Enabling an efficient and dynamic electric vehicle charging network

  • Aug 27, 2018
  • UnitedSiC

By Zhongda Li, NPI Manager at UnitedSiC

The shift to electric traction for vehicles is well under way, with companies such as BMW, Nissan and Tesla all marketing competent cars. The advantages of such vehicles, such as lower pollution at point of use and lower operating costs, come at the cost of an increased burden on the electricity generation and distribution network. In the short term, power companies are adapting to rising demand by installing more generation capacity and reinforcing their grids. In the mid term, though, the distributed storage capacity embodied in a large cohort of electric vehicles could enable grids to get smarter, with spare battery capacity being used to improve the balance of grid supply and demand. Before this can happen, however, we need to build a new generation of efficient in-vehicle chargers.

Building better charging circuits

Electricity grids distribute energy as alternating currents (AC), but batteries store and release that energy as direct current (DC). Converting between the two forms of energy delivery demands the use of circuit topologies such as half- and full-bridge rectifiers and ‘totem pole’ configurations. Meeting the charging time requirements and efficiency requirements for in-vehicle chargers means implementing power factor correction (PFC) circuitry to enable the most effective transfer of energy to and from the battery.

The evolution of power factor correction

Switched-mode AC-DC converters rectify mains voltage, then chop the resultant high-voltage DC at high frequency so it can be converted by a transformer to lower voltages. The current drawn from the mains, though, is far from sinusoidal, because the bulk DC smoothing capacitors used in the converters only get ‘topped up’ with charge at the peak of the mains voltage. The result is a poor power factor – that is, the converter circuits are hard to drive efficiently from the mains. Putting 50 or 60 Hz inductors in series with these converters offers simple PFC, at the cost of increased losses and larger circuit boards (inductors can get big). The standard alternative approach to fixing this problem is to use an active switching circuit that takes a rectified mains voltage and boosts it to a fixed DC level, while controlling the line current to be sinusoidal (see the upper part of Figure 1). This approach can lead to losses at high power ­– the ­circuit topology puts three diodes (D1, D4, D5 or D2, D3, D5 – depending on the mains polarity) into the current path, each of which drops voltage and dissipates energy.

Replacing diodes with MOSFETs

If D5 is replaced with a similar switch to Q1, the circuit can be simplified so that Q1 and Q2 act as a boost switch and synchronous rectifier, swapping their function on alternate mains polarities (see the middle of Figure 1). Now there is only one diode and the drain-source resistance of a switch in series with the current, which cuts the conduction losses. The last step, shown at the bottom of Figure 1, is to replace D1 and D2 with synchronous switches for further efficiency gains.

Figure 1: The evolution of PFC circuits

Managing the impact of body diode behavior

As usual in circuit design, there’s a catch: Q1 and Q2 must always have a ‘dead time’ when neither of them conducts so that the circuit doesn’t suffer catastrophic ‘shoot-through’ currents. During this dead time, the inherent body diode of the MOSFET acting as a rectifier (either Q1 or Q2, depending on the phase of the switching cycle) conducts the full output current. When the device is reverse-biased in the next phase of the switching cycle, a large reverse-recovery current flows, causing power dissipation and electromagnetic interference, and canceling the efficiency gains made. High-voltage MOSFETs can have particularly poor body-diode reverse-recovery characteristics and so bridgeless totem-pole circuits have not been widely used at higher powers.

Applying wideband-gap switches

The availability of wide band-gap switches is making new circuit choices possible. SiC MOSFETs promise low channel conduction losses, high speed and a fast body diode. However, the forward voltage of the diode can be 2.5 – 3 V, leading to high conduction losses. The energy stored (EOSS) in the device capacitance is also typically twice the value for an equivalent Si MOSFET, which adds to the switching losses. Some designers are using enhancement-mode GaN devices, because they don’t have a body diode. However, the ratio of their normalized die area to ON-resistance RDSA is twice that of SiC MOSFETs. They also lack avalanche and short-circuit ratings, so their practical reliability is a concern. Both SiC MOSFETs and enhancement-mode GaN devices also have critical gate-drive voltages, to ensure reliable and efficient operation.

The SiC cascode option

One way to take advantage of the advantages of wide band-gap devices is to use SiC cascodes. These put a high-voltage SiC JFET and a high-performance, low-voltage Si MOSFET in one package. They are normally-OFF devices with defined avalanche and short-circuit ratings. They have low switching losses, due to extremely low input, output and Miller capacitances and stored energy ratings, which all stems from their small die sizes. A SiC cascode has a ratio of normalized die size to On resistance that is about three to four times better than an enhancement-mode GaN or SiC MOSFET, and about ten times better than Si super-junction MOSFETs.

The Si MOSFET in the SiC cascode arrangement introduces a body diode but, because it is a low-voltage MOSFET, the diode can be extremely fast, enabling a low reverse-recovery current and loss. Figure 2 compares the recovery characteristics of a 650 V UnitedSiC UJC06505T and a 650 V IPP65R045C7 silicon super-junction MOSFET, showing about a sixtyfold difference in recovered charge.

Figure 2. Comparing the reverse recovery characteristics of a SiC cascode and a Si MOSFET

SiC cascodes are also easy to drive, with operating levels of typically 0 – 12 V and absolute maximums of +/-25 V.

SiC cascodes in practice

We can explore circuit topology and device characteristics to our hearts’ content, but how good are SiC cascodes in our target application, an onboard electric vehicle charger?

One way of assessing this is to see whether a complete converter can meet the 80PLUS Titanium efficiency standard. This mandates that the converter should exceed 96 percent efficiency at high line and half load. Meeting this standard demands that, if the mains-conversion stage reaches a challenging but possible 97.5 percent efficiency, the efficiency of the PFC stage has to exceed 98.5 percent. It turns out that a 1.5 kW demo board from UnitedSiC, which uses UJC06505K SiC cascodes running at 100 kHz, can meet this requirement with some margin (shown in Figure 3).

Figure 3: Using SiC cascodes and totem-pole PFC to achieve high conversion efficiencies

Enabling a two-way grid with EV chargers

And so we are back to our original premise – that the rise of electrical vehicles is going to create an opportunity for electricity grids to take advantage of many, widely distributed energy storage units (i.e. the vehicle batteries) to sink as well as source power.

Making this a reality will take imagination – and onboard vehicle chargers that can support energy transfer in either direction. The totem-pole PFC stage described above could make this possible. If you look again at the lower circuit diagram in Figure 1 and imagine that the DC bus is the power source and the AC line connection is the load, then if the MOSFETs are driven appropriately the circuit becomes a half-bridge inverter that will turn DC battery power back into an AC feed.

As electric vehicle usage rises, the use of SiC cascodes for PFC will boost the efficiency of onboard chargers. Their ruggedness and adaptability may also mean that they will be used to improve the capacity and efficiency of the electricity grid itself.