CCD sensors
- obsolete technology
The term "CCD" means "charge-couple device" and is an electronic module which is capable to transport electronic charges.
This technology was invented in 1969 in the Bell Laboratories by Willard Boyle and George E. Smith as pure data storage possibility, for storage purposes, however, it never gained technical acceptance. Yet this principle has still successfully been used until today in order to transport charges generated by exposing a semiconductor due to the "inner photon effect" by means of a great number of small steps (vertical and horizontal shift registers) to a central A/D converter, i.e. using a sort of "bucket chain principle". The charge shifting by means of electrodes is forced from the outside by applying electric potentials:
|
|
(Willard Boyle and George E. Smith received the Nobel Prize in Physics in 2009 for this fundamentally important invention 40 years after the discovery of this effect.)
An image sensor which makes use of this principle to shift and read out the charges of the light-active pixels is called "CCD sensor".
The frequency, i.e. how often per second the sensor is capable to transport the charge by one pixel, is called "pixel clock". The frequencies at which CCDs were operated ranged from 25 to 50 MHz.
Simplified functional principle of the sensitivity to light
The inner photoelectric effect is fundamental for the functioning of the sensor.
The atoms of the silicon crystal are located on discrete energy bands, the energetically lower is called valence band, the energetically higher is called conduction band. In the basic state, most of the electrons are on the valence band, however they can be transported to the conduction band by means of excitation from the outside.
The energy required for this is 1.26 eV or more. In case of the CCD sensor, this transport can be induced by light, but also by higher heat supply (dark noise of the sensor). (Note: 1.26 eV approximately correspond to the energy of infrared radiation with a wavelength of 1 µm. Light with longer waves can pass through the silicon without absorption, it is virtually translucent from these wavelengths on and insensitive.)
Due to the excitation, free (negative) electrons and positively charged "holes" in the valence band are created at the same time which separate from each other because of the applied voltage. These charges, however, do not immediately flow off to the outside (like in the case of a photo diode in the CMOS sensor), but are stored in the memory cell itself.
By means of the charge shifting described above, the charge is transported through the entire sensor in many small steps towards a central amplifier / A-D converter.
Typical layout of CCD sensors
In general we differentiate several typical CCD sensor layouts which are described subsequently:
of which mainly the so-called "interline transfer sensor" has gained acceptance in machine vision.
|
|
Only this could record very fast frame rates using an "electronic shutter function" and was therefore used in 90 percent of CCD industrial MV cameras until 2015. After that, better CMOS sensors, especially from SONY, On Semi and other companies, gradually entered the market and CCD product lines were increasingly discontinued.
Frame transfer CCDs and full frame transfer CCDs were mainly used in scientific applications with low image refresh rates where extremely high sensitivity to light was the main criterion.
Advantages and disadvantages of the CCD sensor technology
Since the conversion of charge to voltage for each pixel is done by a single central amplifier/ AD converter, many advantages and disadvantages resulted in comparison to the CMOS sensor. Today, however, the CCD sensor is an obsolete technology, as modern manufacturing steps for consumer electronics (smartphone cameras) have developed CMOS technology to such an extent that its advantages now outweigh its disadvantages. The CCD camera sensor is effectively dead for mass applications.
Advantages were for a long time:
- Higher sensitivity and lower noise due to enhanced surface use (higher fill factor)
- Fewer defective pixels due to the simpler structure
- Better image homogeneity thanks to the central A/D converter
Disadvantages until today:
- Slower readout, as only one central A/D converter digitalises
- No direct pixel access, like in case of the CMOS sensor, as the CCD sensor must be read out serially
- More complex camera layout due to required additional electronics leads to larger and more expensive cameras
- Higher energy consumption of the entire camera
- More smearing and blooming effects when overexposing compared to a CMOS sensor
Due to modern manufacturing technologies, the CMOS sensor has overtaken the image quality of CCD sensors since around 2015. Therefore, manufacturers have gradually switched all camera series to CMOS sensors.