Let’s continue our exploration of the components used in power electronics, with an overlook at those that are employed as switches in the switching regime while using some SPICE simulations to observe their general behavior.
The switching speed, maximum tolerable voltages and currents, and, most importantly, the reduction of the Rds(on) parameter are just a few examples of how the most recent models have improved.
The bipolar transistors observed in the last episode have the drawback of presenting too high switching times, especially at high powers. In this way, they do not guarantee good saturation, so the switching losses are unacceptable. This problem has been greatly reduced thanks to the adoption of “field effect” technology, with switching devices known as Power-mos, or field-effect power transistors. In any case, the most commonly used name to indicate this type of component is that of MOSFET. Power MOSFETs are normally N-channel devices, capable of withstanding voltages of hundreds of volts and currents of tens of amperes. They are biased with a positive VDS voltage, but in the absence of it, only a small leakage current passes through the P-N junction. By adjusting the VGS voltage, it is possible to control the width of the conductive channel and the equivalent RDS resistance of the device, which ranges from very high values (Rds(off)) to very low values (Rds(on)). They are characterized by a high switching speed, with on and off times in the order of tens of nanoseconds, making them hundreds of times faster than BJTs. The terminals are, therefore, the gate, the drain, and the source. The gate is made of polysilicon and is isolated from the whole device by a thin oxide layer. Usually, a free-wheeling diode is inserted in the same device, placed between the drain and source terminals. Figure 1 shows the power supply principle diagram of a MOSFET, with the relative current responses by varying the voltage VGD and the voltage VDS. The MOSFET model used is, in this example, the IRF530, whose SPICE model is as follows:
.model IRF530 VDMOS(Rg=3 Vto=4 Rd=50m Rs=12m Rb=60m Kp=5 lambda=.01 Cgdmax=1n Cgdmin=.26n Cgs=.2n Cjo=.4n Is=52p ksubthres=.1 mfg=International_Rectifier Vds=100 Ron=160m Qg=26n)
If the VDS voltage exceeds the maximum allowed, the current increases sharply and causes an immediate breakdown of the device. MOSFETs behave like variable resistors controlled by the control voltage applied to the gate. When the control voltage exceeds certain values, the Rds(on) parameter is very low, vice versa, if this voltage is zero, the Rds(off) parameter is very high and does not allow any current to flow. The MOSFET has another advantage over the BJT. The resistance between drain and source increases with the increase in temperature, thus limiting the amount of current in transit. In this way, the “avalanche effect” typical of transistors does not occur and the device is not destroyed. The component is very fast in switching and, thanks to its numerous advantages, it can be easily connected in parallel with other units.
It is very simple to calculate the Rds(on) and Rds(off) values of the IRF530 device, respectively, with a static gate voltage of 20 V and 0 V. The first is calculated as follows:
This is, as you can see, an extremely low resistance, probably even lower than that of the same connections, cables and PCB. In this way the thermal dissipation of the device is reduced to a minimum, although the current transit is important. The second is calculated as follows:
The DS channel, in practice, is an open circuit through which only a minimum leakage current passes, on the order of picoamperes.
IGBTs (Insulated Gate Bipolar Transistors) are still widely used as switching devices in power circuits. Converters, inverters, and motor drivers heavily use this type of component. The IGBT is a semiconductor device with four alternating layers (P-N-P-N) and is controlled by the metal oxide semiconductor gate (MOS). Furthermore, the IGBT has a single PN junction. Practically, they are hybrid devices between bipolar transistors and power MOSFETs and can withstand higher voltages and currents, even higher than 1000 V and 1000 A. In practice, these components exploit the advantages of BJT and MOSFET technology. They allow obtaining a low channel resistance even though they are characterized by a high breakdown voltage at the expense, however, of the switching speed. IGBTs are created using one or more BJTs with a VDMOS type field effect transistor. In this way, it has a very high input impedance and a behavior similar to that of the BJT as regards the transit of current, both in conduction and in switching. The terminals of an IGBT device are as follows:
- gate (the control terminal);
Normally, the current between collector and emitter (positive VCE bias) is controlled by acting on the gate with a positive voltage VGE greater than the minimum threshold voltage. Unfortunately, the device is not very fast at switching. Figure 2 shows an example of a practical application of PWM switching through the IKW30N65EL5 IGBT device, the main characteristics of which are as follows:
- collector-emitter voltage (Vce): 650 V;
- DC collector current (Ic): 85 A;
- pulsed collector current: 120 A;
- gate-emitter voltage (Vge): +/- 30 V;
- power dissipation (Ptot): 227 W;
- operating junction temperature (Tvj): between -40° C and +175° C.
The graph in the figure is in the gate voltage domain and is formed by the following curves:
- VGE voltage (Vgate), between 0 V and 30 V. It is this voltage that drives the gate and triggers the conduction of the device. The conduction threshold, in this case, is greater than 8 V;
- collector current (Ic). As you can see, in the Vg range between 6 V and 8 V, the device regulates the current, just like a potentiometer, and it is in the linear area;
- power dissipated by the device (Ptot): this curve represents the worst case of operation, as both the voltage and the current are at very high levels, causing an exponential increase in the dissipated power. In generic applications, it is necessary to avoid making the IGBT work in this condition;
- drain voltage (Vd): it is the voltage present on this terminal. If the device is deactivated and open, the voltage is equal to the value of VCC. If it is activated and in the ON state, the voltage drops to minimum levels.
- efficiency: in static saturation conditions, the device has an efficiency close to 100%.
