# Ballast Transformer

Until just recently, all fluorescent lights such as appearing in Fig. 1(a) had a ballast transformer as shown in Fig. 1(b). In many cases its weight alone is almost equal to that of the fixture itself. In recent years a solid-state equivalent transformer has been developed that in time may replace most of the ballast transformers. However, for now and the near future, because of the additional cost associated with the solid-state variety, the ballast transformer will appear in most fluorescent bulbs.
Fig. 1: Fluorescent lamp: (a) general appearance; (b) internal view with ballast.
The basic connections for a single-bulb fluorescent light are provided in Fig. 2(a). Note that the transformer is connected as an autotransformer with the full applied 120 V across the primary.
Fig. 2: (a) Schematic of single-bulb fluorescent lamp; (b) starter; (c) internal view of ballast transformer
When the switch is closed, the applied voltage and the voltage across the secondary will add and establish a current through the filaments of the fluorescent bulb. The starter is initially a short circuit to establish the continuous path through the two filaments.
In older fluorescent bulbs the starter was a cylinder with two contacts, as shown in Fig. 2(b), which had to be replaced on occasion. It sat right under the fluorescent bulb near one of the bulb connections. Now, as shown by the sketch of the inside of a ballast transformer in Fig. 2(c), the starter is now commonly built into the ballast and can no longer be replaced. The voltage established by the autotransformer action is sufficient to heat the filaments but not light the fluorescent bulb. The fluorescent lamp is a long tube with a coating of fluorescent paint on the inside. It is filled with an inert gas and a small amount of liquid mercury. The distance between the electrodes at the ends of the lamp is too much for the applied auto-transformer voltage to establish conduction.
To overcome this problem, the filaments are first heated as described above to convert the mercury (a good conductor) from a liquid to a gas. Conduction can then be established by the application of a large potential across the electrodes. This potential is established when the starter (a thermal switch that opens when it reaches a particular temperature) opens and causes the inductor current to drop from its established level to zero amperes. This quick drop in current will establish a very high spike in voltage across the coils of the autotransformer as determined by $v_L = L(di_L /dt)$. This significant spike in voltage will also appear across the bulb and will establish current between the electrodes. Light will then be given off as the electrons hit the fluorescent surface on the inside of the tube. It is the persistence of the coating that helps hide the oscillation in conduction level due to the low-frequency (60-Hz) power that could result in a flickering light. The starter will remain open until the next time the bulb is turned on.
The flow of charge between electrodes will then be maintained solely by the voltage across the autotransformer. This current is relatively low in magnitude because of the reactance of the secondary winding in the resulting series circuit. In other words, the autotransformer has shifted to one that is now providing a reactance to the secondary circuit to limit the current through the bulb. Without this limiting factor the current through the bulb would be too high, and the bulb would quickly burn out. This action of the coils of the transformer generating the required voltage and then acting as a coil to limit the current has resulted in the general terminology of swinging choke.
The fact that the light is not generated by an IR drop across a filament of a bulb is the reason fluorescent lights are so energy efficient. In fact, in an incandescent bulb, about 75% of the applied energy is lost in heat, with only 25% going to light emission. In a fluorescent bulb more than 70% goes to light emission and 30% to heat losses. As a rule of thumb, one can assume that the lighting from a 40-W fluorescent lamp [such as the unit of Fig. 1(a) with its two 20-W bulbs] is equivalent to that of a 100-W incandescent bulb.