Design007 Magazine


Issue link:

Contents of this Issue


Page 21 of 71

22 The PCB Design Magazine • October 2015 feature Limitations of the Discipline-Centric Process Looking back just a few years, develop- ers of electronic products primarily competed based on the functionality of their design as determined by unique hardware and software features. Today, with much of the key function- ality of electronic products having been com- moditized in system-on-chips (SoCs) or applica- tion processors, electronics companies are now competing across a wide range of fronts: size, weight, style, battery life and features. The re- sult is a greater need than ever for collaboration across the multiple disciplines that are respon- sible for providing this much broader range of attributes. More than ever before, a multidisci- pline collaborative effort is required to deliver the best design that conforms to marketing re- quirements in the shortest possible time and at the lowest possible cost. Yet the tools used to support product de- velopment and the process itself have not yet evolved to this new reality. In most cases, the marketing department draws up the require- ments documents that then explode into mul- tiple independent paths. The creative team styles the product. The procurement team looks at part cost. PCB designers design the boards one at a time. The manufacturing team decides where to make the product and what processes to use. Mechanical engineers design the enclosure. There may be some limited in - formal collaboration between disciplines us- ing a spreadsheet or flowcharting tool prior to detailed design, but for the most part, each of these disciplines works independently with rel- atively little interaction with the others. This is a critical time where architectural decisions are being made with very little validation. In addi- tion, this is generally a sidebar process whose results are not easy to integrate back into the separate design disciplines or into the detailed design; any communications errors can be di- sastrous. The very nature of multi-discipline design processes suggests that decisions are frequently made that are good for the discipline that made the decision but detrimental to the overall prod- uct. For example, the PCB designer might decide to relocate a connector on the bottom right side of the PCB because it is beneficial from a signal integrity standpoint. But when the industrial designers become aware of this decision they might object because users tend to rest their fin- gers in this location. And the mechanical engi- neers may conclude that having an additional ACCELERATING THE DESIGN CYCLE Figure 1: The multidiscipline design process.

Articles in this issue

Archives of this issue

view archives of Design007 Magazine - PCBD-Oct2015