By Tony Titre, EMEA Automotive Business Development Manager, Future Electronics &
David Woodcock, Centre of Excellence Manager (Power), Future Electronics
Today’s consumer has little patience for delay: smartphone apps such as Uber Eats, for instance, have taught users to expect restaurant meals at home on demand, delivered in minutes. And now Amazon has begun to experiment with the use of drones to find out whether there is a market for the home delivery of online purchases just 30 minutes after placing an order.
The expectation that consumers can have whatever they want, whenever they want it is at odds with the operation of today’s Electric Vehicle (EV) charging systems. If a fully electric vehicle’s battery is deeply discharged, a home charger with a typical power rating of 7kW will take around 5-8 hours to fully charge the battery of a mainstream vehicle such as a Renault Zoe or Nissan Leaf.
So now EV manufacturers have to grapple with consumers’ reluctance to wait any time for anything – let alone waiting a full working day before their vehicle is ready for a new journey. In fact, the car industry is working on the assumption that EV drivers will accept a charging time of no more than one hour (to reach 80% of maximum capacity from a deeply discharged state). This problem of fast charging at home or on the street has to be solved before EVs achieve mass market adoption (see Figure 1).
To date, however, the development effort applied to EV charger hardware for the home has been minimal. This is largely because most European homes have a single-phase mains power supply, and the maximum power capability of the single-phase AC-DC on-board charger inside an EV is around 7kW. In today’s low production volumes, the board design for a simple 7kW AC home charger can be assembled with little more than a few off-the-shelf modular components for the power circuit, controller and memory.
Fig. 1: most home and street-level EV chargers are today limited to a maximum power output of 7kW. (Image credit: Albert Lugosi under Creative Commons licence)
To reach the one-hour charging time benchmark, charger manufacturers will need to develop home chargers that operate from a three-phase mains input and provide a high-voltage DC output directly to the battery – bypassing the car’s on-board AC-DC charger – and that have a power capability of at least 22kW. High efficiency will be a crucial feature of these new home chargers to limit the cost of power losses, and to avoid excessive waste heat generation, easing system thermal management.
This raises three important issues for the industry and policy makers to consider.
1. Upgrading the Grid Infrastructure
While large buildings such as commercial office blocks or apartment complexes might have a high-power three-phase mains supply, most single dwellings in Europe have a single-phase input. The cost of installing a three-phase supply to a home will be considerable – some hundreds of euros per home. This is a big disincentive to purchase a new EV – so will EV manufacturers be expected to subsidise the cost of upgrading the power supply to the home of a new car buyer? And how will the purchaser of a used EV afford the installation of a three-phase supply? Or will government, or the electricity supply industry, be expected to cover the cost of upgrading to three-phase supplies to homes?
Capacity issues also pose a risk to the use of high-power home charging points. Today, the high-speed (>50kW) public charging points installed at highway service stations draw too much power to be supplied directly by the local grid, and instead draw power from local battery power storage.
It is not difficult to imagine that EV owners simultaneously plugging in 22kW charging points after the return journey home from work in the early evening could overload the local power grid transformer. It is also conceivable that market demand for EVs with longer range, requiring a battery capacity of 70kWh-100kWh, will induce car owners to install very high-power charging points rated at 50kW-100kW at home. The operation of such superpowered chargers might require battery back-up supplies either at home or as part of the local grid infrastructure.
2. Availability of Wide Bandgap Semiconductor Components
When converting a three-phase 240V AC input to a DC output at a battery’s voltage of between 200V and 500V DC and at power levels up to 22kW, the scope for losses in conventional silicon switches – MOSFETs or IGBTs – and power components such as diodes is considerable.
The use of wide bandgap semiconductor technology – using either Silicon Carbide (SiC) or Gallium Nitride (GaN) materials – promises a dramatic reduction in both switching and conduction losses compared to losses in equivalent silicon-based parts. Equally important, both SiC and GaN transistors support much higher switching frequencies than silicon equivalents, and this enables the use of smaller magnetic components and capacitors. In a high-voltage power converter using a low-speed silicon switch-based architecture, the large magnetic components can represent as much as half of the total bill-of-materials cost.
A reduction in the size and cost of the magnetics can therefore produce a lower total system cost, even taking into account the higher unit cost of a SiC or GaN component compared to its silicon equivalent.
