The development of the CMOS image sensor in the 1980's brought about the possibility of inexpensive pixel arrays, which because of their lower power demands, low system complexity and the ability to integrate processing on the same chip have become increasingly common throughout imaging technologies. CMOS image sensors come in two ‘flavours’, Passive Pixels and Active Pixel Sensors (APS).

Passive pixels

The Passive pixel array normally consists of a grid of photo-detectors, usually photo-diodes and a switching MOSFET, as shown in the figure to the left. The output is read by an amplifier in each column or a single output for the complete device, as shown. The sensitivity, noise performance and speed of these devices are poor compared to CCD devices largely due to the high capacitance at the input of the output amplifier.

Active pixels

Active pixel sensors differ from passive sensors by the inclusion of an amplifier at pixel level. This improves the noise figure and thus the achievable dynamic range and S/N ration of the array. As in the passive pixel approach the photo-sensing element is usually a diode, however a photo-gate has also been used. The architecture of a simple APS is shown in the figure to the right, it is also common in these systems to have some column level processing in addition to the column buffers. 

Photodiode APS

To the left a simple photodiode-type APS is shown. A source follower buffers the output of the photodiode, and thus a voltage on the column out line represents the signal. The configuration shown is a simple integrating pixel, sometimes known as a reset pixel. 

Integrating pixel

The figure to the right shows the schematic for an Integrating pixel. These devices exhibit a linear response to light. Typically an additional transistor is incorporated into the pixel and acts as a shutter as shown in the figure below. The reset and shutter are set high initially charging the photodiodes capacitance to VDD. At the beginning of the exposure reset falls low, whilst the shutter remains high, the incident light falling on the photodiode causes the capacitance (of the photodiode) to discharge. The rate of the discharge is proportional to the intensity of the light falling on the sensitive area and the capacitance of the photodiode. The larger the capacitance of the photodiode, the more photoelectrons it takes to discharge it, thus making it less susceptible to reset and shot noise.