## 62 DC Regulation and Voltage Conversion

6.2.1 Switching devices

The voltage from all sources of electrical power varies with time, temperature, and many other factors, especially current. Fuel cells, for example, are particularly badly regulated, and it will always be necessary to control the output voltage so that its only varies between set boundaries. Battery voltage is actually quite well regulated, but frequently we will want to change the voltage to a lower or higher value, usually to control the speed of a motor. We saw in the last section that if an electric motor is to be used in regenerative braking we need to be able to boost the voltage (and reduce the current) in a continuously variable way.

A good example to illustrate the variable voltage from a fuel cell system is given in Figure 6.10. It summarises some data from a real 250 kW fuel cell used to drive a bus (Spiegel et al. 1999). The voltage varies from about 400 to over 750 V, and we also see that the voltage can have different values at the same current. This is because, as well as current, the voltage also depends on temperature, air pressure, and on whether or not the compressor has got up to speed, among other factors.

Most electronic and electrical equipment requires a fairly constant voltage. This can be achieved by dropping the voltage down to a fixed value below the operating range of the fuel cell or battery, or boosting it up to a fixed value. In other cases we want to produce a variable voltage (e.g. for a motor) from the more-or-less fixed voltage of a battery. Whatever change is required, it is done using 'switching' or 'chopping' circuits, which are described below. These circuits, as well as the inverters and motor controllers to be described in later sections, use electronic switches.

As far as the user is concerned, the particular type of electronic switch used does not matter greatly, but we should briefly describe the main types used, so that the reader

Stack current/Amps

Figure 6.10 Graph summarising some data from a real 250 kW fuel cell used to power a bus (Derived from data in Spiegel et al. (1999).)

Stack current/Amps

Figure 6.10 Graph summarising some data from a real 250 kW fuel cell used to power a bus (Derived from data in Spiegel et al. (1999).)

Table 6.1 Key data for the main types of electronic switch used in modern power electronic equipment

Type

Thyristor

MOSFET

IGBT

symbol

Max. voltage (V) Max. current (A) Switching time (^s)

4500 4000 10-25

1700 600 1-4

c g s e has some understanding of their advantages and disadvantages. Table 6.1 shows the main characteristics of the most commonly used types.

The metal oxide semiconductor field effect transistor (MOSFET) is turned on by applying a voltage, usually between 5 and 10 V, to the gate. When 'on', the resistance between the drain (d) and source (s) is very low. The power required to ensure a very low resistance is small, as the current into the gate is low. However, the gate does have a considerable capacitance, so special drive circuits are usually used. The current path behaves like a resistor, whose 'on' value is RDSon . The value of RDSon for a MOSFET used in voltage regulation circuits can be as low as about 0.01 Q. However, such low values are only possible with devices that can switch low voltages, in the region of 50 V. Devices which can switch higher voltages have values of RDSon of about 0.1 Q, which causes higher losses. MOSFETs are widely used in low voltage systems of power less than about 1 kW.

The insulated gate bipolar transistor (IGBT) is essentially an integrated circuit combining a conventional bipolar transistor and a MOSFET, and it has the advantages of both. They require a fairly low voltage, with negligible current at the gate to turn on. The main current flow is from the collector to the emitter, and this path has the characteristics of a p-n junction. This means that the voltage does not rise much above 0.6 V at all current within the rating of the device. This makes it the preferred choice for systems where the current is greater than about 50 A. They can also be made to withstand higher voltages. The longer switching times compared to the MOSFET, as given in Table 6.1, are a disadvantage in lower power systems. However, the IGBT is now almost universally the electronic switch of choice in systems from 1 kW up to several hundred kW, with the upper limit rising each year.

The thyristor has been the electronic switch most commonly used in power electronics. Unlike the MOSFET and IGBT the thyristor can only be used as an electronic switch, it has no other applications. The transition from the blocking to the conducting state is triggered by a pulse of current into the gate. The device then remains in the conducting state until the current flowing through it falls to zero. This feature makes them particularly useful in circuits for rectifying AC, where they are still widely used. However, various variants of the thyristor, particularly the gate-turn-off, or GTO thyristor, can be switched off, even while a current is flowing, by the application of a negative current pulse to the gate.

