Power modules gain from important new packaging materials and assembly technologies

As power densities have increased across virtually all electronic products, some power-module applications have approached the limit of what traditional packaging materials and assembly technologies can support. In response, a large number of new packaging alternatives have emerged. In advance of his keynote presentation at PCIM Asia 2012 this coming June, Semikron Elektronik GmbH Chief Technology Officer, Dr. Thomas Stockmeier, generously fielded questions on this topic from Power Systems Design: PSD: Compared to typical periods in our industry's evolution, recently, there seem to be an unusual number of new power-module-assembly technologies coming to market. Do you see these as reflecting demands of disparate applications or are we seeing a fractioning of the market served until now by modules born of more traditional assembly technologies? Dr. Stockmeier: Traditionally, power modules serve markets like industrial applications, automation, welding, power supplies, and other applications, where already a broad spectrum of power modules is necessary to accommodate the different power levels and circuit topologies. However, new applications, like hybrid and electric vehicles, as well as renewable energy have gained quite some momentum, which now rightfully request power modules tailored to their respective needs. Therefore, the market is not really fractioning more than in the past, but new fields are emerging. PSD: Are the distinctions between competing assembly technologies sufficiently reflected in power modules' data sheets or do system designers need to be aware of the assembly technologies exploited by the power modules they specify? If the latter, how should such an awareness affect their module-selection process? Dr. Stockmeier: Data sheets can only serve as a first order selection tool; the specific fit of the selected module has to be intensively tested and qualified. Data sheets will most likely not provide enough information to select the right power module without testing. Whereas the selection with respect to the right power handling capability can be determined with the help of simulation tools provided by module manufacturers, many other parameters, such as EMC, or reliability and durability for a specific environment are not possible to specify fully from the module supplier. PSD: It appears that module manufacturers are exploiting many of the new assembly technologies to accommodate increasing maximum junction temperatures of the modules' various pass elements. Do wide bandgap semiconductors, such as SiC and GaN that claim better power-handling capability than traditional Si, depend on these new module-assembly technologies? Dr. Stockmeier: The great advantage of wide bandgap materials, such as SiC and GaN, are their higher junction temperatures and much faster switching speeds. This may ultimately allow much higher power densities for many power electronic systems. However, traditional power-module-assembly technologies and designs cannot accommodate these needs. In particular, module manufacturers must develop new designs to address medium to high power applications. PSD: If power-module manufacturers are beginning to exploit multiple and diverse packaging materials and assembly techniques, will the market not necessarily see increasing module costs across the board due to falling economies of scale? Dr. Stockmeier: One has always to consider the system cost, which power modules must help decrease. For example, if the power module allows much higher power density and switching frequency, power-subsystem designers can dramatically reduce the size and cost of filtering. Therefore, some shift in cost from filter to power module may become acceptable. Lower system cost also results from functional integration in power devices, eliminating expensive electrical, thermal, and mechanical interfaces. Again, the power module may have higher cost, but the overall system cost is less. Another benefit: fewer interfaces and less material yields higher reliability. PSD: Do you see power modules for specific applications—such as automotive, communications base stations, or computational server farms—clustering around specific packaging and assembly technologies or do applications draw from the full width and breadth of available packaging materials and assembly methods? Dr. Stockmeier: I don´t think that a specific market will exhibit its own specific set of materials and technologies. One example: recently, we completed a study on power modules for the main electric drive in automotive applications. We found all possible ways of applying materials, technologies, and packages. What we have to be careful about is that we don´t re-invent the wheel many times. Precompetitive research and development and co-operation across application fields become a necessity. Initiatives like ECPE (the European Center for Power Electronics) are exemplary. PSD: Often, when a technology sector sees a sudden increase in the number of technologies competing to solve a given problem, there follows a fierce battle between them followed by significant consolidation. Will the recent bumper crop of innovations in power-module packaging follow this classic trend or is there a sufficient market niche for each? Dr. Stockmeier: Has the power-module market ever really consolidated technologies or designs? I see new functions, forms, and fits of power modules at every PCIM show. Very rarely do I see the discontinuation of specific power devices. The reason is that the market is still fast growing. While old designs and technologies serve existing production, which may go on for decades, new projects need and incorporate new technologies for lower cost, higher power densities, and higher reliability. Maybe what we see is not a technology bump, as you call it, but the beginning of an even more diverse (and complicated) world. PSD: Given the growing cost sensitivities in many of the end applications, are specific assembly technologies gaining an edge as the pressures of commercialization come to bear? Dr. Stockmeier: Power modules are very sensitive to material cost. Module manufacturers achieve cost reduction by using less material (such as a module without a base plate), improved thermal conductivity to enable chip shrink, and functional integration to eliminate interfaces. Secondly, large production volumes are required to justify a high degree of automation, to gain productivity, and leverage supply. Any technology that serves these purposes has an edge. PSD: Evidently, reliability concerns have been key drivers for many of these innovations. Of course, power modules are not the only components subject to failure due to high-temperature operation and thermal-cycling stress. How can innovations in power-module packaging inform developments in packaging technologies for complex integrated circuits? Dr. Stockmeier: I don´t think, integrated circuits are our greatest headache in high temperature operation. IC technology provides suitable packages and continues to develop even better ones. Of much greater concern are, for example, capacitors, current sensors, and PCBs. PSD: Are most power-module manufacturers exploiting new materials and assembly methods or are only a few vendors commercializing these new technologies? Dr. Stockmeier: All power module manufacturers very actively research and develop new materials and assembly technologies. This makes our technical community so interesting and makes shows and conferences such as PCIM so vivid. Whereas the overall goals are most likely the same for all manufacturers, the ways to solutions are different. All vendors intend to commercialize these technologies - sooner or later. PSD: If the current power-density trend continues, will it force power-module manufacturers to adopt hermetic packaging methods? How are innovations in non-hermetic packaging staving off that outcome? Dr. Stockmeier: High power density does not require hermetic packaging. Much higher power density can be achieved by liquid cooling technologies, either integrated or discrete. Hermetic packages, on the other hand, are necessary in harsh environments and, there, all other components of power-electronic circuits require consideration, as well. If the power density of modules increases drastically, and designs can considerably reduce the filter size, the compartment that houses the entire power electronic system may become much smaller. Therefore, it may be more useful to look at a suitable sealing for the entire (small) compartment, rather than just looking for hermetically sealed power modules. PSD: Thank you, Dr. Stockmeier; we look forward to your keynote address at PCIM Asia 2012. www.semikron.com www.powersystemsdesign.com