Efficient heat management design for modern electric vehicle on-board chargers

In this technical white paper, Infineon Technologies examines the challenges facing on-board charger designers and takes a closer look at the role that semiconductor packaging plays in creating a solution. The paper also explores an innovative approach to top-side cooling that can be used across a range of high-performance components for designers to choose from.

The automotive industry is advancing at a rapid pace, with every aspect of automotive design improving from chassis to powertrain, from infotainment to connectivity and automation. However, the long-standing issue of charging time for electric vehicles (EVs), especially when charging on the go, is a huge inconvenience that has hindered the popularization of electric vehicles, so the design of on-board chargers (OBCs) may become an area of ​​concern. As with most engineering challenges, designers are looking to advanced technologies to provide solutions using modern silicon superjunction (SJ) technology and wide bandgap (WBG) materials such as silicon carbide (SiC) and gallium nitride (GaN). But semiconductor materials are only part of the solution. For any on-board charger design to fully realize its potential in terms of power density and energy efficiency, efficient heat dissipation design is essential. Infineon Technologies examines the challenges facing on-board charger designers and takes a detailed and in-depth look at the role of semiconductor packaging in creating solutions. This article also explores an innovative approach to top-side heat dissipation that can be used in a range of high-performance components for designers to choose from.

Modern Electric Vehicle On-Board Chargers—Design Challenges

The role of the on-board charger is to convert AC power from the grid into DC power for charging the power battery. The on-board charger performs this function only when the car is parked and connected to the charging cable. When the car is driving, it can only carry this heavy object all the way, so the size and weight of the on-board charger must be minimized to reduce its impact on the driving range while achieving fast and efficient charging.

Another challenge is the rapid increase in the power level of on-board chargers. A few years ago, 3.6 kW was the state-of-the-art technology, but in the near future, the power will be increased by about three times, that is, the power can reach 11 kW in the same space. On-board charger designers face five major interrelated challenges. Among them, improving power density is particularly important because it means reducing size and weight, which helps to extend the range of electric vehicles. Improving energy efficiency not only reduces the heat accumulation inside the on-board charger (which reduces the requirements for thermal management, thereby reducing the size of the on-board charger and correspondingly increasing the power density), but also provides more power from the grid to charge the power battery, thereby shortening the actual charging time. The voltage of the power battery continues to increase, and the typical voltage has increased from 400 V to 800 V. This is mainly to reduce the current and related I2R losses transmitted in the cable when charging and delivering power to the main drive motor.

Figure 1 OBC designs present a number of challenges for power electronics designers The requirement for bidirectional operation presents another challenge for on-board charger designers. As electric vehicles become more popular, the pressure on the grid will increase significantly, especially as people may charge their cars at the same time (for example, charging at night after the daily commute). Electricity providers recognize that the large amount of energy stored in electric vehicles can be used to stabilize the AC grid and to power homes during peak hours to reduce peak demand. In addition, when the AC grid fails (blackouts), electric vehicles can act as "house batteries." However, to do this, the on-board charger needs to be able to feed power from the traction battery in addition to receiving power. To meet these challenges, the choice of topology and technology is important, especially for the switching elements. In most cases, WBG solutions will help provide the required performance advantages. However, while knowing the benefits of WBG technology, designers must also consider that improved thermal performance is critical to achieving these important goals.

Top Cooling - Overview and Advantages

The automotive environment presents many hazards to electronic components, including dust, dirt and liquids, so most electronic systems in electric vehicles are protected by sealing. This does not allow for forced air cooling, so thermal management is usually to conduct the heat generated by high-power components to the coolant in the electric vehicle. Generally, the heat conduction path for high-power SMD components is from the power device down to the PCB, which is bonded to the heat sink. This approach is called "bottom-side cooling" (BSC). In applications where the heat dissipation task is difficult, the power device can be mounted on an insulated metal substrate (IMS), which can optimize the heat dissipation performance because the thermal conductivity of the IMS is better than FR4 with thermal vias. However, the bottom-side cooling approach always involves a trade-off between heat dissipation performance and board space utilization. Through innovative packaging, Infineon has developed top-side cooling (TSC) technology for power discrete devices and power ICs. This technology has many advantages, all of which can benefit on-board charger designs and other similar applications. Bottom-side cooling usually uses a heat sink to be mounted on the bottom of the PCB/IMS to dissipate heat. This leaves one side without components, thus reducing the power density by half. Bonding semiconductor devices to the PCB to dissipate heat means they will operate at the same temperature. FR4's Tg is lower than the operating temperature of many modern power devices, which limits these devices from operating to their full potential.