Despite the fact that the switching is achieved by just a pulse of current, the energy needed to effect the switching is much greater than for the MOSFET or the IGBT. Furthermore, the switching times are markedly longer. The only advantage of the thyristor (in its various forms) for DC switching is that higher currents and voltages can be switched. However, the maximum power of IGBTs is now so high that this is very unlikely to be an issue in electric vehicle systems, which are usually below 1MW in power.4

Ultimately the component used for the electronic switch is not of great importance. As a result the circuit symbol used is often the 'device-independent' symbol shown in Figure 6.11. In use, it is essential that the switch moves as quickly as possible from the conducting to the blocking state, or vice versa. No energy is dissipated in the switch while it is open circuit, and only very little when it is fully on; it is while the transition takes place that the product of voltage and current is non-zero, and that power is lost.

6.2.2 Step-down or 'buck' regulators

The 'step-down' or 'buck' switching regulator (or chopper) is shown in Figure 6.12. The essential components are an electronic switch with an associated drive circuit, a diode and

Figure 6.11 Circuit symbol for a voltage operated electronic switch of any type

4 Electric railways locomotives would be an exception to this, but they are outside the scope of this book.

Inductor fYYY

Inductor fYYY

Current path when switch is ON

Current path when switch is ON

Inductor

Inductor

Current path when switch is OFF

Figure 6.12 Circuit diagram showing the operation of a switch mode step down regulator

Current path when switch is OFF

Figure 6.12 Circuit diagram showing the operation of a switch mode step down regulator an inductor. In Figure 6.12(a) the switch is on, and the current flows through the inductor and the load. The inductor produces a back EMF, making the current gradually rise. The switch is then turned off. The stored energy in the inductor keeps the current flowing through the load, using the diode, as in Figure 6.12(b). The different currents flowing during each part of this on-off cycle are shown in Figure 6.13. The voltage across the load can be further smoothed using capacitors if needed.

If V1 is the supply voltage, and the 'on' and 'off' times for the electronic switch are iON and iOFF, then it can be shown that the output voltage V2 is given by:

t0N + toFF

It is also clear that the ripple depends on the frequency: higher frequency, less ripple. However, each turn-on and turn-off involves the loss of some energy, so the frequency should not be too high. A control circuit is needed to adjust iON to achieve the desired output voltage; such circuits are readily available from many manufacturers. The main energy losses in the step-down chopper circuit are:

• switching losses in the electronic switch;

• power lost in the switch while on (0.6 x I for an IGBT, or RdSon x I2 for a MOSFET);

 Off ^On"""''' Off Off

Current supplied by fuel cell during the ON time

Current supplied by fuel cell during the ON time

Current circulating through diode during the OFF time

Current through load, being the sum of these two components

Figure 6.13 Currents in the step down switch mode regulator circuit

Current through load, being the sum of these two components

Figure 6.13 Currents in the step down switch mode regulator circuit

• power lost due to the resistance of the inductor;

In practice all these can be made very low. The efficiency of such a step-down chopper circuit should be over 90%. In higher voltage systems, about 100 V or more, efficiencies as high as 98% are possible.

We should at this point briefly mention the 'linear' regulator circuit. The principle is shown in Figure 6.14. A transistor is used again, but this time it is not switched fully on or fully off. Rather, the gate voltage is adjusted so that its resistance is at the correct value to drop the voltage to the desired value. This resistance will vary continuously, depending on the load current and the supply voltage. This type of circuit is widely used in small electronic systems, but should never be used with traction motors. The voltage is dropped by simply converting the surplus voltage into heat. Linear regulators have no place in systems where efficiency is paramount, such as an electric vehicle.

6.2.3 Step-up or 'boost' switching regulator

It is often desirable to step-up or boost a DC voltage, regenerative braking being just one example. This can also be done quite simply and efficiently using switching circuits.

