Selecting passives for SiC topologies in electric vehicle chargers

The shift from powering vehicles with internal-combustion engines to using electric motors is well underway. Although Tesla held the limelight for a long time, traditional automobile makers such as Ford and Volvo are now bringing electric vehicles (EVs) to market to compete. Beyond personal transportation, electric versions of taxis, buses, and trucks are also in development or already on the road.

Although these vehicles usually offer all sorts of attractive features, such as advanced driver aids and electronic cockpits, consumers still cite vehicle range and charging rates as among the key concerns that stop them buying or leasing an EV. Until it’s as easy and quick to find a charger and top up a battery as it is to find a garage and fill up a tank, EVs will continue to be seen as a limited alternative to traditional vehicles.

This is understood by EV makers, who work hard to strike the right balance between cost, battery capacity, charging rates and efficient usage of stored energy. Infrastructure providers are also stepping up their game. Basic public chargers are proliferating, and Tesla is steadily increasing the charging power of its proprietary network. IONITY, a consortium including BMW, Daimler, Ford, and Volkswagen, is building a network of 350kW chargers along some of Europe’s major highways. Energy company E-On is cooperating with Danish mobility service provider CLEVER to build a similar north-south, ultra-fast charging network.

This may not be enough to properly service commercial vehicles such as buses and trucks, which will need much bigger batteries than automobiles before they can take over the work of their internal-combustion-engined forebears. Fortunately, the development of much more powerful charging solutions is already underway. CharIN, a global organisation that promotes a standard for charging vehicles, has been working on megawatt charging systems since 2018. It envisages the standard it is developing, based on the Combined Charging System, using a single plug to carry up to 1500V and 3000A, to enable commercially acceptable charging times for large vehicles.

There are three major ways of charging an EV: using an onboard system to rectify an AC input, such as household mains electricity, to charge the battery at relatively low power; using an offboard system to deliver large amounts of DC power conditioned for direct storage by the battery; and wirelessly, through inductive coupling between a coil mounted on a garage floor and a receiver under the car. Whichever approach is taken, the efficiency with which the input electricity is converted into a form that the battery can absorb is critical. A single percentage-point improvement in the efficiency with which a 350kW charger delivers its power means both an extra 3.5kW reaching the battery, and 3.5kW less waste heat to manage.

CoolSiC­­­™ Schottky diodes from Infineon

One increasingly important weapon in the EV power-system designer’s arsenal is the availability of key devices such Schottky diodes and MOSFETs built using wideband semiconductor materials. A wideband semiconductor such as silicon carbide (SiC) has a breakdown voltage ten times that of silicon (Si), which means it can be used to implement high-voltage circuits more simply than Si devices. The wide bandgap of SiC also means that the channel of any device built using the material can be more heavily doped than is possible in Si, and driven at a higher gate-source voltage, both of which reduce its resistance and so the conductance losses through it. SiC devices can also be driven with a higher gate-source voltage than Si equivalents, which further reduces the channel resistance.

SiC can operate effectively at higher temperatures (up to 200ºC) than Si, and has much greater thermal conductivity. This gives two advantages. The first is that SiC devices can run at much higher temperatures than Si equivalents without cooling, where this is practical, but if cooling is necessary, it takes less work to do it. This enables simpler, lighter, smaller, lower-cost cooling systems.

SiC devices can also be switched more quickly than Si equivalents, improving power conversion efficiency. It also means that passive components, such as the inductive and capacitive parts used for electromagnetic interference (EMI) control or as energy reservoirs in power-conversion circuits, can be smaller. Higher frequency switching also creates less powerful ripples in a converter’s output, simplifying DC conditioning circuitry.

SiC’s high conductance means that devices made with it can be smaller than Si devices of equivalent performance. This reduces the manufacturing cost of the SiC part, its self-inductance, and the capacitive coupling it experiences with its package. This all enhances SiC’s ability to switch at higher frequencies.

This combination of characteristics adds up to a number of systemic advantages for SiC, including simpler high-voltage circuitry than Si equivalents, higher temperature operation without cooling, and lower cooling effort when that is required. Higher switching speeds also mean that smaller inductances and capacitances are needed, which usually mean smaller component sizes.

