What is the actual risk of tin whiskers?

• One of the more controversial issues in regards to Pb-free. Device manufacturers using palladium-based plating, some companies involved in high-end, high-density designs, and military/avionics OEMs promote tin whiskering as a major risk. The majority of the electronics industry is proceeding on the assumption that the risk of tin whiskering is relatively low, but with various degrees of uncertainty.
• The decision on how to mitigate your risk is most dependent upon the degree of control you maintain over your supply chain and the level of resources available in-house. For more information, please see our webpage on tin whiskers.

Can I reflow Pb-free BGAs through an eutectic reflow process?

• During the initial wave of interest in Pb-free solder, several organizations demonstrated potential issues with this approach. However, several organizations, primarily telecommunications and military, have decided to take advantage of exemptions or customer indifference and stay eutectic as long as possible. Unfortunately, area array devices with eutectic solder balls are becoming obsolete. Rather than deal with reballing or redesign, these OEMs have decided to reflow Pb-free BGAs using eutectic solder paste through a modified eutectic reflow process.
• Will this work? Extensive testing has indicated that the solder joint must experience a minimum peak temperature of 232^oC (more). More extensive information on assembly and reliability is currently being established by DfR and its partner companies. For more information, feel free to contact us by phone or e-mail.

Where can I get the latest information on wave soldering?

• Wave soldering, unfortunately, has become the poorer cousin of electronics manufacturing. While understanding of the appropriate reflow profile has reached the point to where it is almost standardized in J-STD020D, wave soldering has received much less attention. This is unfortunate, as the issues involved with wave soldering are much more severe. They include the capital cost of retrofitting your existing machine or purchasing a new one (Pb-free solder rapidly corrodes the standard stainless steel tanks and clogs nozzles). It also requires changing the spacing between the two standing waves. It is our experience that the best sources for information on this issue are Engent and Speedline Technologies.

Which board plating should I choose?

• The current trend in the industry is a movement to primarily immersion tin or organic solderability preservative (OSP). This is primarily due to the uncertainty regarding black pad and electroless nickel/immersion gold (ENIG) and champagne voiding from immersion silver. However, the trend is not consistent across all companies and for the most part, OEMs will tend to use the board plating recommended by their electronic manufacturing services (EMS) provider or board manufacturer.

Will my current qualification test plan need to change?

• The short answer is no. Most companies simply test to specification, and none of the current industry test specifications have changed in response to the change to Pb-free solder.
• However, if your organization takes a best practice approach to reliability and tailors your product qualification to a test-to-life approach, then the answer is yes. Pb-free solders have different behaviors and therefore a different acceleration factor than eutectic solder. In addition, there are some current concerns regarding the observation of Kirkendall voiding between the copper plating and intermetallic layer (UIC).
• Here is a quick overview:
  • Preconditioning: Because of this observation of intermittent voiding, some degree of elevated temperature preconditioning may be appropriate. Initial experiments found voiding at levels sufficient to allow for premature failure in drop testing after 100 hours at 150^oC or 20-40 days at 100^oC. Assuming Arrhenius behavior, which is somewhat valid because this is a diffusion-driven mechanism, this is equivalent to 1 to 5 years at 40^oC (2 to 15 years at 25^oC).
  • Temperature/Power Cycling: Requires an acceleration factor. At this time, to the best of DfR‘s awareness, the only organization offering the ability to compute an acceleration (which takes into account dwell time) is EPSI.
  • Vibration: Based on preliminary investigation by DfR.
  • Temperature/Humidity/Bias: Very important, given the need for more aggressive flux formulations. However, the actual conditions can still be the same, which should be 40^oC/93% RH for approximately 3 to 7 days (longer if any conformal coating or potting compound is used)
  • Mechanical Shock: Most mechanical shock test specifications assume failure through an overstress condition. That is, the product either fails or it doesn‘t, but it retains no memory/damage from the event. From this viewpoint, the test does not need to change, but this is the one test in which preconditioning should seriously be considered.

to top

Will I need to redesign my board for Pb-free manufacturing?

• As with the prior questions, the short answer is no. The current bond pad designs should provide a nominal solder joint in well qualified reflow and wave processes. Board level plating will be exposed, as the Pb-free solder will not spread like eutectic. For most applications, this is not considered a reliability issue. If you are uncomfortable with the presence of exposed plating (or copper, in the case of OSP), you could shrink the current bond pad area to ensure more uniform coverage.
• However, a more accurate answer is maybe. The maximum ceramic capacitor that should be subject to wave soldering should be reduced by one case size (Sn/Pb: 1206; Pb-Free: 0805). In addition, movement from HASL to OSP may require a redesign to ensure adequate test coverage (ICT should be performed on test points rather than on the bond pads covered with OSP). Talk to your board shop or test engineer about this issue.

to top

Do I have to choose the SnAgCu alloy as my Pb-free solder?
Not necessarily. There are other alternatives on the market. The primary alternatives tend to include bismuth (Bi), indium (In), or nickel (Ni). Examples include:

• SnAgBi
• SnAgIn
• SnAgBiIn
• SnNiCu (SN100)

The exact alloy compositions can vary widely, except for SnNiCu. It is dependent upon the desired characteristics, primarily the melt temperature and degree of “shininess”. Companies wanting a “drop-in” replacement have selected SnAg-based alloys with higher concentrations of Bi and In (up to 6 wt%). Lower concentrations result in melt temperatures closer to SnAgCu, but with potentially a better look and feel. SnNiCu has especially shown improved manufacturability and shininess in regards to wave soldered interconnections (more).

Except for SnAgCu, however, none of the alloys have the test-to-field correlation necessary to provide some degree of assurance of design life. If solder joint robustness is not a concern for your product, then other alloys can definitely be considered. However, if solder joint wearout is something you wish to avoid and may experience, the other alloys are currently too risky to implement without the necessary reliability models.