"Electronic resistor"

"Electronic resistor"

lil

r

■4-

Control circuit

Figure 6.14 Linear regulator circuit

Inductor

JYY\

Figure 6.14 Linear regulator circuit

Inductor

JYY\

Current paths while switch is ON

Current paths while switch is ON

Inductor

STf\

Inductor

STf\

Current flow while switch is OFF

Figure 6.15 Circuit diagram to show the operation of a switch mode boost regulator

Current flow while switch is OFF

Figure 6.15 Circuit diagram to show the operation of a switch mode boost regulator

The circuit of Figure 6.15 is the basis usually used. We start our explanation by assuming some charge is in the capacitor. In Figure 6.15(a) the switch is on, and an electric current is building up in the inductor. The load is supplied by the capacitor discharging. The diode prevents the charge from the capacitor flowing back through the switch. In Figure 6.15(b) the switch is off. The inductor voltage rises sharply, because the current is falling. As soon as the voltage rises above that of the capacitor (plus about 0.6 V for the diode) the current will flow through the diode, and charge up the capacitor and flow through the load. This will continue so long as there is still energy in the inductor. The switch is then closed again, as in Figure 6.15(a), and the inductor re-energised while the capacitor supplies the load.

Higher voltages are achieved by having the switch off for a short time. It can be shown that for an ideal convertor with no losses:

t0FF

In practice the output voltage is somewhat less than this. As with the step-down (buck) switcher, control circuits for such boost or step-up switching regulators are readily available from many manufacturers.

The losses in this circuit come from the same sources as for the step-down regulator. However, because the currents through the inductor and switch are higher than the output current, the losses are higher. Also, all the charge passes through the diode this time, and so is subject to the 0.6 V drop and hence energy loss. The result is that the efficiency of these boost regulators is somewhat less than for the buck. Nevertheless, over 80% should normally be obtained, and in systems where the initial voltage is higher (over 100 V), efficiencies of 95% or more are possible.

For the regulation of fuel cell voltages, in cases where a small variation in output voltage can be tolerated, an up-chopper circuit is used at higher currents only. This is illustrated in Figure 6.16. At lower currents the voltage is not regulated. The circuit of Figure 6.15 is used, with the switch permanently off. However, the converter starts operating when the fuel cell voltage falls below a set value. Since the voltage shift is quite small, the efficiency would be higher.

Voltage

Voltage range

Output of step-up converter

Step-up converter kicks in here

Voltage of fuel cell

### Current

Figure 6.16 Graph of voltage against current for a fuel cell with a step-up chopper circuit that regulates to a voltage a little less than the maximum stack voltage

It should be pointed out that, of course, the current out from a step-up converter is less than the current in. In Figure 6.16, if the fuel cell is operating at point A, the output will be at point A', a higher voltage but a lower current. Also, the system is not entirely loss-free while the converter is not working. The current would all flow through the inductor and the diode, resulting in some loss of energy.

These step-up and step-down switcher or chopper circuits are called DC/DC converters. Complete units, ready made and ruggedly packaged are available as off-the-shelf units in a wide range of powers and input and output voltages. However, when they are used as motor controller circuits, as in the case of electric vehicles, the requirements of having to produce a variable voltage, or a fixed output voltage for a variable input voltage (as in the case when braking a motor using regenerative braking), then such off-the-shelf units will not often be suitable. In such cases special circuits must be designed, and most motors can be supplied with suitable controllers. As we have seen, the circuits required are, in principle, quite simple. The key is to properly control the switching of an electronic switch. This control is usually provided by a microprocessor.

### 6.2.4 Single-phase inverters

The circuits of the previous two sections are the basis of controlling the classical DC motor. However, the motors to be considered in the next section require alternating current (AC). The circuit that produces AC from DC sources such as batteries and fuel cells is known as an inverter. We will begin with the single phase inverter.

The arrangement of the key components of single phase inverter is shown in Figure 6.17. There are four electronic switches, labelled A, B, C and D, connected in what is called an H-bridge. Across each switch is a diode, whose purpose will become clear later. The load through which the AC is to be driven is represented by a resistor and an inductor.