The problem for any manufacturer which plans to enter the market for high-power home EV chargers is that, while SiC or GaN components’ attributes are highly attractive, supply is growing but constrained. Car manufacturers purchase much of the market’s capacity for use in EVs’ traction power systems. Suppliers of SiC components such as STMicroelectronics, ON Semiconductor, Littelfuse and Microchip are ramping up production fast, and supply will of course eventually balance with demand. Likewise, manufacturers are investing heavily in facilities to increase production of GaN High Electron-Mobility Transistor (HEMT) devices. But in the short term, the use of wide bandgap semiconductors in high-power charging points will require careful management of the supply chain. The shortage of SiC and GaN parts is a problem that a distributor with a global supply chain, such as Future Electronics, can do much to eliminate.
3. Availability of High-Power Systems Design Expertise
The third important challenge facing manufacturers eyeing the home charging-point market is assembling the specialist expertise required to implement a complex power-conversion architecture using GaN or SiC components, with which few seasoned power-system designers are familiar.
Fig. 2: schematic of a Vienna rectifier, which implements power factor correction in a typical unidirectional three-phase EV charger’s AC-DC converter circuit. (Image credit: Uwe Drofenik under Creative Commons licence)
The charging point designer is likely to use one of two architectures, depending on the application requirement. The charger might be unidirectional – that is, drawing power from the grid to the EV battery, but never supplying power in the reverse direction.
Three-phase unidirectional charging points will typically adopt a two-stage architecture consisting of a Vienna rectifier to provide an output of around 800V DC, and a step-down DC-DC converter secondary stage to produce a battery voltage in the range 200V-500V DC with a constant-current/constant-voltage charging profile (see Figure 2). Microchip provides an excellent reference design for a Vienna rectifier based on the use of SiC MOSFETs and diodes, providing Power Factor Correction (PFC) in EV chargers rated for up to 30kW power. The MSCSICPFC/REF5 design achieves 98.5% efficiency at a power output of 20kW .
The application might, however, require bi-directional capability, in which the vehicle battery can both sink power from the grid, and source it when required, for instance to support local grid-balancing schemes. Indeed, grid operators around the world are keen to explore the potential for EVs to smooth power fluctuations through the day by ‘lending’ power to the grid at times of high demand, and drawing it back at times of low demand.
A bi-directional charging point provides for this two-way flow of power, but is a more complex architecture than a unidirectional charging point. Topologies which will support efficient bi-directional conversion lend themselves well to the use of GaN or SiC transistors. GaN technology under development will in future support operation at power levels as high as 22kW. Further advanced in its commercialisation, SiC MOSFET technology for bi-directional charger topologies is available today.
In the short term, the commercial focus of charging point manufacturers will be on unidirectional chargers for home and street use at power levels up to 22kW. And while a Vienna rectifier will take care of the PFC stage, EV charger manufacturers are entering uncharted territory in the high-voltage DC-DC conversion stage (see Figure 3). Certainly, this DC-DC stage will call for the use of SiC components because of their fast-switching capability – reducing the size of the magnetics – and their low switching and conduction losses.
There is today, however, no consensus on the most appropriate topology to be used in this stage: literature and reference designs showing how to implement such a design are scarce. Topologies including a dual active bridge are in use by some manufacturers, as they offer efficient operation. Others are working to adapt resonant DC-DC topologies to make extra efficiency gains.
Fig. 3: functional blocks in a 22kW home EV charger circuit. (Image credit: STMicroelectronics)
Given the lack of a standard approach today, EV charger manufacturers will be looking for third-party support and guidance on the integration of SiC components into high-voltage designs.
Here, Future Electronics has two particular advantages: first, as a global franchised distributor it maintains close relationships with suppliers of SiC components such as STMicroelectronics, ON Semiconductor and Microchip. This gives it privileged insight both into the performance and reliability data that these companies generate, and into the supply chain for these scarce components.
Second, Future Electronics’ power-specialist Centre of Excellence in London, UK has developed rare expertise in the development of production-ready custom power circuits for customers for challenging applications that consume more than 1kW. It has considerable experience in the implementation of wide bandgap semiconductors in high-efficiency systems. This expertise can be deployed to help charger manufacturers get to market with an efficient, isolated DC-DC power stage at the 200V-500V level required by EV batteries. The CoE’s position as part of a global distributor also means that designs created by Future Electronics are based on components that are commercially available in production volumes.
Expertise and Parts Availability
EV charger manufacturers, then, are studying how to implement a new generation of home and street chargers operating at power levels of 22kW and higher. As this article has argued, manufacturers should be looking carefully at the availability of SiC components, and at the need for expertise in using them in isolated, high-efficiency DC-DC converter stages.
Wider grid infrastructure issues also remain unresolved, and manufacturers will do well to press grid operators, regulators and governments to address them as early as possible.
 Microchip MSCSICPFC/REF5 reference design information available here.