Ask the Experts
November 6, 2018
BGA reballing question
We are a mil contractor and have requirements to have BGA components reballed. The BGA components are supplied with SAC305 solder and will be reballed using tin-lead solder. Our reball service supplier uses a single dynamic solder wave to strip the lead-free solder balls prior to application of new tin-lead solder balls. This dynamic wave contains 63-37 tin-lead solder.
Is a single wave stripping process acceptable, or should the reball service supplier run the BGA components through a second pure tin-lead wave to ensure the any lead-free solder residue has been sufficiently purged prior to reballing?
Expert Panel Responses
In my opinion, the additional thermal cycle would have much more negative impact to the BGA than would the residual Silver and Copper from the original SAC305 interconnects. The residuals would not be significant.
Principal Product Engineer
Benchmark Electronics, Inc.
27 years experience working with electronic and electro-mechanical manufacturing and design (medical, automotive, military, computer, and industrial controls). Military veteran - served as a Combat Engineer with the United States Marine Corps.
The concern of course is, as the alloy in the stripping pot gets contaminated with Pb-free solder, the remaining solder will also be contaminated. One way to look at this would be to examine the extreme condition where the entire contents of the stripping pot were replaced with the Pb-free alloy. Making some generic assumptions on pad geometry and ball size using typical dimensions for a 0.050" pitch BGA:
Based on the above assumptions, I calculate that in the worst case the final ball composition would be:
- Pad diameter of 0.018"
- Average solder thickness on pad after removal of 0.0007"
- Replacement ball diameter of 0.025"
The contamination level for Ag would be at the maximum per J-Std-001, and the level for Cu would still be still acceptable. The composition of Sn and Pb would shift more than allowed under J-Std-006, however. The maximum allowed deviation for Sn and Pb is 0.5%. The observed shift would be a little over 2 times the allowed shift, so the pot would need to be contaminated severely (About 40% Pb-free alloy) in order to take the final alloy out of specification. This is conservative, as I have (intentionally) not accounted for the solder paste volume in final assembly which would shift the alloy back toward nominal to a small degree. Additionally, if the re-baller is using an air knife, the remaining solder thickness may be smaller than in the above example.
Fritz's career in electronics manufacturing has included diverse engineering roles including PWB fabrication, thick film print & fire, SMT and wave/selective solder process engineering, and electronics materials development and marketing. Fritz's educational background is in mechanical engineering with an emphasis on materials science. Design of Experiments (DoE) techniques have been an area of independent study. Fritz has published over a dozen papers at various industry conferences.
I have not seen this method of removing excess solder prior to reballing but the use of flux sprayed to the surface and dynamic wave action will remove any excess solder. To do it twice is something that needs to be checked out to see what qualify process control they use to determine a single pass is sufficient.
President and Founder
Contract Manufacturing Consultants, Inc.
Robert Fried helps leading electronics OEM's develop world-class sourcing strategies for PCBA, cables, precision metals, plastics, modules and complete end-products. Other service areas are supplier risk assessment, comprehensive outsource ...
The SAC305 (96.5Sn/3.0Ag/0.5Cu) overwhelmingly consists of tin. The material that primarily creates the bond between lead free and leaded solder is Sn. So in your supplier's process, the material you are putting back on the BGAs is going to have a different percentage of Sn/Pb. As the BGAs are being processed the percentage of Sn/Pb in the pot is changing. The volume of solder in the "pot" compared to the volume of solder on the BGAs is critical, because if all things were equal your new solder material would be 79.75% Sn, 18.5% Pb and 1.75% others. But they are not equal, because the solder bath volume is much higher than the combined volume of all the BGA balls.
If you are concerned, I recommend getting the balls (not the pot) material tested post processing to verify Sn/Pb + composition. I would test them at the end of the run, as the composition of the solder is changing during processing. Hope this helps.
Mr. Harman is an SMT Engineer at ACDi and worked at Oven Industries as a Manufacturing Engineer. H's worked at Philips manufacturing ultrasound probes/circuitry. He holds three patents, one for an ultrasound probe design, and two for innovative rat/mouse zapper circuits. He attended Pennsylvania State University for Electrical Engineering.
What needs to be accomplished is the creation of a proper SN63 intermetallic. I would think that one pass will be sufficient if the wave has enough agitation to "scrub" the surface. Really the only way to tell for sure is cross sectioning and inspection of the intermetallic.
Kelly Atay, Sunstone Circuits
While a single dynamic wave can be used to remove lead-free BGA solder balls before the re-balling process, it is recommended to utilize a dual dynamic wave process to ensure that all lead-free solder is completely removed along with any organic residues, and that flux buildup and accumulation of gold are not transferred to the second pot. The ANSI/GEIA-STD-0006 standard mandates that all automatic solder dipping apparatus and semi-automatic solder dipping apparatus shall use a dual solder pot system for this specific reason to avoid cross-contamination.
Carlos Bouras is the General Manager of Nordson SELECT and has over 30 years of experience in the electronics manufacturing industry. Carlos's expertise is in process engineering, product development and manufacturing operations. For the past 15 years Carlos has focused specifically on automated assembly issues and is the holder of several US patents for non-contact dispensing and precision dispensing of adhesives for the packaging of microprocessor devices.