Advanced Through-Hole Rework of Thermally Challenging Components/Assemblies: An Evolutionary Process, Part 2

Estimated reading time: 9 minutes

Rework Study

A rework study was conducted on the PCBRM100 to assess its capability to effectively rework challenging PTH components on high thermal mass assemblies. The test vehicle (TV) was a 180 x 200mm (7.1x7.9”), 3.3mm (.130”) thick 12-layer board with 13 ounces of copper and an OSP finish. PTH components on the TV include electrolytic caps, headers and two 140mm (5.5”) DIMM connectors as shown in Figure 15. The DIMM connectors were chosen for the rework study due to their size, thermal mass, number of pins and known issues regarding connector body temperature.

Figure 15: Test vehicle used in reword study.

SN100C solder was used due to its previously documented lower copper dissolution rate during mini-pot rework compared to SAC305. SN100C has a melting temperature of 227°C which is ten degrees higher than SAC305 (217°C). The solder pot temperature used in the study was 272°C which is 45 degrees above the melting temperature. Kester RF771 tacky flux was used as its formulation is designed specifically for rework. The solder contact time during PTH rework is multiple times longer than the contact time during wave soldering, and therefore the fluxes typically used for wave soldering are not designed to withstand the rework process.

Multiple thermocouples were attached to the bottom side and top side DIMM joints as well as to the DIMM body. A baseline thermal profile and an alternate thermal profile were developed. The baseline profile preheated the entire board to 150°C. The instrumented DIMM was then immersed in solder until all top side joints reached 240°C, which took 45 seconds.

During the alternative profile, the board was preheated to 150°C just like the baseline process. However a three-minute focused convective heating (FCH) stage took place prior to immersion in solder. During the FCH stage, top side heating from the nozzle and bottom side heating from the universal heating blades increased the top and bottom side joint temperatures by approximately 55°C. The FCH stage acts like an extended soak stage where the DIMM temperature is increased and stabilized and where the core temperature of the board near the DIMM is maintained.

Immersion in solder occurred after the FCH process was complete, however the required contact time to achieve 240°C top side joint temperature was significantly shorter, in this case 20 seconds versus 45 seconds for the baseline process (see Figure 16).

Figure 16: Thermal profiles.

Initial assembly of the test vehicles was done on a standard wave soldering system using SAC305. Kapton tape was used to protect 20 bottom side joints on one end of each DIMM site from wave soldering. These unsoldered joints represented the “as-received” copper thickness for each DIMM site. An additional 20 joints were protected from the rework processes. These joints represented the “post-wave” copper thickness for each DIMM site.

A total of 26 DIMM sites were subject to a complete rework cycle that included removal, barrel cleaning and replacement using either the baseline or alternate process. Reworked TVs were sent to an independent laboratory for cross-section analysis. Twelve cross-section measurements of the bottom side knee were taken in the as-received, post-wave and post-rework sections on each DIMM site. See Table 1.

Table 1: Baseline vs. alternate (FCH) process copper dissolution results. X represents 10 data points.

Table 1 shows that the as-received copper thickness varied widely from a low of 1.6 mils to a high of 3.0 mils which in turn resulted in a wide variation of both the post-wave and post-rework copper thicknesses.

Table 2 is a summary of the average copper thickness and average dissolution based on the data points in Table 1. The key point in Table 2 is that the average post-rework copper dissolution for the alternate (FCH) process was 0.5 mils compared to 0.9 mils for the baseline process. Convectively heating the DIMM prior to solder immersion resulted in a 45% (0.5 vs 0.9 mils) reduction in copper dissolution.

Table 2 also shows that the average post-rework copper thickness for the FCH process was 1.5 mils, which is significantly above the minimum standard of 0.5 mils (12.7 microns). Comparatively, the average post-rework copper thickness for the baseline process was only 0.7 mils which is just above the minimum standard. In addition, 22% of the baseline copper thickness measurements taken fell below the 0.5 mil minimum standard while none of the FCH process copper thickness measurements fell below 0.5 mils. Tables 1 and 2 clearly show that the FCH rework process significantly increases the lead-free PTH rework process window for high-complexity assemblies.

Table 2: Baseline vs. alternate (FCH) process copper dissolution summary, in mils.

Figure 17 shows one of the worst-case results from the baseline process where 100% dissolution occurred at the knee and significant dissolution occurred at the pad and in the barrel. In contrast, Figure 18 shows the typical copper thickness after the FCH process where minimal copper dissolution has occurred at the knee as well as in the barrel and on the pad.

Figure 17: Baseline process (100% Cu dissolution at knee).

Figure 18: Alternate process (minimal Cu dissolution at knee).

In addition to copper dissolution, iNEMI also cited barrel fill as a key concern for lead-free PTH rework on Class 3 assemblies. Laboratory analysis of barrel fill on post-wave solder joints varied widely from a low of 44% to a high of 100% with an average barrel fill of 76%. In addition, 31% of the post-wave joints did not create a complete fillet where solder climbs the pin (see Figure 19). It is important to note that the pin protrusion on the TVs was virtually zero which is perhaps a worst-case scenario for barrel fill and fillet formation. However, despite this fact, the post-rework barrel fill and fillet results were excellent. Barrel fill of 100% was measured on all but four of the 156 measurements taken. In addition, a positive fillet was formed on every post-rework joint that was analyzed (Figure 20).

Figure 19: Negative fillet from wave soldering.

Figure 20: Positive fillet from rework process.

