Tin WhiskersTheyre Back!
The Galaxy IV, a $250 million communications satellite, suddenly stops working. Its failure silences about 40 million pagers and interrupts millions of dollars worth of banking transactions. It becomes, as one NASA engineer puts it, a doorstop in space.
The US Food and Drug Administration initiates a Class I Product Recall of heart pacemakers; the US Air Force experiences problems with radar equipment in its F-15 fleet; the Millstone power station near Waterford, Conn, shuts down just as if a steam line had rupturedbut no plant failure actually occurs: There was no ruptured steam line, no drop in pressure, and no problem with the reactor itself. It was a false alarm, a computer glitch.
Close up view of a Tin Whisker "growing" from a pure tin-plated lead of an installed component. Eventually, it can grow and attach to an adjacent lead, causing a short.
What do these seemingly unrelated events have in common? What powerful post-Y2K force could kill both a quarter-billion-dollar satellite and an even more expensive nuclear power plant? Like the classic black-and-white movie version of Orson Welles War of the Worlds, all succumbed to the smallest of thingsnot a virus, but a sliver of metal, thinner than a spiders strand, called a tin whisker.
A tin whisker is a strand of pure tin that grows as a crystalline structure from the surface of pure tin electroplate and coatings. It is typically only 1 or 2 microns in diameter, but occasionally it is as large as 10 microns in diameter. A tin whisker can be seen with the naked eye only when struck by light at just the right angle. It may grow straight, kinked, or spiraled to as long as 1¼2 inch, can carry tens of milliamps before fusing and vaporizing, and has a resistance of around 50 ohms. A single whisker can grow from one component leg to another, from a component to ground, or on pure-tin-plated items within the device, and then break off and land inside the device. Rarely does a single tin whisker grow from a pure tin or a pure-tin-plated surface. Surfaces capable of producing tin whiskers can produce dozens, hundreds, or (in the case of metal covers) thousands of individual whiskers.
A Little History
Where did they come from? How did they get in there? That mystery has almost 60 years of history behind it. The story starts further back than most electronics technicians can rememberin fact, before most of todays biomedical equipment technicians (BMETs) were even born.
The story of tin whiskers began in 1946, when the cadmium-coated plates of a variable condenser (capacitor), used to tune the radio in which it was installed, sprouted whiskers long enough to short the adjacent plates. Two years later, channel filters (used in multichannel telephone lines) failed for no apparent reason. Bell Telephone Corp investigated and found the root cause to be cadmium whisker formations within the filters. Further research revealed that whiskers spontaneously formed, not just on cadmium but also on zinc, tin, certain aluminum casting alloys, and, in certain environments, silver electroplate. Between 1948 and the early 1960s, various researchers delved deeper into the phenomenon of tin-whisker growth, publishing numerous papers on the subject. They concluded that:
Each whisker was a single filament stack of individual crystals, growing up from the base and not out from the tip.
Whiskers grew from both solid metals and electrodeposited coatings.
Although growth rates varied, whiskers grew most quickly when the host metal was exposed to stress, caused by anything from cooling of the coating to the physical pressure of mounting hardware.
Whiskers grew straight, bent, kinked, or in combination within the same whisker colony.
Whiskers at times grew toward a differential voltage.
Whiskers stop growing as suddenly as they start.
Whisker mitigation could be achieved by alloying the final finish tin with 3%10% of lead.
For the next 45 years, tin whiskers were unheard of, since manufacturers of electronic components used a tin-lead alloy as the final component coating. This combination was well-received by the electronics industry since it took solder well and was compatible with the rosin-core 60/40 tin-lead solder used throughout the industry. Meanwhile, research into the whys and hows of tin-whisker formation continued, albeit at a slower pace, since the industry-wide use of tin-lead coatings solved the problem.
Laboratory experiments and empirical data indicate that tin whiskers may sprout from pure-tin surfaces months or years after manufacture, or not at all. Although the complete pathology of tin whiskers is not fully understood, one major factor in their growth is mechanical stress. Stress can come from the uneven cooling of plating and finishes; physical pressure (compression) of pure tin-coated surfaces caused by press-fit assembly; bending during component assembly; and mounting hardware such as nuts, screws, and flat washers. In short, tin-whisker growth is unpredictable, and tin whiskers themselves are a failure waiting to happen.
If the problem was solved in 1959 using a tin-lead final coating and has remained in check since then, why did the satellite, pacemaker, radar, and power-plant failures occur? The answer lies in Europe, almost half a world away. In 2003, the European Union published earth-friendly legislation, requiring articles sold in member countries after July 1, 2006, to be free of lead, mercury, cadmium, hexavalent chromium, and numerous other metals and chemicals. China is considering a similar prohibition. Although the United States is not considering similar legislation, the Environmental Protection Agency added both lead and lead compounds to its list of persistent, bioaccumulative, and toxic chemicals and changed reporting requirements from 25,000 pounds per year to only 100 pounds per year.
Reacting to this global trend by marketing more green products, component manufacturers have shifted to using pure-tin coatings (again) as a final finish on their parts. Their parts go into printed circuit boards supplied to original equipment manufacturers (OEMs). To compound the problem, the Department of Defense (DoD) dictated a re-examination of the use of military specifications and standards. In 1994, the DoD mandated the greater use of performance and commercial specifications and standards. The new policy stated that performance specifications shall be used to the maximum extent practicable when purchasing new systems, and allowed the use of other nongovernment standards to capitalize on efficiencies obtained in the private sector. In essence, where military specifications (commonly called Mil Specs) called for tin-lead coatings and banned the use of pure-tin coatings as final finishes, manufacturers and industry could now do whatever cost the least and provided the highest profit margin. As an indirect result of this policy, lead-free whiskering components ended up in the Galaxy IV communications satellite, with disastrous results years later.
