| Previous | Table of Contents | Next |
As you read CCD literature, you will encounter new terminology. Here are a few important terms.
The quantum efficiency (QE) metric is the ratio of detected electrons divided by the number of electrons expected. The expected count depends on the number of photons and the number of electron-hole pairs the photon should create. Some incident photons may reflect off the CCD surface or may impact outside of any photosite. Others, depending on their frequency, create more than one electron-hole pair. CCDs usually have a QE ranging from 50 percent to 90 percent. In comparison, human eyes have less than 1 percent QE. Photographic film has a 5 percent to 20 percent QE.
The charge transfer process is the process of extracting photosite charges. It is sometimes called “clocking the gates” and can take several seconds to complete before another exposure is possible.
The charge transfer efficiency (CTE) measures the charge transfer efficiency which is the percent of charge transferred from one pixel to the next. Typically, the CTE is greater than 99.999 percent per photosite, where 100 percent is perfect. The CTE is important because on a 1024 x 1024 photosite CCD, a charge may transfer from two to 2,048 times and the loss is cumulative.
With a front-illuminated device, photons pass through gate structures and a protective layer to generate a charge. Unfortunately, some photons may be absorbed in these layers and not contribute to the signal.
With a back-illuminated device, incident photons strike a specially thinned CCD from the rear, bypassing the front surface electrodes and passivation layers. An optional enhancement layer forces photo-generated electrons toward the collection areas, and an optional anti-reflective coating may reduce reflections.
Any detector with high dynamic range can simultaneously detect low and high intensity objects. CCDs have a dynamic range sensitive to object intensities differing by a factor of 10,000. In contrast, photographic film has a dynamic range of about 100.
Dark current is an undesirable charge that accumulates in CCDs whenever they operate at temperatures above absolute zero (always). Dark current contributes sampling errors since it is indistinguishable from photon-generated charges. Dark current is similar to inadvertently exposing film to light. Sufficient CCD cooling (e.g., liquid nitrogen) effectively eliminates the problem.
CCDs have a uniform response characteristic. Because a CCD detector is permanent, it is straightforward to determine the behavior of each photosite, and differences can be compensated for using a “flat field” frame. This is not possible with photographic film which is always variable.
Saturation and blooming occur in all CCDs. If individual photosites comprise an “electron well,” saturation occurs when the well becomes full. The maximum charge which a photosite can store is principally determined by its area. As a photosite approaches its maximum sustainable charge, it loses its ability to accommodate additional charge which then blooms (spreads) into adjacent pixels. This causes reporting or cascades the saturation problem. The blooming manifests itself in the image as a white streak or blob. Controlling saturation and blooming depends on the CCD architecture.
Binning combines adjacent pixel charges during the charge transfer process. This trades spatial resolution for improved signals and extends the CCD dynamic range. It also accelerates the charge transfer process. Parallel binning combines two or more rows before serial shifting. Serial binning combines two or more charges in the output amplifier before charge digitization. The CCD binning factor specifies the number of pixels combined.
A full frame CCD uses the parallel register array for exposure, charge integration, and charge transport. A shutter controls exposures and blocks light during the charge transfer process, preventing image smearing.
A frame transfer CCD divides the parallel register into two equal size areas. The image array is located farthest from the serial register and is where the images are focused. The remaining area is covered with an opaque mask and provides temporary storage for image area charges. After exposing the image array, the collected photosite charges rapidly shift to the storage array. This allows the image array to capture another image during the storage array charge transfer. Frame transfer CCDs can operate continuously at high rates without requiring a shutter.
Active Pixel Sensor (APS) is a relatively recent solid-state imaging technology that uses a more prevalent fabrication process. (Where CCDs use NMOS, APS uses CMOS.) This allows building sensors in existing semiconductor fabrication facilities. Advocates stress the following APS advantages:
However, as attractive as APS appears, CCD technology still has a promising future. CCD scientists are developing features to offset some APS advantages. For the near term, CCDs enjoy significantly higher resolutions because their photosites do not require the large readout amplifier that APS requires at each photosite.
CCDs promise ubiquitous digital imaging and are in perpetual evolution. Daily developments and product announcements clearly indicate we will be increasingly immersed in its output. For more information on CCD technology, see:
| Previous | Table of Contents | Next |