It can be hard to tie together calls for energy savings in electronics when that means only a few cents cut in your electric bill or some small contribution to reducing global CO2 emissions, but when better efficiency is extracted from electric vehicles, the effects are more real – scope Better, lighter weight and lower operating costs. Now, advances in battery and power conversion technology in cars have made them so viable that sales of internal combustion engine (ICE) vehicles are set to be banned in some countries1 and most new vehicle development is focused on electric vehicles and their powertrain. In the search for the perfect switch, electric vehicles are filled with electronics that need power, from traction inverters to built-in chargers and auxiliary supplies. In all cases, to achieve high efficiency, transformer mode technologies are used to generate voltage bars, relying on semiconductors operating at high frequency. The ideal switch for application has a resistance close to zero when turned on, no leakage when turned off, and a high breakdown voltage (Fig. 1). When you switch between the two states, there should be little transient power dissipation and any residual losses should result in minimal switching overheating. Over the years, semiconductor technologies were introduced whose performance was closer to ideal, but expectations also changed and the search for the perfect switch continued. Figure 1: Perfect Switching of Candidates for the Perfect Switch Today’s switch selection is diverse: very high power IGBTs are preferred for low conduction losses while low to medium power MOSFETs dominate, with fast switching to reduce associated component size and cost, particularly magnetic. MOSFETs have traditionally used silicon technology, but now, silicon carbide can be used for its own benefits of low dynamism and plug-and-play losses at high temperatures. It’s a step closer to that elusive ideal switch, but there’s another way better than that – a SiC JFET jointly assembled with a low-voltage silicon MOSFET in code arrangement, together called a “SiC FET”. Simply put, Si MOSFET provides an easy and non-critical gate driver, while typically converting on JFETs to unnatural cascode with a host of advantages over Si or SiC MOSFETs. Figure 2 shows the basic construction of IGBT, planar SiC MOSFET, and JFET in a SiC FET, all in the 1200-V class. Figure 2: Construction of IGBT, SiC MOSFET, and SiC JFET (1,200-V class) It is clear from Figure 2 that the higher critical breakdown voltage of SiC in a MOSFET or JFET allows for a much thinner drift layer, about one-tenth of this for silicon in IGBT, with a resistance less interview. Silicon IGBTs achieve their low resistance by injecting a large number of carriers into the thick drift layer, and this results in 100× stored charge, which must be driven in and out of the drift layer at each switching cycle. This results in a relatively high switching loss and a large power requirement for the gate motor. SiC MOSFETs and JFETs are unipolar devices in which the charge movement is in and out of the device capacitances, resulting in much lower dynamic losses. Now comparing SiC FETs with SiC MOSFET, the electron movement in the channel is much better, as the SiC FET allows a much smaller template for the same resistance with lower capacitance and thus faster switching or lower resistance (RDS(ON)) same template region A. Therefore, A is A key metric and indicates the potential for more die per chip for a given performance and subsequent cost savings or less delivery loss for a given die area. COSS defines the interaction between the resistance and the output capacitance, which is switched for a given voltage rating to give a more or less shunt loss. The win-win situation of more dies per chip with faster switching, all other things being equal, is mitigated by the need to now remove heat from a smaller area. SiC has 3 times better thermal conductivity than silicon, which helps, and also has the ability to operate at higher and higher average temperatures, but to take advantage of these advantages, the latest generation of SiC FETs, “GEN 4”, features a thinning foil to reduce Its electrical and thermal resistance and silver sintering dies for 6 times better thermal conductivity than welding – the net effect is enhanced reliability due to lower junction temperatures by a wide margin to absolute maximum. The advantages of SiC FETs over SiC MOSFETs are broad and application dependent but can be summarized in a radar diagram of FOMs and their main characteristics (Fig. 3). Figure 3: Advantages of SiC FET in different applications. The plots are normalized with the properties of GEN 4 SiC FETs from UnitedSiC. The plots were normalized according to the properties of UnitedSiC GEN 4 SiC FETs, showing superior all-round performance at both high and low temperatures. Practical results confirm the promise of SiC FETs The efficacy of SiC FETs has been demonstrated by UnitedSiC with the design of a totempole PFC stage operating in continuous conduction mode with ‘steady’ switching, which would be typical of a front-end EV charger on board. The converter is rated at 3.6 kW with an 85- to 264-VAC input and a 390-VDC output using 18 mV or, alternatively, 60 m3 Generation 4 SiC FETs in TO-247-4L packages, switching at 60 kHz . The system efficiency plots are shown in Fig. 4, with a peak value of 99.37% achieved at 230 VAC, with an 18 m3 SiC FET for the high-frequency, high and low-side switch positions. At a full 3.6 kW output, these SiC FETs together dissipate 16 W or 0.44% inefficiency, requiring minimal heat dissipation. Figure 4: A totem-electrode PFC stage achieves an efficiency of 99.37% using SiC FETs. In electric vehicles, there is also a down-conversion stage with isolation from the traction battery voltage to 12 volts, usually implemented with the LLC transformer, which is currently the preferred structure for high efficiency. LLC transformers are resonantly switched at high frequency for optimum performance, and SiC FETs are, again, a good choice. At 3.6 kW, switched at 500 kHz, a pair of GEN 4 750-V 18 MOSFET modules show a dissipation of less than 6.5 watts, each including conduction, switching, and body diode loss. Traction transformers are where the most energy can be saved, and SiC FETs can replace IGBT for a real increase in efficiency. The switching frequency is usually kept low at 8 kHz, even with SiC devices, since the magnetic component is the motor, which does not directly shrink in size as the switching frequency of the inverter increases. For a significant gain, a single IGBT can be replaced in addition to a parallel diode, for example, six parallelograms 6 MΩ SiC FET resulting in a 1.6% increase in semiconductor efficiency to 99.36% at a 200 kW output, which is more From 3 x cut-off in power loss or 3 kW At light loads, in which vehicles typically operate, the improvement is even better, with losses 5x to 6x lower than IGBT technology – all with the advantages of very low gate drive power and no “knee” effort to better control light loads. Lower losses, of course, mean savings in size, weight, and cost of the heatsink as well as better range, so any additional cost for a semiconductor module will soon be overcome. Have we reached perfection? No semiconductor manufacturer would dare claim their switch is perfect, but now that the efficiency in converting power to decimal points has fallen above 99%, we’re getting closer. SiC FETs enable this, and you can try it out for yourself using the SiC FET-JET calculator tool on the UnitedSiC website, which calculates losses for a variety of AC/DC and DC/DC topologies. References 1 https://thedriven.io/2020/11/12/the-countries-and-states-leading-the-phase-out-of-fossil-fuel-cars/ 2https://info.unitedsic.com/ fet- jet Sign up for Roadmap to Next-Gen EV & AV Virtual Conferences to view this and other on-demand presentations. Written by Anup Bhalla, Vice President of Engineering, UnitedSiC Please visit the e-book for the full article.