• head_banner_01

Inside a hybrid inverter combining SiC MOSFET and IGBT

Inside a hybrid inverter combining SiC MOSFET and IGBT

        After decades of silicon’s dominance, silicon carbide (SiC) is replacing silicon as the gold standard for high-voltage power electronics, including the traction inverters that form the basis of electric vehicles (EVs). In addition to being one of the key building blocks of electric vehicles, wide bandgap semiconductors play a key role in power inverters for solar, wind and other renewable energy systems, as well as power distribution in data centers.
        The driving force behind this change is the fact that SiC MOSFETs can reduce switching power losses and occupy less space than silicon IGBTs at the same voltage rating. The disadvantage is that they are more expensive than IGBTs. Power electronics company executives warn they are being pressured by automakers and other customers to cut prices on silicon carbide power devices as quickly as possible, and in some cases even sooner.
        STMicroelectronics is trying to solve this problem with a hybrid switching architecture that uses SiC power FETs in parallel with trench gate IGBTs to reduce costs without affecting the high efficiency of the inverter. The company is in the early stages of testing this concept and is also exploring various other innovations for electric vehicle powertrains. But as SiC MOSFET prices fall over time, hybrid inverters may be a temporary solution.
       The company promoted this concept at APEC 2024, where it demonstrated a galvanically isolated gate driver that can be used with a hybrid power stage consisting of a SiC MOSFET and an IGBT.
        The traction inverter is the basic building block of electric vehicles as it converts the high DC voltage from the battery into AC voltage, which controls the torque and speed of the electric motor. Energy conversion efficiency directly affects the range of electric vehicles.
        The inverter is based on power transistors (usually SiC field-effect transistors or IGBTs) that control the flow of electricity from the high-voltage battery to the electric vehicle’s motor. These devices must be protected throughout their operation, so their temperature, voltage and current must be carefully monitored. The MCU in the system is used to provide pulse width modulation (PWM) signals that control the high and low side of the power inverter.
        In electric vehicles, the inverter uses three-phase power transistors. They are assembled into multiple power stages driven by isolated gate drivers that act as an interface between the microcontroller and the power field-effect transistor, or IGBT. Each power stage is connected to a high-voltage DC bus, decoupled by several large capacitors to output a three-phase AC signal that drives the electric vehicle’s engine. The topology of the power stage usually depends on the system output power requirements.
        In a half-bridge circuit, the phases are assembled from a pair of power switches. The full bridge topology consists of four power switches and is suitable for higher power levels from 10 to 100 kW or more. Managing such large amounts of energy in the limited space of modern electric vehicles or data centers is not easy. Generally speaking, IGBTs must be physically larger to handle the higher voltages used in electric vehicles and other power systems.
        One of the most valuable properties of SiC is that it can withstand higher breakdown voltages than silicon, ranging from hundreds to thousands of volts. Because SiC MOSFETs are more reliable, they can handle the same voltages as silicon IGBTs in a smaller space.
        These chips also have higher electron mobility in the channel (the area under the MOSFET gate that transfers current from source to drain). Therefore, they turn on and off much faster than IGBTs, thereby saving energy. In addition to superior power characteristics, SiC power FETs have higher thermal conductivity than IGBTs, allowing for more efficient cooling. They can also operate safely at higher operating temperatures.
        However, the cost of SiC power field effect transistors is higher. These chips lack the enormous economies of scale of silicon IGBTs and are more prone to defects during the manufacturing process, driving up the price of defect-free chips.
        ST said it is testing a hybrid architecture using SiC MOSFETs and IGBTs, rather than using the same power devices in all aspects of the inverter. In its configuration, the inverter uses a pair of full bridges per phase. Each phase can output up to 10 kW, equivalent to a total power of 30 kW. Inside a full bridge, the transistors are arranged in a 3:1 ratio, with each SiC MOSFET corresponding to three IGBTs. IGBTs are placed in parallel with MOSFETs to handle large peak currents.
       It’s important to note that ST specifies that the power transistors are rated up to 1200 V. As a result, they can safely handle bus voltages up to 800 V, which is becoming the industry standard for EV batteries.
