Why the name? The “charge-coupled” part refers to the fact that the device works by moving, or “coupling”, charges between two adjacent locations.

For small numbers of pixels, one can imagine connecting up the wires for each silicon pixel, and then recording the intensity (e.g., sensing whether the Sun is out or not). However, even for very basic digital cameras today, there are over 1,000,000 (one million) pixels (a mega-pixel). Clearly, the number of wires would be very large, and makes this impractical. Several schemes have evolved to solve this problem to make the low-cost devices we are so fond of. One very popular solution, invented in the 1960’s, is the charge-coupled device or CCD.

In a CCD, a set of charges is generated by the silicon in a grid, and held in place using a set of electrodes at particular voltages that “fix” the charges at the intended pixel locations. This works on the principle that the photoelectrons have negative charge, so they are attracted to positive voltages, and repelled from negative voltages. If the electrodes are placed so that a pixel is surrounded by a negative voltage, then the photoelectrons will accumulate in the middle of the pixel.

After collection, these sets of photoelectrons need to be measured so that we know how many photoelectrons ended up in each pixel. A CCD does this by manipulating the voltage in the electrodes to “march” the photoelectrons in the pixels out to a corner of the image. Then, by looking at the amounts of charge in the pixels after they arrive at the corner, we can reconstruct what the image looked like. Thus, it allows us to read the entire image out by looking at just one set of wires connected to the corner. The electrical signal on this wire changes with time to reflect the sequence in which the charges are “marched” out of the CCD. This process of moving charge from one pixel location to an adjacent one is often compared to a water “bucket brigade”.

By manipulating the voltages, we make the parallel registers shift once, to move a line of charges into the serial register. The serial register is then shifted however many times it takes to move the line of charges to a corner. At the corner, there is an electrical circuit (“read-out amplifier”) that converts the charge into a voltage that can be fed to some electronics for further processing (such as feeding into a computer). By repeating the parallel register shifts, we can read out the entire CCD.

This is the simplest form a CCD can take. There are now many fancy variations in the read-out method, including having multiple read-out amplifiers, having adjacent pixels sense different colors, etc.