Authored By:
Tae-Kyu Lee, Gnyaneshwar Ramakrishna
Cisco Systems
CA, USA
Jonghyun Nam, Daljin Yoon, and Heera Roh
SK Hynix
Icheon, Korea
Summary
The increase of functionality and performance in electronic systems and devices induce more complex printed circuit board design features with more Cu layer counts, increased thickness, and with a demand for less impact to the high-speed signals. These combinations drive the PCB design to configurations including via-in pad platted over (VIPPO) with back-drill process, which resulted in a mixture of various stress distributions in a given printed circuit board.
Recent studies revealed that these mixed design features can induce failure during a second reflow at the solder joint interface known as solder separation. Through a series of in-situ resistance measurements during reflow, the accurate separation location and temperature are identified per design configuration and solder joint material combination. A 10x15mm BGA component with 10x20 array solder balls were assembled on 28 Cu-layered 140mil thickness boards, which contain a mixture of VIPPO and VIPPO+back-drilled pads compared to a full VIPPO pad design as a baseline.
The assembled boards were subject to in-situ resistance measurements during second reflow. The segmented daisy chains per component allowed the detection of solder separation during reflow at the exact solder joint location, which are associated with the failure symptoms. The failure signature was detected at 210-212oC during heating cycle at solder joint locations, which are adjacent to solder joints with higher z-axis expansion, in this case full VIPPO location adjacent to back-drilled VIPPO pad locations. Three different solder alloys were selected to investigate the sensitivity to solder separation defects along with the evaluation per long-term thermal cycling performance.
The results indicated that the separation defect in SAC305 can be mitigated by using SAC305+SnBiAg low melting temperature (LTS) hybrid solder joint configuration and full LTS configuration. The presented study elaborates the parameters associated with the solder separation failure per board design configuration, and provides the potential solution to mitigate the failure. An in-depth analysis on the solder joint microstructure utilizing Electron-backscattered diffraction (EBSD) supports the test results leading to the mechanism associated with the solder separation defect failure.
Conclusions
Through a series of in-situ resistance measurement during second reflow, the accurate separation time and temperature are identified per design configuration and solder joint alloy material combination. The solder separation defect induced failures detected at 210-212oC at full VIPPO solder joint locations, which are adjacent to solder joints with higher z-axis expansion, in this case, back-drilled VIPPO and dog-bone/floating pad locations. Three different solder alloys were selected to investigate the sensitivity to solder separation defects along with the evaluation per long-term thermal cycling performance.
The results indicated that the separation defect in second reflowed SAC305 joints can be mitigated with using the SAC305+SnBiAg hybrid solder joint configuration and full LTS configuration since it has lower than 212oC assembly peak temperature. Thermal cycling performance after second reflow also shows the hybrid LTS is not affected by the uneven VIPPO distribution induced stress conditions. The presented study elaborates the parameters associated with the solder separation failure per board design configuration, and also provides the potential solution to mitigate the failure. An in-depth analysis on the solder joint microstructure utilizing Electron-backscattered diffraction (EBSD) supports the test results leading to the mechanism associated with the solder separation failure.
|
