabilities of the cameras include setting the number of cycles for clearing, setting slow or fast line dumping, reading out of any one of the four CCD chip output ports, and setting of certain voltages for radiation damage compensation. While it is possible to use any of the CCD output ports and their associated electronics lines for an image readout, the camera design has resulted in the readout electronics associated with the CCD chip "D" output port having the best noise performance, less than 1 DN (Digital Number) without the CCD in place. The best output port on the CCD chips generally has a noise performance of 1 DN also, but this is not necessarily from the chip D port. The readout line with the best combined readout electronics noise and CCD port noise performance will be used until a failure occurs, forcing the use of another port.

The analog signal processing amplifies the output of the CCD (about 1.5-2.0 x 10E-6 volts per electron) by a factor of about 30, and uses the usual double-correlated sampling technique to sample the photoelectron charge packet. In this technique, the CCD output is measured just prior to injection of the electron charge packet, establishing a reference level. The CCD output is again sampled after the charge packet is injected, and the difference from the reference level is measured. The analog signal is digitized to 14 bits by an analog- to-digital converter, with a quantization step of about 15-20 electrons. The noise of the entire process is less than 10 electrons, so that the entire dynamic range of the system is the full 14 bits, or about 16,000. Note that this does not cover the full dynamic range that the CCD can achieve. Since the total system noise has been measured to be about 25 electrons, the noise will be dominated by photon noise.

While the CCD chip has an imaging area of 1024x1024 pixels, output of any image line from the readout register will contain 20 additional pixels at the beginning which are not used for imaging, but provide useful calibration information. These pixels in the readout register are called underscan pixels. The pixel addresses assigned to the image pixels will include these underscan pixels, so that the first true imaging pixel column is 21. In addition, calibration information on the horizontal and vertical charge transfer efficiency can be obtained by commanding the readout of virtual pixels beyond the imaging area, called overscan pixels, both horizontally and vertically. Normally, only the imaging pixels are downlinked to the ground, but the first image column is still numbered 21. The pixel numbering convention is always referenced to the CCD readout port, and column 21 will always be the column closest to the readout port being used. If the readout port changes, for example from A to D, the image as read out will be inverted. This will be corrected during the data reduction process.

Return to Title Page or Go to Next Chapter