About CCD Technology

A Charge Coupled Device (CCD) senses light that is incident on its silicon surface. This surface is divided into discrete square cells, each several microns across (e.g. 5.25um). In a document scanner, the CCD will have a line of, say, 5000 cells (or pixels). For color scanning, the CCD will have 3 or 4 closely-spaced parallel lines of pixels, each line being overlaid by a different color filter.

CCD scanner cross section

CCD scanner cross section.

Now consider the geometric requirements for a CCD scanner. To produce a scanner with 400 dpi optical resolution, the scanner manufacturer needs to design an optical system which focuses light from a 1/400" square pixel at the object (document) onto a 5.25 um square pixel at the CCD silicon surface, i.e. a reduction of approximately 12:1. This is done using photographic enlarger lenses and, most likely, mirrors to fold the optical path into a reasonable space. The use of such lenses and mirrors will introduce some geometric distortion and some limits on focus, thereby effecting geometric accuracy and resolving power. The magnitude of these effects will be determined by the performance of the lens and the mirrors and how much the manufacturer is prepared to pay for these key components.

The manufacturer of a CCD scanner must also design an illumination system, generally using one or two fluorescent tubes. Because the manufacturer controls this illumination, and is able to apply proprietary processing to the CCD output, it is possible to optimise the color gamut and the dynamic range of a CCD scanner. The color content is determined by using a "white" light (controlled by the scanner manufacturer) combined with colored filters within the CCD (controlled by the CCD manufacturer).

The use of fluorescent tubes does mean that a warm-up period is necessary, to allow the illumination levels to stabilise as the fluorescent tubes heat up. The need for frequent re-calibration of a CCD scanner is also mainly due to variations in illumination.