In this installment we will continue with the subject of fluorescent lighting systems and how they work but will be concentrating on hardware necessary to make them work as opposed to the lamps themselves (which was covered in part 1 of this series).

Unfortunately you can not just hook a fluorescent lamp up to an electrical outlet and expect it to work as you can an incandescent bulb. The principal problem is that it normally takes a voltage greater than the typical line voltage (110 Volts AC) to start a fluorescent lamp (i.e., initiate the electrical arc in the tube) in fact it can be several hundred volts depending on the specific lamp design. The second problem with operating a fluorescent lamp is that once the electrical arc is started in the lamp its electrical resistance drops due to the ionization of the gases in the lamp’s tube. This lower resistance would in turn cause higher current flow in the lamps electrodes causing more ionization and even lower resistance in the lamp causing even higher current flow an on and on till the lamp would quickly burn out. So to successfully operate a fluorescent lamp you have to provide two things, first a high enough starting voltage to initiate the electrical arc in the tube and then secondly a means of limiting the current flow in the tube to keep it from burning itself out. The normal way of accomplishing both of these tasks is through use of an added piece of hardware called a ballast. There are two fundamental types of ballasts, standard or iron core ballasts and newer electronic ballasts. We’ll start with the more common iron core ballast and its variants and then contrast its operation and performance with the newer electronic ballasts.

The two most common iron core ballasts used today are called "rapid start" and "instant start" ballasts.

Figure #1 shows a simplified wiring diagram for a instant start ballast for a pair of lamps and consists of an auto-transformer, chokes, a capacitor and of course the lamps themselves. The auto-transformer is used to convert the input 110V AC line voltage to the much higher voltage

(500-600V) required to start the electric arc in the fluorescent lamps. The chokes are just inductors made up of windings on an iron core and are used to limit the current to the lamps. This current limiting by the choke is accomplished by acting as a high resistance element to the 60 Hz line frequency. When the line voltage is first applied to the input of the auto-transformer of the ballast no current initially flows in the lamp because it is very high resistance. As the voltage across the lamp increases it eventually reaches a critical value where the electrical arc starts to flow in the lamp. When the electrical arc or current starts to flow in the lamp its resistance drops to a much lower value and would cause too high a current flow to the lamp except that the chokes higher resistance in series with the lamp limits the maximum current flow to the lamp. You will also remember there is a capacitor in line with the choke and the lamp, this capacitor is used to improve the efficiency of the ballast by increasing its "power factor".

Power factor refers to the phase or timing relationship of the voltage and current waveforms in the ballast. The actual real power delivered to the lamp is given by:

Where the Cosine term is referred to as the "power factor".

AC line voltage consists of a 60 Hz (or cycles/second) sinusoidal waveform and when this signal is applied across a simple resistance its corresponding current waveform is in step with its voltage waveform or said to be in phase (see figure #2). If you now apply this AC voltage across an inductor such as our choke you cause the voltage waveform to lead the current waveform by 90 degrees due to magnetic fields generated by the inductor (see figure #2). If you were to apply the AC voltage across a capacitor rather than an inductor you would see that the voltage now lags the current waveform by 90 degrees due to electric fields generated by the capacitor (see figure #2). By combining the capacitor in series with the inductor or choke you help bring the voltage and current waveforms back in phase more similar to that seen with a simple resistance and in doing so make the Cosine term or power factor in the power equation to approach unity, giving the maximum power for a given applied voltage and current. This circuit configuration is often referred to as a "lead-lag" circuit corresponding to the effects of the inductor (choke) and capacitor have on the voltage and current waveforms phase relationships.

The circuit shown in figure #1 is referred to as a parallel configuration since each lamp can operate independent of the other. There exist other configurations for multiple lamp instant start ballasts in which the lamps are in series with each other (figure #3), this has the advantage of reducing the ballast size since the same current flows in both lamps and a single choke can be used to limit the current. The down side to this configuration is that if any one lamp burns out the other lamp will not operate either and the starting circuitry is more complex since the lamps have to start in sequence to keep the starting voltage requirements down. Another problem with instant start ballasts be they parallel or series is that the high starting voltages required to start the lamps tend to reduce their lifetimes, one way to address this weakness is with a "rapid start ballast".

The most common form of ballast used today for standard, HO and VHO lamps is the "rapid start" ballast. It was said earlier that a very high voltage (500-600V) is required to start the electric arc in most fluorescent lamps. This very high starting voltage assumes that you are trying to cold start a lamp, if you could preheat the lamps electrodes slightly you could partially ionize some of the internal lamp gases and reduce the voltage required to start the lamp. This is exactly what a rapid start ballast does through the use of special cathode heater windings and in turn allows for a smaller ballast since it does not require as high a voltage to start the lamps (250-400V).

