ECWC 2014: The Metal-base PCB Technology Session
At the 13th Electronic Circuits World Convention in Nuremberg, Germany, the session on metal-base PCB technology was well-attended, with delegates interested to learn about life cycle cost optimisation, latent short circuit failure, and a particularly innovative alternative approach to manufacturing electronic assemblies. The session was introduced and moderated by EIPC Chairman Alun Morgan.
Dr. Viktor Tierderle from RELNETyX Consulting discussed the optimisation of life cycle cost by selection of PCB laminates, including metal-based materials, using physics-of-failure methodology. For printed circuits to perform reliably under harsh conditions of thermal and mechanical stress, it was important to choose the appropriate laminate from the many available in the market, and to select the best balance between cost and performance to fulfil the requirement.
Dr. Tierderle demonstrated a procedure which began with a CAD data input and took into account all of the relevant issues: Temperature cycling, sustained temperature, humidity, corrosives, power cycling, electrical loads, electrical noise, mechanical bending, random vibration, harmonic vibration, and mechanical shock, using a series of simulation and three-dimensional modelling techniques to predict reductions in life cycle costs. The outcome was the ability to reduce field failure rates and the number of testing cycles required, and to enable the selection of an adequate-for-purpose material with an optimised cost-performance relationship.
Stan Heltzel from the European Space Agency emphasised the critical importance of cleanliness in processing PCBs to avoid latent in-service short circuits. A consortium of space agencies, ESA-qualified PCB manufacturers and leading European aerospace OEMs, had worked on revision 12C of the ECSS-Q-ST-70 Space Product Assurance specification, shortly to be published, which would define new design rules for PCBs. Lack of cleanliness in PCB laminates had been demonstrated in several case studies to be a cause of failure due to electromigration, and Heltzel described the inspection procedures used by ESA on materials, manufacturing processes, and finished PCBs. He discussed risk mitigation strategies, cleanliness control plans, and test methods that could detect reduced insulation due to contamination and electromigration, and presented proposals to tighten the requirements on base materials for the manufacture of high-reliability PCBs for space applications.
The presentation that attracted most interest and provoked plenty of interactive discussion came from Joe Fjelstad of Verdant Electronics who described an alternative approach to the manufacture of electronic assemblies, which used aluminium as a substrate and eliminated the soldering process. Aluminium had many attributes which made it attractive as an alternative circuit substrate: Abundance, low cost, thermal conductivity, light weight, dimensional stability, thermal expansion comparable with copper, ease of processing, and capable of being anodised or electrophoretically coated.
Ironically, the main reason aluminium had not been considered for traditional PCB applications was its high thermal conductivity and heat-sinking ability, which limited the practicability of the soldering process. Except in applications such as LED substrates, designers had preferred to specify resin-based laminates and then deal with thermal management issues after completion of assembly. And even in the knowledge of the thermal issues and intrinsic reliability problems associated with the introduction of lead-free soldering, human inertia--“better the devil you know” attitudes--had tended to preclude the consideration of radical alternative concepts.
Fjelstad advocated reversing the assembly process to eliminate the solder issue, by positioning and bonded all the components on an aluminium carrier with leads facing up, encapsulating the components in place, exposing the terminations, and interconnecting them by additive or semi-additive PCB fabrication techniques or alternative direct interconnection methods. Because solder lands were not required, conductor routing density could be higher so that fewer layers would be required. The basic manufacturing infrastructure was already in place, semi-additive fabrication processes were well-established, and the concept was appropriate for all classes of products including flex.
The benefits of solderless assembly were many: No PCB was required, no soldering was required, component concerns were reduced, circuit design layout was easier, design security was increased, thermal management was integral, reliability was enhanced, and multiple novel structure options were possible--modular array assemblies, aluminium-core rigid-flex, direct-write prototyping. The limits were more likely to be defined by the imagination of the designer than by the fundamental limitations of the technologies that Fjelstad had described. “Change may happen slowly, but change always comes…”