Data Center Journal


Issue link:

Contents of this Issue


Page 23 of 32

THE DATA CENTER JOURNAL | 21 i n World War II, many airborne radios became essentially unusable owing to static whenever the tun- ing knobs were turned. e cause of the failure was unknown, and it took the scientists at Bell Laboratories to discover that there were microscopic metal "hairs" apparently growing from one metal surface to a nearby metal surface in the meshing metal plates of the tuning ca- pacitors. It turned out that certain metals commonly used as corrosion-protection plating had this strange tendency to grow these hairs—called whiskers because they mostly stand straight out from a surface, as may the whiskers on a man's face. For the metal surfaces that were plated with tin, the solution was to add a little lead to the tin, and for reasons unknown the whiskers tended then not to grow much at all. at pretty much solved the problem, as it (usually but not always) took years for the whiskers to grow, and even then less than an eighth of an inch, and electrical spacings were fairly large. As electronics became smaller, the size of electronic components also de- creased, and when transistors shrank into integrated circuits, the leads were spaced an eighth to a sixteenth of an inch apart. But all the leads were plated with the tin- lead alloy so they would not corrode. ey could easily be joined to printed-circuit boards by controlled melting of a small amount of that same tin-lead solder to make the electrical connections. e sizes of these "integrated circuits" continued to shrink until today, when the leads on their packages may only be spaced a hundredth of an inch apart. Still, everything worked fine until certain politically inspired "en- vironmentalist" activity in Europe led to a law in that region, called RoHS (Restric- tion on Hazardous Substances). is law banned import of any electronics contain- ing lead. e electronic-component and product manufacturers changed from tin-lead-plated parts to RoHS-compliant pure-tin-plated parts. In many cases this change owed to not knowing any better— specifically, the lessons of three genera- tions ago, that pure-tin-plated surfaces grow whiskers. With component leads now so close, the chance of failure in about five years from a tin whisker growing from one lead to another may be about 30%. e "consumer" didn't care, much less even know, having been trained to toss the old appliance when something new and shiny was offered by manufacturers that needed to maintain their cash flow. us, most folks didn't own anything electronic very long. With products con- taining electronics upgrading so fast, the consumer might have one only a few years before buying a newer and better model— and so never noticing the rare failures that occurred for no understandable reason. Mass-produced consumer products failing in only a few years became common, but with costs being so low, it became "nor- mal" to toss a device when it quit and just buy new. Besides, as the tin whiskers oen made only erratic contact, the "failures" only appeared intermittently. us, the microcircuit leads (above) emerging from package; scale graduations = 0.01" Typical zinc whiskers from Data Center floor tile; scale graduations = 0.01". note longest whisker (>0.14") can easily bridge seven microcircuit leads

Articles in this issue

Links on this page

Archives of this issue

view archives of Data Center Journal - VOLUME 41 | DECEMBER 2015