These advantages enable designers to build smaller, denser, lighter, simpler EV charging systems that have greater charging efficiency and reliability than Si alternatives.

Given the expectations of growth in the market for EV chargers, companies such as Infineon have invested heavily in developing SiC production capacity, device architectures and circuit topologies, packaging technology, reference designs, and engineering support. The company makes CoolSiC-branded Schottky diodes and CoolMOS-branded MOSFETs, and also packages them in modules that employ patented technology to improve cooling efficiency.

The basic architecture of an EV charger involves three stages. In the first, an AC supply (usually three-phase) is filtered, undergoes power-factor correction and is then rectified to DC. This DC is then switched rapidly enough to excite a transformer whose turns ratio sets the output AC voltage to match the battery’s operating voltage. The AC voltage is then rectified and filtered again into a DC voltage suitable for charging the vehicle battery.

To simplify charger design, Infineon recommends that high-power chargers are built out of multiple lower-power modules working in parallel. To improve charger reliability, designs may be built with (n+1) modules to allow for redundancy. The table below shows Infineon’s recommendations for choosing between discrete and modular implementations, using Si or SiC parts, for charging powers of up to 350kW.

Recommendations on discrete and modular Si or SiC implementations (Source: Infineon)

Infineon offers many examples that show the advantages of using SiC devices in power conversion circuits. The DC-DC converter design below shows the benefits of SiC’s high operating voltage. The SiC design replaces 16 650V Si super-junction MOSFETs with eight 1200V SiC MOSFETs, halves the number of inductive components used, and doesn’t need three out of four of the discrete capacitances used in the Si design.

DC-DC converter design: Si to SiC comparison (Source: Infineon)

Working with SiC devices does make extra demands on the supporting passive components used, in terms of their size, operating voltages, frequencies, and temperatures. The block diagram and CAD rendering below show a reference design for a bidirectional EV charger. As you can see, the passive components such as the power-factor correction inductors, DC link capacitors, and EMI filters make up a substantial proportion of the volume of the overall design.

A block diagram and CAD rendering of a bidirectional EV charger (Source: Infineon)

KEMET makes a range of parts optimised for these sorts of demanding applications. Some of the key specifications for these parts include the equivalent series resistance (ESR) of a capacitor’s terminals and electrodes, which controls the amount of energy lost as heat when it is subjected to ripple currents. ESR decreases with increasing frequency. Similarly, inductors should be chosen to minimise their DC resistance to reduce self-heating and power loss. Designers also need to pay attention to the inductor’s core losses, which are created by the combination of hysteresis losses and eddy current losses. These can lead to self-heating, which undermines the inductor’s performance.

ESR of KC-LINK capacitors (Source: KEMET)

KEMET’s KC-LINK capacitors have been designed for fast-switching applications such as EV charging systems using SiC parts, and operate at higher voltages, temperatures, and frequencies than similar parts that have not been optimised for this application. The parts can operate at up to 150°C, and can be mounted close to fast-switching transistors in high power-density applications.

The graph to the right shows the very low and relatively stable ESR of KC-LINK capacitors over a wide operating frequency range, which makes it possible for them to operate at very high ripple currents with no change in capacitance with applied DC voltage. They also show little change in capacitance versus temperature.

KEMET also offers automotive-grade inductors for charging battery banks and implementing power trains. KEMET's METCOM MPXV1 metal composite power inductors will operate at up to 155°C, while the MPEV series operates up to 180°C. They have a higher saturation flux density than ferrite-cored inductors, which allows them to sustain a stronger magnetic field with low magnetic flux leakage and high saturation. These are ideal characteristics for use in designs that need stable inductance across temperature and current.

EVs are coming and the opportunities presented by building the necessary charging infrastructure are huge. Shifting to SiC semiconductors will enable smaller and denser charging units, which operate with greater efficiency than Si equivalents while requiring less cooling. But these advantages are only possible if the supporting passive components have been uprated to match.