The familiar geometry of fluorescent lights involve a long narrow glass tube with two electrical connections on each of the metal caps which seal the ends of the tube. The tube is filled with noble gases such as argon, neon or krypton to suppress chemical reactions resulting from the electrical discharges in the tube. According to Bloomfield, these gases are at a pressure of about 0.3% of atmospheric pressure. A few drops of mercury is placed in the tube, and the vapor pressure is sufficient to constitute something like one in a thousand of the atoms inside the tube.
If electrons are removed from the mercury atoms by collisions with high speed electrons, they can produce light by emission (see mercury spectrum) when electrons make transitions downward to fill the electron gaps produced. One key to getting light out of the fluorescent tube is then to produce the high speed electrons which can "excite" the mercury atoms so that they will produce light. This is done in most fluorescent tubes by heating a filament in the end of the tube, which frees some electrons . Other types use high voltages to eject electrons from electrodes at the ends. Once freed, the electrons can be accelerated in the tube by the applied AC voltage. Some of the electrons transfer energy to the mercury atoms in collisions, so that one or more of the electrons of the mercury atom is elevated to an excited state. Once the electrical discharge in the tube is started, the current must be controlled to maintain a steady light source. This is done by coil arrangement called a fluorescent light ballast.
The process of producing light from the mercury atoms is fairly efficient, but a large part of it is in the ultraviolet rather than the visible range. The final transition of the electrons to the ground state of the mercury atom produces light at 254 nm, considerably below the blue limit of human vision at about 400 nm. The ultraviolet light does not get through the glass envelope of the tube, but because of its high quantum energy it can be used to advantage in producing visible light. To produce light in the visible range, the inside of the tube is coated with a phosphor powder. When the ultraviolet light strikes the powder, it produces excitations of the electrons of the phosphor which then produce visible light by a process called fluorescence. Ultraviolet photons associated with the 254 nm uv light have quantum energies of 4.9 electron volts, whereas the energy range for the visible photons which we can see is from 1.6 to 3.1 eV. Since there are many intermediate levels for the electrons to drop to after being excited by the uv photons, they can produce visible photons of light throughout the visible range, producing nearly white light.
The production of white light is a challenging undertaking. In the case of the fluorescent lights, a carefully selected and blended set of phosphors is used so that the wavelengths at which it fluoresces are distributed evenly over the visible range. Current fluorescent lighting uses six standard phosphor blends: cool white, deluxe cool white, warm white, deluxe warm white, white, and daylight. The daylight phosphor which was used in early fluorescent lights tips the light toward the blue end and is criticized as being "cold". The "cool" phosphors resemble daylight and are more nearly color neutral, while the "warm" phospors tip the spectrum toward the red end and resemble incandescent lighting.
Practical electric circuits
Fluorescent Light Ballast
The operation of an incandescent light bulb is fairly simple and self-regulating. You apply the full electrical line voltage to the bulb and the current heats up the filament until it glows. The heating of the filament increases its electrical resistance, and that resistance limits the current to a controlled value.
You can't just apply the full voltage to a fluorescent tube; you must provide for starting the electrical discharge and then controlling the resulting current in the arc discharge in the bulb. Many different strategies and approaches have been employed - for detailed information you need an industry source like that maintained on the web by Summit Electrical.
Starting the lamp is the ballast's first task. The major types of starting strategies are (1) preheat, (2) slimline instant start, and (3) rapid start. If you need current technical information then you should be aware of two more recent types, (4) modified rapid start and (5) instant start of rapid start lamps.
The "preheat" strategy was the original method used for fluorescent lights. The filaments in the lamp are heated for a few seconds before applying the full operating voltage to the lamp. This is done by having a switch in parallel with the gas tube which shunts current around the gas discharge path and through the filament heaters. After a few seconds the filaments reach the temperature necessary for them to emit electrons, and the switch is opened, applying the operating voltage to the tube to start the arc discharge in the gas. The ballast must then employ the current regulating circuitry discussed below.
The "slimline instant start" system produces light instantly by employing a transformer in the ballast to produce a voltage about three times the normal operating voltage to "strike the arc" in the bulb. Preheating the filaments is not required for this kind of system.
The "rapid start" system is reportedly the most popular in the U.S. currently. These ballasts provide for continuous heating of the filaments for electron supply. They require that the fixture be properly grounded and that the lamps are within 1-2 cm of the metal fixture for proper starting. Because of the continuously heated filaments, these units do not require the high starting voltage of the slimline instant start types . The bulbs light immediately at low brightness and are fully lighted in about two seconds.
Once the bulbs are lighted, the ballast must control the current. An arc discharge is an inherently variable thing, and could be subject to high surge currents. The main mass of the ballast consists of a large coil wound around a laminated steel core to produce a large inductor, or "choke" as they are often called in the industry. The coil also acts as a transformer. The nature of an inductor is to limit the rate of change of current, so the large inductance of the ballast acts to suppress current spikes. The laminated-core coil is often "potted" in a material like asphalt to help with heat dissipation, and the combination is housed in a steel case.
There are also electronic and hybrid ballasts which perform the regulation tasks. Description of those systems may be added here. Comments and suggestions are welcomed. If you have detailed schematics of ballast operation, I would be interested - I have not found any.
Practical electric circuits