        But more important is the switching process of the power stage, which changes depending on the load conditions. With a light load, the power SiC transistor turns on earlier than the IGBT connected in parallel to it. Since SiC MOSFETs have lower conduction and switching losses than silicon IGBTs, their inclusion primarily improves the energy efficiency of the inverter. This is useful when low load current is the normal operating condition. For example, most uninterruptible power supplies (UPS) in data centers operate at 20-40% of full load.
        Use the same switching sequence at moderate loads to minimize the power loss encountered in the hybrid switch. When the SiC power transistor turns on, the voltage flowing through the device drops. ST claims that the sudden voltage drop causes the IGBT connected in parallel with the SiC MOSFET to achieve zero voltage switching (ZVS) (also known as soft switching). Therefore, most of the switching losses occur in SiC MOSFETs, whose switching losses are significantly lower than those of IGBTs at the same current.
       Subsequently, during the switching process, the SiC MOSFET in the hybrid power stage is first turned off by hard switching, and then the IGBT can be turned off by soft switching.
        When the current flowing into the load approaches its peak value, the switching sequence is reversed. IGBTs share current with SiC power transistors, which are susceptible to permanent damage under short-circuit and overcurrent conditions. At high loads, when the current exceeds the safe operating area (SOA) of the SiC MOSFET, the IGBT turns on first. To keep switching losses to a minimum, the SiC MOSFET is then turned on and off via soft switching.
       The company says that by replacing all but one of the SiC MOSFETs in the power stage with IGBTs, the hybrid inverter reduces costs while losing only a small amount of energy efficiency (at least at light loads).
        ST said preliminary testing shows that the hybrid power stage only loses about 0.5% energy efficiency at light loads compared to a power stage based entirely on SiC MOSFETs. The company said it has started focusing on light loads because the inverter is expected to spend most of its time in such conditions.
        Unlike switching multiple identical power transistors in parallel, switching hybrid power devices is a more complex process. Because power SiC MOSFETs switch at a higher frequency than IGBTs and the switching sequence varies with load current, the on and off phases of the SiC MOSFET or IGBT must be delayed slightly to allow the power transistors to switch. synchronously.
        At APEC, ST demonstrated a monolithic galvanically isolated gate driver specifically designed to drive the IGBTs and SiC MOSFETs at the heart of these hybrid switches. STGAP can control turn-on and turn-off timing to ensure that the first power transistor turned on or off using hard switching is fully turned on or off before other power transistors in the topology are turned on or off using hard switching. Soft switching process.
       Gate drivers can independently drive MOSFETs and IGBTs to reduce unnecessary power loss and prevent unintended consequences, including short circuits that could cause permanent damage to the power supply.
        Designed with a dual output architecture, STGAP is capable of accepting and delivering current up to 0.6A and delivering current of 2.5A from one output pin, while the other output pin can accept and deliver current of 2.4A and 0.6A. A. source. The current used to drive the power field effect transistor or IGBT determines the time required to turn on and off. The longer the transition between on and off states takes (called “dead time”), the more energy is lost during the switching process.
        ST said the gate driver is designed to be placed in front of a separate MOSFET and increase the current to control the turn-on and turn-off timing of the SiC MOSFET. Compared with IGBTs, SiC MOSFETs have special control and control requirements. It also integrates an independent isolated flyback power supply controller into the system, which can provide positive and negative voltage to turn on and off the power field effect transistor or IGBT.
       Gate drivers can be programmed via the SPI bus, making them flexible for use in a variety of applications, from traction inverters and electric vehicle charging stations to solar inverters or data center power supplies.
        While hybrid switches have the potential to be a valuable tool in the power electronics toolbox, they may not be a long-term solution. Silicon carbide is becoming one of the key building blocks of electric vehicle powertrains as it provides longer range and faster charging times. ST said that as economies of scale reduce the cost of power SiC FETs, all-SiC designs are likely to remain the first choice in the long term.
       Check out our coverage of APEC 2024. Also read more articles on TechXchange: Silicon Carbide (SiC).
        James Morra is a senior editor at Electronic Design magazine, covering the semiconductor industry and emerging technology trends. He also talks about the business behind electrical engineering, including the electronics supply chain. He joined Electronic Design in 2015 and lives in Austin, Texas.


Post time: Jul-08-2024