The Engine Control Unit (ECU) for small engines is facing challenges with regard to performance, size and cost. In many instances, the customer requirements often contradict each other. Examples include higher performance at lower cost or smaller size, both of which can cause thermal challenges. In order to meet varying performance requirements in a platform approach, the ECU must provide a wide range of functionality. Providing a solution that can meet these flexible requirements will result in an increased component count and larger ECU size. An optimized feature set in the right package can help alleviate these issues. The ECU must be impervious to a wide range of environmental conditions, such as temperature, humidity and vibration. Restricted air flow must also be considered when designing an ECU. Existing approaches often apply the use of large aluminum housings to provide a strong mechanical support with good thermal performance. As the market pushes for lower cost solutions, housing materials are being evaluated as cheaper alternatives become available. Alternative housing materials will drive the electronics design to either reduce the power dissipation or improve thermal performance.
When ECU designs do not meet customer expectations for cost, size, reliability or performance we need to look for ways to optimize the system design in order to meet the variety of system requirements. To improve the system design we must evaluate the system requirements to find an optimal level of feature integration that can be achieved economically. At the semiconductor level we need to determine the optimized features and functions that can integrated in a single piece of silicon or package to help reduce the size of the ECU without degrading the thermal capabilities of the design. However, the operation modes and environmental conditions of the system must be understood prior to the definition of the semiconductor device.
This paper demonstrates the process of optimization and identifies the trade-offs between performance, cost and size based on a system-based thermal simulation model. The simulation model approximates the environment, wire-harness, ECU housing and PCB with heat dissipating features using a detailed thermal model for the semiconductor devices. Contrary to other simulation models, which only consider static conditions, this model and simulator can also provide responses to transient thermal events. With this methodology the system can be thoroughly analyzed with respect to the established system requirements providing a thermally optimized solution.