Silicon (Si) based power electronics has long dominated the power electronics industry. Due to its important advantages, silicon carbide (SiC) has gained a lot of space in the market in recent times. With the implementation of new materials, electronic switches have significantly improved their static and dynamic electrical characteristics. An ideal switch has the following characteristics:
- has an infinite switching speed;
- can pass high currents without voltage drops;
- can handle high voltages;
- its current passage channel (usually D-S) has zero resistance;
- it does not cause energy losses in the transition between the two logic states.
Silicon does not allow superior performance and devices made with this material do not exhibit high efficiency. The SiC MOSFETs combine all the characteristics of an almost ideal switch, allowing you to operate with very high-performing devices. Its main advantages include increased efficiency and reliability, reduced thermal problems, and a reduced physical footprint. Thanks to the reduced switching losses, the final system can work at a lower temperature, allowing for a lighter and more economical circuit since the thermal conductivity of SiC is much greater than that of silicon. In other words, a low-power SiC system can replace a higher-power silicon system with the same performance. Furthermore, the switching frequencies can be significantly elevated, allowing large reductions in the size of the circuits. SiC devices can work up to temperatures of 175 ° C. The characteristics of SiC Mosfets do not vary much according to temperatures and currents (as happens, however, with silicon). Thanks to all these advantages, SiC Mosfets are used in a variety of applications:
the transport of energy, in anticipation of the use of high voltage DC;
e-mobility, to drive electric car engines and battery recharging circuits;
railway sector, for driving motors with a power of millions of Watts;
photovoltaic sector: for driving loads and for recharging accumulators.
Figure 3 shows a typical use of activating and deactivating a SiC MOSFET by means of a PWM signal. The SiC MOSFET used is the UF3C065080T3S model, with the following basic characteristics:
- package: TO-220-3L;
- drain-source voltage (VDS): 650 V;
- gate-source voltage (VGS): -25° C to +25° C;
- continuous drain current (ID): 31 A;
- pulsed drain current (IDM): 65 A;
- power dissipation (Ptot): 190 W;
- maximum junction temperature (Tjmax): 175° C.
The example compares it with a BJT power transistor. When the two electronic switches are activated, a current of about 4.8 A passes through the load. The driving frequency is quite high, at about 100 kHz. It is interesting to note that, at each signal period, the activation of the SiC MOSFET occurs in only 30 nanoseconds, while the saturation of the BJT occurs in about 500 nanoseconds, an unacceptable time for this type of application. Precisely for this reason, BJTs have been abandoned in high frequency power solutions. The high speed of the device allows its low dissipation. The solution with SiC MOSFET, in fact, dissipates on average a power of 1 Watt, while the solution with BJT dissipates on average a power of 12 Watts.
Gallium nitride is a semiconductor material with a direct band gap and its most important characteristics are the ability to handle very high voltages at high temperatures. These types of devices ensure greater efficiency and fewer switching losses in switching applications. Gallium Nitride offers better thermal conductivity, higher switching speeds, and allows for the construction of physically smaller devices than traditional silicon devices. In other words, there are low power losses during the charge and discharge cycles, and they take up less PCB space. With GaN MOSFETs, they increase the energy efficiency and reliability of the final solutions. GaN components are expected to radically change the world of power electronics, and the cost and reliability of electronic components made with the new semiconductor materials are increasingly approaching those of silicon components. A GaN device can be turned on and off much faster than other types of electronic switches. In fact, its average turn-on times are about 4 or 5 times shorter than traditional MOSFETs. GaN devices need a driver to ensure they turn on and off perfectly. To conduct a GaN device, it is always advisable to supply the gate terminal with its maximum tolerable voltage. In this way, the ON state is clear and decisive. An important advantage in the adoption of GaN-based devices is an important reduction in Rds(on), or the internal resistance of the device when it is in the conduction state. Furthermore, the large band gap improves performance up to higher temperatures than those of silicon, so much so that in recent years the number of applications using GaN MOSFETs is increasing exponentially. The following examples concern the EPC2032 model, a specimen equipped with some protuberances to allow welding and with very relevant characteristics, including:
- drain to source voltage (VDS, continuous): 100 V;
- drain to Source voltage (up to 10,000 5 ms pulses at 150˚ C): 120 V;
- continuous current (ID): 48 A;
- pulsed current: 340 A;
- drain to source On resistance (Rds(on)): 3 milliOhm;
- gate to source voltage (VGS): from -4 V to 6 V;
- very high switching frequency;
- operating temperature (TJ): from -40° C to +150° C.
A first observation concerns the determination of the Rds(on) of the device, according to the application scheme in static regime of figure 4. This resistance, in static regime, is extremely low (only 0.002853 Ohm) and allows almost zero dissipation of the electronic switch equal, in the example, to only 1.29 W, against a load that dissipates well 1997 Watts, with an equivalent efficiency of 99.94%.
Temperature always affects any electronic component. Fortunately, GaN devices are not very affected by thermal changes and although the Rds is relatively variable, the efficiency of the circuit is always very high. The two graphs in the figure show, respectively, the trend of the Rds parameter as a function of the voltage Vgs (top graph) and the junction temperature (bottom graph). The temperature coefficient of the Rds(on) is positive, i.e. it increases with increasing temperature.
In this article, we very broadly looked at some essential parts of power electronics. There are also other components on the market that combine the advantages of those previously seen and eliminate some of their negative aspects. Among these we can include, for example, GTOs and GCTs, special thyristors that can withstand voltages of many kV and currents of a few kA. They can be switched on and off via the gate terminal. Materials with a large band gap, such as GaN and SiC, now make it possible to lower design costs, also reducing the size of power solutions. The band-gap of a material depends on the strength of the chemical bonds between its atoms. And the new materials allow designers to achieve very important results in terms of system performance from all points of view.
You can find our Power Electronics Course series here.