Bob Landman’s comments to my tin whisker posts appear below. Friendly dialogue such as this helps us to all learn more and is appreciated. Thank you Bob, keep me honest!
However, Bob’s comments do not change my position, which is:
1. Tin whiskers exist and can cause failures
2. However, there is yet no data that suggest that there are numerous tin whisker failures, or that a significant reliability risk exists due to tin whiskers in RoHS-compliant products. NASA's TW website notes only 26 fails.
3. Although not completely understood, tin whiskers can be created in the lab, and mitigation (not elimination) and reliability test techniques exist and have been demonstrated.
4. With well over $1 trillion in RoHS-compliant electronics manufactured since the early 2000s, there have been no significant reliability issues as compared to tin-lead solder.
5. Long term lead-free reliability (> 5yrs) has not been demonstrated. Hence, mission critical products should not use lead-free electronics at this time.
6. Tin-lead solder does not assure defect-free electronics with perfect reliability.
As I type this post, I am surrounded by more than 20 RoHS-compliant products, some dating from 2005. Outside my office, at Dartmouth’s Thayer School of Engineering, is our computer center. The thousands of RoHS-compliant products that the computer center buys each year (they get me my laptop, etc) are almost all RoHS-compliant. No unusual reliability issues have been noted.
Bob mentions that CALCE reports that 31% of laptops fail in 3 years. This number actually seems low to me. Upon reading the paper, one finds that over 10% of the 31% is due to accidents.
A study of 100,000 hard drives at Google suggests that hard drive fails are in the 5% range per year, which may account for much of the 20% of fails in 3 years. But what solid conclusion can be made from these data? Nothing, unless failure analysis is performed.
The sky is not falling. Lead-free has process-ability and reliability challenges, such as graping, head-in-pillow, voiding, etc. With data-driven process optimization at all steps in the manufacturing of the ICs, components and assemblies, good lead-free yield and reliability can be achieved.
Lead-free is here to stay. It is up to us to perform the experiments and develop the techniques to assure that RoHS compliant products have acceptable reliability.
Bob's comments follow:
My source for the dead vehicles that arrive at car dealers having whisker problems, comes from my former professor of physics, Dr. Henning Leidecker at NASA Goddard Space Flight Center in Greenbelt MD. Dr. Leidecker said that in the last four years his office has been contacted by seven major suppliers of automotive electronics inquiring about failures in their products caused by tin whiskers. He said his office has contacted Toyota offering to help analyze its acceleration problem, but hasn't heard back. For full context, read the rest of the article [http://wtop.com/?nid=108&sid=1898265].
Ron Lasky confirms that parts plated in pure tin will grow tin whiskers "with a certain amount of aging". According to NASA, whiskers can grow in hours, days, weeks, months or years. It depends on at least six factors; the quality of the tin plating, the residual stress in the coating, was the coating annealed or not, grain uniformity, temperature, humidity, and unknown other factors we don’t yet understand which is what makes it so difficult to stop whiskers from growing and is why there are so many papers published on the subject (as you can clearly see at John Barnes website) yet we still do not understand why or how they grow.
So yes, is entirely within the realm of possibility that "new" products have failed due to tin whiskers or perhaps dendritic growth.
NASA cannot tell us who the manufacturers are who reported these events due to confidentiality agreements. Dr. Leidecker says they get these calls from other industries as well and most request a non-disclosure agreement. NASA feels it’s better to get some information rather than none, don't you agree?
Last week at CALCE at UMd. it was reported that 31% of all laptops fail within 3 years. This is the link to the report http://www.squaretrade.com/pages/laptop-reliability-1109/ No information is given as to what has failed. Is it due to whiskers? We do not know.
What we do know is that the laws of physics have not been repealed. Tin will most certainly grow whiskers so using leadfree solder and tin plated components has to result in tin whiskers growing.
NASA continues to log failures. NASA Goddard is now studying the Toyota incidents for NHTSA. Again, a non-disclosure statement has been signed so they cannot comment on the study at this time.
Dr. David Gilbert of Southern Indiana University has demonstrated that a low resistance or shorted input between the wires from the pedal electronics to the electronics control module will cause Toyotas to open their throttles full. Perhaps the problem is due to leadfree manufacturing (which Toyota admits it began in 2002-3)? Perhaps it is software? We don't yet know but we can be reasonably certain that not all the accidents are caused by the owners of the vehicles. You can see pictures of the Toyota parts at my website [www.hlinstruments.com//RoHS_articles/Toyota/] The pedal has a pc board layout that I would have been comfortable with. In particular, the SOIC part that converts the signals from the Hall effect sensors (that sense pedal position) into 1-5Vdc signals sent to the electronic control module is very close to the edge of the board. The board has serrated edges which indicates it was snapped out of a large panel of these boards after the parts were soldered to it. It's possible a trace or lead has fractured or one of the capacitors or resistors. We know that leadfree solder is more brittle than tin-lead. Perhaps a few boards are marginal and over time a lead opens or becomes intermittent?
The EU was warned that tin whiskers and brittle joints would result if lead was banned from electronic assemblies but went ahead and banned lead from tin-lead solder and platings on parts. They acknowledged the possibility of reduced reliability under intense pressure from hi reliability industries and did exempt some products (military, aerospace, etc...). What difference did it make since the majority of component manufacturers refused to continue to offer tin-lead plated leads? That is why NASA replates it's components with tin-lead at Corfin Industries and uses only tin-lead solder.
The image is a Toyota accelerator pedal position sensor board from Toyota Gilbert Report.