A 100% barrel fill was a key objective of the alternate (FCH) process. It was expected that focused convective top side heating of the component would significantly improve barrel fill by creating a heated upward path for the solder to follow. However 100% barrel fill was also achieved in the baseline process where no FCH was used. It was surmised that there were two reasons for the significant improvement in barrel fill during PCBRM100 rework compared to wave soldering. First, the solder contact time during rework is multiple times longer than in the wave soldering process. Second, flux was applied to the bottom side of the board, the top side of the board and onto the replacement component pins during the rework process compared to just the bottom of the board during wave soldering.

A post-rework void analysis was also performed on the 156 joints that were analyzed. Of these, 42% of the joints analyzed had zero voiding, 44% has a worst-case void diameter of 10% or less, and 14% had a worst-case void diameter of over 10%.

Unfortunately, the DIMM connector on the TV did not lend itself well to the single-cycle rework approach as the three DIMM locating holes had annular rings which collected solder. This caused the locating pins to remain behind during component removal which prevented any attempt to immediately re-insert a replacement component. A future design recommendation is for DIMM locating holes to not have annular rings.

Summary and Conclusions

The solder fountain or mini pot has been the industry standard for tin-lead PTH rework as well as lead-free rework of PTH components on low and mid-complexity assemblies. The solder fountain process has been optimized for lead-free rework by the use of lower dissolution solder alloys and by the addition of integrated preheating systems. In addition, BGA rework systems with convective and IR heating systems have been successfully used to remove lead free PTH components for applications where the optimized solder fountain process does not meet the rework objectives.

The current “lead in solder” exemption for Class 3 applications including server, storage and network infrastructure equipment is set to expire in 2014. Alternative rework solutions, including convection, IR, vapor phase and laser have been proposed, however none of the existing technologies was designed with lead-free rework of PTH components on large, high thermal mass assemblies in mind.

The PCBRM100 is a “clean sheet of paper” design approach to solving copper dissolution and barrel fill issues on Class 3 assemblies. In design and beta testing for over three years, initial production shipments will begin in the first quarter of 2013. The key to the 100 is the top and bottom focused convective heating system which significantly reduces the required solder contact time which in turn significantly reduces copper dissolution. A two-phased DIMM connector rework study on the 100 demonstrated that 100% of DIMM connectors reworked with the FCH process showed excellent results in regard to post-rework copper thickness, barrel fill and fillet formation. Void analysis showed excellent results on 86% of the joints analyzed; however some large, random voiding did occur. The combined phase one and phase two processes were based on the complete rework (i.e., removal, barrel cleaning and replacement) of 26 DIMM connectors with 52 cross-sections, 936 copper thickness measurements taken and 156 barrel fill and voiding calculations made and fillet formations assessed.

Acknowledgements

The author would like to thank the following individuals:

• Chuck Richardson (iNEMI) for use of the iNEMI 2013 Technology Roadmap
• Gordon O’Hara and Larry Yanaros (Flextronics)for test vehicle access and wave solder assembly
• Craig Hamilton (Celestica) and Alan Donaldson (Intel) for key reference articles on lead-free PTH Rework
• Brett Pennington (Endicott Interconnect Technologies) for performing all of the laboratory analysis for the rework study
• Mike Berry (Celestica) for providing initial test vehicles, lab analysis and machine design input
• Bob Farrell (Benchmark) for machine design input
• Yogesh Patel, Tan Tran and Himanshu Deo (Flextronics) for beta site testing and machine design input
• All at Air-Vac Engineering who helped him on this project.

References 

1. Joint response from EICA, AeA Europe and EECA Esia to the General and Specific Questionnaires Relating to Exemption 7b (March 2008)

2. Ibid.

3. R. Prasad, “Unresolved Issues in Lead-Free Through-Hole Soldering and Surface Finishes”

4. P. Austen (ECD), “Through Hole Component Phase Out,” Circuitnet, October 2012

5. iNEMI 2013 Technology Roadmap: Rework and Repair Sections

6. C Hamilton, “A Study of Copper Dissolution During Lead-Free PTH Rework,” May 2006

7. C Hamilton, “A Study of Copper Dissolution During Lead-Free PTH Rework,” May 2006

8. A. Donaldson, et al “Comparison of Copper Erosion at PTH Knees in Motherboards using SAC305 and SnCuNiGe Alternative Alloy for Wave Soldering and Mini-Pot Rework”

9. IPC-6012B “Qualification and Performance Specification for Rigid Printed Boards,” December 2006

10. L. Ma, et al, “Reliability Challenges of LF PTH Mini-Pot Rework” IPC/JEDEC LF Reliability Conference, Boston, April 27

11. Morose, et al, “Long Term Reliability Analysis of Lead Free and Halogen Free Electronic Assemblies”

12. iNEMI 2013 Technology Roadmap: Rework and Repair Section

13. C Hamilton “A Study of Copper Dissolution During Lead-Free PTH Rework,” May 2006

14. iNEMI 2013 Technology Roadmap: Rework and Repair Section

15. Bielick, et al, “High Thermal Mass, Very High Lead Count SMT Connector Rework Process,” 2010

16. R. Prasad, “Unresolved Issues in Lead-Free Through-Hole Soldering and Surface Finishes,” SMT Magazine

Brian Czaplicki is the director of technical marketing programs at Air-Vac Engineering. In 2012, Czaplicki participated in the development of the Rework and Repair Section of the 2013 iNEMI Technology Roadmap, which identified the key future rework and repair challenges facing the electronics industry. This article is one result of that research. He can be contacted at Brian.czaplicki@air-vac-eng.com.

.