INVERTER SWITCHING DEVICES
As far as the user is concerned, it does not really matter what type of switching device is used inside the inverter, but it is probably helpful to mention the four most important families of devices in current use so that the terminology is familiar and the symbols used for each device can be recognised. The feature which unites all four devices is that they can be switched on and oV by means of a low-power control signal, i.e. they are self-commutating. We have seen earlier that this ability to be turned on or oV on demand is essential in any inverter which feeds a passive load, such as an induction motor.
Each device is discussed brieXy below, with a broad indication of its most likely range of application. Because there is considerable overlap between competing devices, it is not possible to be dogmatic and specify which device is best, and the reader should not be surprised to Wnd that one manufacturer may oVer a 5 kW inverter which uses MOSFETs while another chooses to use IGBTs. The whole business of power electronics is still developing: there are other devices (such as those based on silicon carbide) that are yet to emerge onto the drives scene. One trend which continues is the integration of the drive and protection circuitry in the same package as the switching device (or devices). This obviously leads to considerable simpliWcation and economy in the con- struction of the complete converter.
Bipolar junction transistor (BJT)
Historically the bipolar junction transistor was the Wrst to be used for power switching. Of the two versions (npn and pnp) only the npn has been widely used in inverters for drives, mainly in applications ranging up to a few kilowatts and several hundred volts.
The npn version is shown in Figure 2.18: the main (load) current Xows into the collector (C) and out of the emitter (E), as shown by the arrow on the device symbol. To switch the device on (i.e. to make the resistance of the collector–emitter circuit low, so that load current can Xow), a small current must be caused to Xow from the base (B) to the emitter. When the base–emitter current is zero, the resistance of the collector– emitter circuit is very high, and the device is switched oV.
The advantage of the bipolar transistor is that when it is turned on, the collector–emitter voltage is low (see Figure 2.3) and hence the power dissipation is small in comparison with the load power, i.e. the device is an eYcient power switch. The disadvantage is that although the power required in the base–emitter circuit is tiny in comparison with the load power, it is not insigniWcant and in the largest power transistors can amount to several tens of watts. This means that the complexity and cost of the base-drive circuitry can be considerable.
Metal oxide semiconductor field effect transistor (MOSFET)
Since the 1980s the power MOSFET has superseded the BJT in inverters for drives. Like the BJT, the MOSFET is a three-terminal device and is available in two versions, the n-channel and the p-channel. The n-channel is the most widely used, and is shown in Figure 2.18. The main (load) current Xows into the drain (D) and out of the source (S). (Confusingly, the load current in this case Xows in the opposite direction to the arrow on the symbol.) Unlike the BJT, which is con- trolled by the base current, the MOSFET is controlled by the gate- source voltage.
To turn the device on, the gate-source voltage must be comfortably above a threshold of a few volts. When the voltage is Wrst applied to the gate, currents Xow in the parasitic gate-source and gate-drain capacitances, but once these capacitances have been charged the input current to the gate is negligible, so the steady-state gate drive power is minimal. To turn the device oV, the parasitic capacitances must be discharged and the gate-source voltage must be held below the threshold level.
The principal advantage of the MOSFET is that it is a voltage- controlled device which requires negligible power to hold it in the on state. The gate drive circuitry is thus less complex and costly than the base-drive circuitry of an equivalent bipolar device. The disadvantage of the MOSFET is that in the ‘on’ state the eVective resistance of the drain source is higher than an equivalent bipolar device, so the power dissipation is higher and the device is rather less eYcient as a power switch. MOSFETs are used in low and medium power inverters up to a few kilowatts, with voltages generally not exceeding 700 V.
Insulated gate bipolar transistor (IGBT)
The IGBT (Figure 2.18) is a hybrid device which combines the best features of the MOSFET (i.e. ease of gate turn on and turn oV from low-power logic circuits) and the BJT (relatively low power dissipation in the main collector–emitter circuit). These obvious advantages give the IGBT the edge over the MOSFET and BJT, and account for their dominance in all but small drives. They are particularly well suited to the medium power, medium voltage range (up to several hundred kilo- watts).
The path for the main (load) current is from collector to emitter, as in the npn bipolar device.
Gate turn-off thyristor (GTO)
The GTO (Figure 2.18) is turned on by a pulse of current in the gate- cathode circuit in much the same way as a conventional thyristor. But unlike an ordinary thyristor, which cannot be turned oV by gate action, the GTO can be turned oV by a negative gate-cathode current. The main (load) current Xows from anode to cathode, as in a conventional thyris- tor. The twin arrowed paths on the gate lead (Figure 2.18) indicate that control action is achieved by both forward and reverse gate currents. (In US literature, a single gate lead with a short crossbar is used instead of the two arrows.)
The gate drive requirements are more demanding than for a conven- tional thyristor, and the on-state performance is worse, with a forward volt-drop of perhaps 3 V compared with 1.5 V, but these are the penalties to be paid in return for the added Xexibility. The GTO has considerably higher voltage and current ratings (up to 3 kV and 2 kA) than the other three devices and is therefore used in high-power inverters.