In the late 1970s advances in semiconductor technology allowed the manufacture of light-sensitive imaging detectors known as chargecoupled devices. They came into widespread use in astronomy during the 1980s, and recent advances have expanded their employment in commercial applications. The same detecting devices now found in consumer digital cameras and video cameras are located at the focal point of most of the world's research telescopes. The CCD is a grid of light-sensitive elements (called pixels), each of which records the detection of a photon by emitting an electron as a result of the photoelectric effect. The electrons that pile up in a given pixel are eventually "read out" of the CCD and recorded on a computer. Each pixel then has an electron count that is directly proportional to the number of photons that struck it.
The rapid advances that have occurred in CCD technology within the past 10 years have decreased their physical dimensions, improved their sensitivity, and lowered their noise (visual interference) characteristics. As fabrication techniques have improved, the typical size of a CCD has increased from 32 pixels on a side to 4,096 pixels on a side. The largest-format CCD cameras now employ several CCD chips aligned side by side to produce a larger detecting area. CCDs are supremely sensitive, currently producing an electron for each photon that strikes their surface. Early models were not this efficient, often losing electrons, thereby decreasing sensitivity.
The images CCDs produce can be immediately processed electronically and improved in quality. The main drawback of CCDs has been that they are relatively small and therefore cannot make wide-field images (which sample a large piece of the sky). The Hubble Space Telescope, for example, uses only four detectors (CCDs) in conjunction to produce images from its WF/PC2 and only samples a fraction of the total focal plane of the telescope. While the size of CCDs is a limitation, there is no alternative detector with equal sensitivity. The problem with CCDs generally is that they are small, and the area in a telescope's focal plane is large. As an example, Schmidt plates (the photographic plates that can cover about 95 percent of the focal plane of a typical Schmidt camera with a diameter of 0.5 m) are about 12 inches square. A fantastically large CCD is 1 inch on a side. A huge number of these large-format CCDs (about 144) would be required to cover the area of the focal plane. Take 144 (the number of CCD chips required) and multiply it by 4,096 X 4,096, and you get the number of pixels produced by a single snapshot. Limitations in the data transmission rates prohibited (at least at the time the HST was built) a fully sampled HST focal plane from making the design cut.
Was this article helpful?