The basic operation of the inverter is quite simple. First switches A and D are turned on, and a current flows to the right through the load. These two switches are then turned off; at this point we see the need for the diodes. The load will probably have some

Figure 6.17 H-bridge inverter circuit for producing single phase alternating current

Figure 6.18 Current/time graph for a square wave switched single-phase inverter

inductance, and so the current will not be able to stop immediately, but will continue to flow in the same direction, through the diodes across switches B and C, back into the supply. The switches B and C are then turned on, and a current flows in the opposite direction, to the left. When these switches turn off, the current 'free-wheels' on through the diodes in parallel with switches A and D.

The resulting current waveform is shown in Figure 6.18. The fact that it is very far from a sine wave may be a problem in some cases, which we will consider here.

The difference between a pure sine wave and any other waveform is expressed using the idea of harmonics. These are sinusoidal oscillations of voltage or current whose frequency fv is a whole number multiple of the fundamental oscillation frequency. It can be shown that any periodic waveform of any shape can be represented by the addition of harmonics to a fundamental sine wave. The process of finding these harmonics is known as Fourier analysis. For example, it can be shown that a square wave of frequency f can be expressed by the equation:

1111

v = sin(<wi) — - sin(3&>i) + — sin(5&>i) — — sin(7&>i) + — sin(9&>i)...

So the difference between a voltage or current waveform and a pure sine wave may be expressed in terms of higher frequency harmonics imposed on the fundamental frequency.

In mains-connected equipment these harmonics can cause a wide range of problems, but that is not our concern here. With motors the main problem is that the harmonics can increase the iron losses mentioned back in Section 6.1.5. We saw that these iron losses are proportional to the frequency of the change of the magnetic field. If our AC is being used to produce a changing magnetic field (which it nearly always will be) then the real rate of change, and hence the losses, will be noticeably increased by these higher harmonic frequencies. For this reason the simple switching pattern just described is often not used,

Figure 6.19 Pulse width modulation switching sequence for producing an approximately sinusoidal alternating current from the circuit of Figure 6.17

in favour of a more complex system that produces a more smoothly changing current pattern. This method is known as pulse width modulation.

The principle of pulse width modulation is shown in Figure 6.19. The same circuit as shown in Figure 6.17 is used. In the positive cycle only switch D is on all the time, and switch A is on intermittently. When A is on, current builds up in the load. When A is off, the current continues to flow, because of the load inductance, through switch D and the free-wheeling diode in parallel with switch C, around the bottom right loop of the circuit.

In the negative cycle a similar process occurs, except that switch B is on all the time, and switch C is 'pulsed'. When C is on, current builds in the load; when off, it continues to flow (though declining) through the upper loop in the circuit, and through the diode in parallel with switch A.

The precise shape of the waveform will depend on the nature (resistance, inductance, capacitance) of the load, but a typical half-cycle is shown in Figure 6.20. The waveform is still not a sine wave, but is a lot closer than that of Figure 6.18. Clearly, the more pulses there are in each cycle, the closer will be the wave to a pure sine wave, and the

weaker will be the harmonics. Twelve pulses per cycle is a commonly used standard, and generally this gives satisfactory results. In modern circuits the switching pulses are generated by microprocessor circuits.

### 6.2.5 Three-phase

Most large motors, of the type used in electric vehicles, have three sets of coils rather than just one. For these systems, as well as for regular mains systems, a three-phase AC supply is needed.

This is only a little more complicated than single-phase. The basic circuit is shown in Figure 6.21. Six switches, with free-wheeling diodes, are connected to the three-phase transformer on the right. The way in which these switches are used to generate three similar but out of phase voltages is shown in Figure 6.22. Each cycle can be divided into six steps. The graphs of Figure 6.23 show how the current in each of the three phases changes with time using this simple arrangement. These curves are obviously

Figure 6.21 Three-phase inverter circuit

## Electric Car Craze

Electric Cars Are A Savior From Pollution And Can Be A Laudable Investment For You. Find Out How Electric Car Conversion Shrinks Your Driving Expenses.

Get My Free Ebook