CCD in Digital Camera - Turning an Image into a Bunch of Electrons
How it is manufactured and what happens in it when we press the shutter release button is of interest to any good photographer, because it may contribute in taking better photos.
Moreover, basics of lingo are useful for the understanding of characteristics of a camera, especially if you are going to buy one.
A pixel element of a CCD, in its simplest form, is basically a MOS capacitor, i.
a semiconductor structure made of three layers: Metal-Oxide-Semiconductor.
The semiconductor part of the MOS is simply silicon while the oxide is silicon dioxide, acting as an insulating layer.
The metal is not a real metal substance, but rather a heavily doped polysilicon layer and is named "gate".
Basically, the image is formed within the semiconductor part of the MOS system and the metal gate is used for shifting such image.
When we press the shutter release button on our camera, light starts hitting the silicon in the MOS structure, and hole-electron pairs are created via the photoelectric effect.
While the holes are drained by the grounded substrate and not utilized, electrons are collected in the semiconductor part of the MOS.
We are interested in these electrons, because they will build up our photograph.
The region where electrons are collected is referred to as the "channel" and, of course, corresponds to the electron lowest potential energy.
In order to get better images, basically all CCD manufacturers use the so called "buried channel structure", where the electrons are collected not at the oxide-semiconductor interface, but a little bit distant, within the semiconductor (hence the name "buried").
In order to confine the photo-generated electrons in the channel, "channel stops" are created in the MOS structure, isolating each MOS element from the adjacent ones.
The number of photoelectrons collected is linearly proportional to the intensity of the photon flux and to the time this flux hits the pixel ("integration time").
Consequently, we now know that the brighter the registered scene, the greater the number of electrons.
The efficiency with which incident photons are detected is known as "quantum efficiency"; a typical value is 40%.
This value is pretty low because of the absorption caused by the passivation (protection) layer present on the integrated circuit and the presence of the polysilicon gate that must be passed through by the incoming photon.
A single pixel of our digital camera is made up of a structure similar to the MOS capacitor outlined above.
It's a little more complicated, though, because we don't want electrons to be confined where they are generated forever, but want to sense how many electrons have been generated.
In order to accomplish this, each CCD pixel is made up of 3 parallel gates (although variations are possible) and, perpendicularly to these, a channel stop on both sides.
By keeping the central electrode (gate) at a higher potential than the other two, electrons will be attracted there and so collected.
The CCD is then made up of a matrix of millions of these pixels, arranged in thousands of rows and thousands of columns.
So, upon pressing the shutter release we start the acquisition phase, and this ends after the exposure time set by the photographer (typically 1/2000s to a few seconds).
At this point we must read the number of electrons collected in each pixel.
The greater the number of electrons, the brighter the pixel.
For this purpose, a charge transfer process must take place from each pixel to a sensing circuitry.
To achieve this, a shift phase occurs, exploiting the 3 gate structure described above, after which the electrons of the first row are shifted into an array of serial registers located at the edge of the CCD matrix, the second row electrons are shifted to the first row and so on.
The efficiency with which the transfer process occurs is measured by a parameter called "Charge Transfer Efficiency", a typical value of which is 99.
999% per pixel.
The serial register shifts its content into a charge detection output amplifier one pixel at a time.
The output amplifier converts the electrons' charge to a voltage.
The order of magnitude is an output voltage of about 1 microvolt per electron and this is a linear relationship.
The slope of this curve is referred to as the "output sensitivity" or "conversion gain".
The higher the voltage, the brighter the pixel.
Once all the pixels of the first row are read by the output amplifier, the shifting phase takes place again and the whole sensing process is repeated.
This is so until all the pixels in the matrix are read out.
This is the end.
Nothing else happens in the CCD chip.
All the rest of the image processing is done off-chip.
In particular the voltage read out by the CCD is first amplified and then converted into a digital value by an off-chip analog to digital converter.