With additional research, which continues today, several promising tin-whisker-remediation methods have been developed to replace the tin-lead coatings in use. Two of them are the use of a conformal coating and the replacement of pure-tin coatings with other metals and alloys. Each has its own unique disadvantage, and neither works as well as the tin-lead alloys currently in use by the electronics industry. Conformal coatings need to be applied thickly; otherwise, tin whiskers will grow right through them. This adds another cost to the production of circuit boards and other subassemblies, but it does not ensure complete mitigation. Heat from conventional soldering equipment destroys the coating. Therefore, recoating in the field is necessary to restore and maintain future tin-whisker mitigation. Unfortunately, this requires chemicals and industrial processes unsuited for typical biomedical equipment maintenance shops. Other suitable metal finishes, not prone to whisker growth, are more expensive than the current tin-lead alloys in use and require both higher soldering temperatures and special solders. These temperatures are high enough to destroy silicon-based components like transistors, integrated circuits, and multilayer printed circuit boards. To date, the most cost-efficient, trouble-free mitigation method remains the 45-year-old practice of using a tin-lead alloy as the final coating on component cases and leads.
As evidenced by tin-whisker-related failures, lead-free components are finding their way into globally marketed electronics. Although contractually banned from Galaxy IV, they managed to slip through anyway. A nonwhisker, but related, incident occurred in December 2001, when the Netherlands government refused entry of 1.3 million Sony PlayStation® game consoles because their cables contained cadmium. (the Netherlands has strict laws against products with cadmium.) Since OEMs do not want to lose international market share (as Sony did because of this incident), they are already using lead-free components and remain oblivious to the impending tin-whisker problem. Why? Because the engineering community has forgotten about tin whiskers because, like so many childhood ailments, whiskers were thought to have been cured. As a result, information about them no longer appears in textbooks or is discussed in university lectures. Since a large part of their target market (Europe, in this case) bans the sale of lead in products, manufacturers started using lead-free components exclusively in their products. The alternative would have been to stock and maintain two sets of componentsone using tin-lead finishes, the other using pure tin finisheswhich is not only confusing, but also reduces profits.
Importance to BMETs
BMETs need to know about the problems and symptoms caused by tin whiskers, the remediation methods and their associated problems, and the identification of tin whiskers and tin-whisker-related failures. Articles such as this are a good start. Some of the best tin-whisker information is at the NASA Goddard Space Flight Center Literature References page at http://www.nepp.nasa.gov/whisker. This site contains references and links to more than 180 tin-whisker events and documents. Tin whiskers are a phenomenon that once again is gaining the notice of the upper echelons of the electronics industry (satellites, missiles, and other high-cost items), but little of this knowledge has yet trickled down to the medical-electronics subset of the electronics industry.
Until a permanent solution to the tin-whisker phenomenon is found, the best remedy is good preventive maintenance. At the maintainer level, the tin-whisker-remediation method first recommended in the 1960s by Northern Electric Corp and again by General Electric in 2000 is still the best. They both endorse vacuuming the equipments interior to remove loose whiskers and those forming on cabinet components, and vacuuming printed circuit boards to remove growing tin whiskers before they can cause a failure. More than 40 years ago, the Tin Research Institute recommended the use of a nonconductive brush attachment on the vacuum hose.
A Growing Problem
As previously mentioned, tin whiskers can grow not only from circuit components, but also from washers, nuts, screws, and other mechanical fasteners used in the construction of medical devices. If these fasteners are inside the case, they can break off and land on circuit boards and across dual inline-pin-switch contacts, or they can work their way inside moving components and cause shorts. Other possible sources of tin whiskers are pieces of components themselves that stay inside the component in which they grow. Small switches and potentiometers using pure tin-plated parts in their manufacture have succumbed to internal tin-whisker failurescausing both temporary and permanent internal shorts. The temporary shorts might explain why some devices result in unable to reproduce problem notations on the work order.
When the bench technician finds that a new componentsay, a capacitorwill not take solder, the technician should check with the manufacturer to determine if the final coating is one of the new nonwhiskering alloys and whether a particular soldering technique should be used. Another effect of these new platings, besides their reluctance to bond with common tin-lead solder, is solder brittleness and cracking. A probable replacement for common tin-lead solder might be tin-silver-copper solder. This and other new solders require temperatures around 470°F500°F, versus approximately 365°F for the solder commonly used today. Additional heat sinking will be required to protect components from the higher heat. These new solders may contain components that are even more toxic than lead. Therefore, additional protection may be required when using them.
Some BMETs advocate applying immediate action to resolve a probable tin-whisker failure: a swift, sharp, lateral shock to dislodge the offending whisker and get the equipment working again. While it might work fine for your personal equipment, in the privacy of your own home, out of sight of physicians and nurses, it is not the kind of remedial maintenance I want regularly applied to equipment used on my significant other or myself. Certainly, medical-equipment technology and our troubleshooting skills have advanced past this now that were in the 21st century.
Robert M. Dondelinger, CBET-E, MS, is the medical equipment manager at the US Military Entrance Processing Command in North Chicago, Ill.