Overview
Power Module Reliability (PEMREL) addresses the clean sky’s call “Sample PEM (Power Electronic Module) construction for testing, characterisation and manufacturability assessment. This proposal aims to compliment other research into the development of an all SiC 10kW inverter, implemented in a double-side cooled, non-hermetic sandwich packaging technology (no bondwires), able to withstand a nominal ambient temperature range of -60°C to +200°C.
Key developments will include FMMEA, and the development of physics-of-failure reliability models for both type 1 & 2 modules. These modules will be tested for electrical and thermal characterisation. Power and passive cycling of the modules will be undertaken to generate reliability data for validating the developed reliability models.
Funding
Results
Final Report Summary - PEMREL (Sample power electronic module construction for testing, characterisation and manufacturability assessment)
Power electronic modules (PEMs) are self-contained power electronics components that are widely used in aerospace, automotive and alternative energy generation & distribution applications. They play an important role in the conversion, control and delivery of electrical power. At present the vast majority of power electronics modules are packaged using solders and wirebonds. PEMs have highly inhomogeneous structures. They consist of semiconductors, ceramic, copper, aluminium, polymers and sometimes composite materials. These materials are assembled together in the packaging manufacturing process using soldering, direct bond copper (DBC), wirebonds, and pressure contact interconnection techniques. This is the traditional approach to packaging of IGBT inverter modules.
Present state-of-the-art manufacturing for plastic packaged IGBT modules involves the solder attachment of the dies to a substrate followed by ultrasonic bonding of IGBT emitter and gate and diode anode connections, there being up to 600 individual bonds in a large module. Creation of such a large number of individual bonds is expensive, time consuming and a source of yield loss. During operation, the wires are a known reliability weak point and in the event of device destruction the bond wires typically fail open circuit by a high energy fusing process. This failure process is inherently unpredictable and the high energies dissipated within the module can lead to rupture of the module housing. In addition, the typical bond wire and bus-bar assembly leads to relatively high levels of parasitic inductance which compromises device switching performance leading to increased switching loss and/or over-rating of device blocking voltage. Finally, the thermal management arrangements in conventional modules remove heat from just one die surface, restricting the maximum heat flux that can be sustained for a given maximum junction temperature.
Early failures in a power electronic module are usually due to manufacturing failures that show up quickly. These failures can quickly be screened out with stress burn-in techniques. The second period of the product reliability has a constant failure rate until it reaches the time when wear out mechanisms result in an increase in the failure rate. The constant failure rate is due to random events. Up until the 1990’s the exponential, or constant failure rate (CFR), model, had been the only model used for describing the useful life of electronic components and was the foundation of the military handbook for reliability prediction of electronic equipment's known as the MilitaryHandbook-217. This became the de facto industry standard for reliability prediction.