Figure 2: TSC allows components to be placed on both sides of the board, doubling the power density

These issues are addressed by bonding the heatsink to the top of the power components, allowing components to be placed on both sides and the WBG devices to operate over their entire operating temperature range. While IMS offers better thermal performance than FR4, it also adds complexity. In fact, many IMS solutions become multi-board assemblies, where IMS is used only for power devices and FR4 is used for drivers and passive components. This makes design and manufacturing extremely complex. However, a recent teardown report shows that in reality, this assembly uses 169 connectors - while an equivalent top-side heatsink design only requires 41. ①

Figure 3: The number of connections required for a simple TSC assembly can be reduced by up to 76%.

Changing to a single-board TSC design can reduce the use of 128 connectors, which not only saves costs and reduces complexity, but also virtually resolves the reliability issues caused by these connectors. It also saves the cost of IMS. According to the analysis of the disassembly report, the assembly cost is reduced by one third. The key parameter of heat dissipation design is the thermal resistance between the semiconductor junction and the heat sink, because this parameter defines the ability to conduct heat. Heat dissipation simulation shows that the thermal resistance of top heat dissipation on FR4 is 35% better than that of bottom heat dissipation on FR4, and even slightly better than that of bottom heat dissipation on IMS, while the cost is greatly reduced.

Figure 4: Despite the lower cost, TSC outperforms the bottom-cooled IMS design The thermal limitations of the FR4 itself are relevant here, as this is a safety requirement. In the bottom-cooled solution, the MOSFET is bonded to the FR4, which means that the temperature of the FR4 is very close to the semiconductor junction temperature. The temperature limitations of the FR4 mean that the operating temperature of the MOSFET is also limited, and therefore cannot reach its full potential. In the top-cooled solution, the MOSFET is not bonded to the FR4 to dissipate the heat, so the MOSFET can run at a higher temperature. When using IMS, it is usually necessary to mount the driver and passive components on a separate FR4 PCB, so there may be a large distance between the gate driver and the MOSFET, which inevitably aggravates the parasitic effects and thus causes ringing.

Figure 5: SMD power device with top cooling

Short gate traces reduce parasitics TSC allows all components to be placed on the same double-sided PCB, so the driver can be placed directly below the corresponding MOSFET, significantly reducing PCB-induced parasitics. This will improve system performance and produce cleaner waveforms, thereby reducing electrical stress on power components.

Assembly Considerations

As discussed previously, a typical TSC assembly is usually simpler than an equivalent bottom-side cooling solution, not least because it uses only one board and requires significantly fewer connections.

Heat sinks are mounted directly to the heat-generating MOSFET package on the top of the PCB to dissipate heat. Thin components are also placed on this side, and thick components are placed below. When developing the HDSOP series, Infineon has ensured that each component has a nominal height of 2.3 mm. This uniform height greatly simplifies the heat sink, eliminates the need for machining, and allows for a more optimized heat sink to be used even when different power semiconductor technologies are used in the same design.

Figure 6: All HDSOP devices are uniformly heighted, greatly simplifying heat sink design and assembly There are multiple ways to bond the MOSFET package to a heat sink for heat dissipation. Generally speaking, the simplest method is to place a thermally conductive gap filler pad between the MOSFET and the heat sink. The optimized gap filler height can achieve the best thermal performance, but the prerequisite is that the gap filler is filled without leaving any gaps. In addition, liquid gap fillers can be used in fully automated production lines.