Power Semiconductor Considerations for Automotive OEMs and Suppliers

The power semiconductor industry is amid a revolution, with device vendors racing to harness cutting-edge technologies to tackle new applications that bring both immense challenges and exciting opportunities. Paramount amongst these is the electric vehicle, where effective use of power semiconductors can increase range, reduce the size and weight of powertrains, reduce charging times, and lower battery costs.

There are now a diverse range of solutions for automotive OEMs and tiers ones who are looking to navigate through their own rapidly evolving industry. The power semiconductor options can be broadly classified by the materials they are fabricated from, each of which has benefits and weaknesses. 

The traditional silicon remains a compelling option. Silicon IGBTs are a workhorse. They have been refined over multiple generations to combine good performance with excellent reliability and cost. They are also available from a broad range of suppliers globally making them a robust choice and the standard against which other solutions must be judged.

Silicon carbide MOSFETs are the high-performance device of choice. These MOSFETs have higher efficiency than silicon IGBTs and can result in small system size, but this comes at a significantly higher price point. Since Tesla first introduced SiC to the powertrain inverter its use has expanded to include many vehicles in the mid to high end range.

Finally, there is gallium nitride, a material may be familiar from its use within power adapters but is yet to penetrate the automotive market in any appreciable way. Gallium nitride also has excellent efficiency, and its price is much closer to that of silicon. However, it cannot scale as easily to high voltage and has a prevailing reputation for unreliability. This reputation is largely unwarranted but its lack of field data in automotive applications is sometimes a barrier to adoption. 

So which material should powertrain engineers opt for? Silicon should remain the standard, and it is only full electric vehicles which should consider an alternative. Whilst improved electric range and charging would benefit plug-in hybrid vehicles, the back-up option of an internal combustion engine makes it more difficult to justify the extra chip and engineering cost. 

For fully electric vehicles, silicon carbide is beating out silicon in many cases. The fully electric vehicle market is diverging down two separate paths – vehicles with a smaller range designed to cover average day to day usage, and those with a larger range which account for all driving profiles including longer journeys. 

As things stand, those targeting a smaller range should stick with silicon IGBTs within their traction inverter. Range just needs to be sufficient, and cost is the more crucial factor. Alternatively, instead of improving range silicon carbide or gallium nitride could be used to reduce battery size and make cost savings there. Currently this calculation doesn’t come out on the side of the new materials, but it is possible to imagine a scenario soon where it does. 

When a car manufacturer is trying to maximise range silicon carbide is the obvious solution. Vehicle manufacturers are banking on customers being willing to pay a premium to get the range they need, and these vehicles are typically not economy vehicles. More and more of these vehicles are utilising 800 V battery systems and silicon carbide is a key enabler of this transition, with 1200 V devices readily available. Silicon carbide has also matured to the point where there are several suppliers of automotive qualified parts with their own varied and diverse supply chains. Previous silicon carbide chip supply constraints have loosened. 

So, when is gallium nitride the right option? It is expected that gallium nitride will first be used in on-board chargers beginning potentially as early as 2025. This will help to mature the automotive qualified gallium nitride device ecosystem and could potentially pave the way for its inclusion within inverters. Whilst 900 V and 1200 V gallium nitride devices are available, they are still very much in their infancy, therefore a shift to three-phase topologies may be needed for compatibility between 800 V batteries and the widely available 650 V devices. This is currently on the roadmap of many manufacturers due the expected improvement in performance, but silicon carbide also has 600 V devices available, meaning it is not a guarantee that gallium nitride would be used. 

Once the ecosystem matures sufficiently, gallium nitride devices will provide a significant performance edge over silicon counterparts with only a slight price increase, making the cost-benefit analysis favor gallium nitride. In this case we may see gallium nitride inverters used in economy vehicles and perhaps even plug-in hybrids.

The target market, battery choice, and design topology all impact which semiconductor material should be chosen. Silicon carbide has already found its way in this increasingly complex product area, whilst gallium nitride has a lot of potential but a long road ahead of it. Omdia can provide support to powertrain designers looking to understand the supply landscape, or device vendors looking to plot their course through it.