Figure #4 shows a typical series configuration rapid start ballast for a pair of lamps. You can see two new components from earlier instant start ballast circuits, the cathode heater windings mentioned earlier and a starting capacitor. The starting capacitor is only needed for dual lamp ballasts and is there to allow sequential starting of the lamps, remember that sequentially starting lamps in series reduces the maximum voltage required to start them. The secondary windings are the same as the chokes in the instant start ballast circuit and also used to limit the current to the lamps once they are operational. The second capacitor is also there for the same reason seen before or to improve the ballast’s power factor and thus electrical efficiency.

One potential issue with a rapid start ballast is that they often need to have their lamp(s) mounted close (< 1") to a grounded reflector to start reliably. This requirement is due to the lower voltages used to start the lamps. By placing a ground plane such as a grounded reflector close to the lamp(s) the ease of ionizing the gases in the lamp is enhanced due to capacitive coupling between the lamps glass tube and the near by ground plane allowing these lower starting voltages to better initiate the electrical arc in the lamp.

In many respects an electronic ballast works the same way an iron core ballast works but does it at a much higher frequency. Remember that line voltage is an alternating current (AC) with a frequency of 60 Hz, an electronic ballast takes this 60 Hz waveform and transforms it into a higher frequency waveform, typically 20,000 Hz or higher. The means of doing this frequency transformation is through a special circuit call an inverter/switcher. Operating a fluorescent lamp at this higher frequency offers several advantages:

1. Allows use of smaller chokes and capacitors making for a smaller and lighter ballast.
2. Electrical operating losses in the ballast are also reduced by running at the higher frequency allowing it to dissipate less heat.
3. Quieter operation of the ballast since it is operating at frequencies above our hearing (no 60/120 Hz hum).
4. Less flicking of fluorescent lamp again due to higher frequency of operation.
5. Lower operational costs (10-20% lower depending on design) due to improved efficiencies.
6. The lamp’s efficiency itself is improved when operating at higher frequencies creating about 10% more light for the same power input or alternately allowing the input power to be reduced by 10% to give equivalent light output and longer lamp life.

This last listed advantage of higher lamp light efficiencies needs further explanation. It may not have been obvious from the earlier discussion but fluorescent lamps actually start, stop and restart 120 times a second corresponding to the twice the 60 Hz line frequency, this is typically fast enough that we do not notice it and it appears the lamp is on all the time. By increasing this frequency of on/off you actually increase the amount of light that is generated by the lamp for the same power input due to better efficiencies in the mechanisms generating the light in the lamp.

There are some down sides to electronic ballasts and some of these are listed below:

1. Depending on ballast design may cause more electrical interference in other electrical equipment due to high frequency circuit operation.
2. Higher initial cost over electromagnetic ballast (but operational cost savings may offset this initial higher cost over time).
3. Higher "inrush" current for electronic ballasts on start up possibly causing breaker or fuse tripping.

The concept of "inrush" current is simply that when a ballast is first powered up there is a momentary (typically < 10 milliseconds) surge of current to the ballast. This initial current surge for conventional iron core ballasts can be several times (2-3X) its full operational current but an electronic ballast due to it having an initial lower input impedance from its special circuitry may have 4-6X inrush currents. The end result of all this is that care must be taken when using electronic ballasts (an to a lesser extent iron core ballasts) by making sure any electrical breakers or fuses used will be able to handle the short duration current surges without tripping or blowing. Additionally any switches or relays used to turn the ballast on and off need to be able to handle these same short duration higher currents to operate reliably.

The last topic for discussion in this installment is dimming of fluorescent lamps. While dimming systems exist for iron core ballasts they typically are not very efficient and often noisy when dimmed so for best operation if you need dimming capability for fluorescent lamps its best to stick with electronic dimmable ballasts.

Though there are several dimming approaches to electronic ballasts the most common is one that uses an external control voltage of 0-10 V DC to adjust the dimming level of the lamp. This light level adjustment normally runs from 100% to 20%, much lower than this and reliable operation of the lamp may suffer. Another issue is that only rapid start type ballasts can be dimmed as you normally need to keep the cathodes or filaments powered at fixed or higher current levels during dimming and this can not be done with instant start ballasts as they have no cathode heating windings. The dimming of the light itself is accomplished by reducing the current flowing through the lamp. A properly designed dimming electronic ballast should not adversely effect the life of a suitable fluorescent lamp (some lamps such as energy savings lamps are not recommended for use with dimming ballasts).

This wraps up fluorescent lighting systems and in the next and last article in this series we will tackle HID lighting systems as well as other more exotic or alternative lighting systems.