Dennis Lau

A new method to evaluate BGA pad cratering in lead-free soldering

Pad cratering is one of the most important issues in lead- free soldering. The PCB materials change and the increased reflow temperature account for the degradation of the drop test performance of the PCB materials. This paper uses both finite element analysis (FEA) and experimental testing to study PCB pad lift and adhesion. The pad lift force in pull and shear tests has been determined with the test data, and a new cratering test using ball shear test was proposed for qualifying PCB BGA pads incorporating multiple reflows and mechanical loading.

Experimental Testing and Failure Prediction of PBGA Package Assemblies under 3-Point Bending Condition through Computational Stress Analysis

Reliability test is a very important step for any electronic products before they can be sold to the market. In the previous decades, various types of test have been developed. One of the widely used testing methods is the board bending test. A major issue of performing a reliability test is that numbers of testing samples are required since the test has to be repeated until consistent and satisfactory results can be obtained. A long time will be spent on the specimen design, the assembly processes and the testing procedures. In case the reliability test is not satisfied, the specimen will need to be redesigned, reproduced and retested. This is costly and not efficient from the commercial point of view. To simplify the reliability testing process, computational simulation is proposed to predict the reliability of printed circuit board assemblies (PCBA) at the early design stage. In the present study, 3-point bending test is the reliability test adopted. Finite element model validation and two case studies regarding the application of the model are the objectives of this work. Comparison between experimental results and computational results is carried out for model validation. The physical properties (stiffness and strain) of the model show a good agreement with the experiments at both the elastic deformation stage and the plastic deformation stage under the 3-point bending condition. The maximum von Mises stress at the solder joints of the model at which the samples are found to be failed is also very close to the strength of the solder. This further validates the model. For application, this stress data is used as a failure criterion to indicate whether a PCBA is reliable under different bending conditions

Thermal-fatigue life prediction equation for wafer-level chip scale package (WLCSP) lead-free solder joints on lead-free printed circuit board (PCB)

A new thermal-fatigue life prediction equation for a class of lead-free solder alloys, 95.5wt%Sn3.9-4.0wt%Ag0.5-0.6wt%Cu is proposed in this investigation. The test vehicle consists of a lead-free solder-bumped WLCSP, lead-free PCB, and lead-free solder paste. The coefficients of the fatigue equation presented herein are determined by best fitted with the isothermal fatigue data of the test vehicle. Failure modes and locations of the failed samples are discussed.

Impact Performance of Microvia and Buildup Layer Materials and Its Contribution to Drop Test Failures

This paper introduced a new method to investigate the fracture dynamic response of the PCB materials. Ball impact is used to evaluate the performance of PCB laminate and prepreg materials. Various PCB materials and microvia structures are evaluated which compares halogen free (HF) vs. non-HF, and leadfree compatible vs. non-leadfree compatible, and RCC (resin coated) vs. laser drillable prepreg. It is found that pad cratering including cracking of buildup layer and copper pad delamination are effectively screened out by the impact test. Opportunity or severity of cracking and delamination are dependent to the impact energy and as well as materials of PCB. Materials especially halogen free (HF) and non-filler buildup layers are much more tendency to crack.

Computational model validation with experimental data from temperature cycling tests of PBGA assemblies for the analysis of board level solder joint reliability

A previous experimental study was conducted to characterize the thermal fatigue life of PBGA solder joints under temperature cycling condition. The temperature cycling profile was one-hour cycles with a temperature range from -40/spl deg/C to 125/spl deg/C. The thermal cycling test lasted for more than 3,000 cycles and the Weibull characteristics were obtained. In the present study, a computational model is established for the analysis of solder joint thermal fatigue life. The current approach is a finite element based analysis and the previous experimental data are used for model validation. In this paper, the results from two types of plastic ball grid array assemblies are presented. Both eutectic Sn-Pb and Pb-free solders are studied. The comparison between computational and experimental results shows that the two dimensional finite element analysis appears to be sufficient for estimating the thermal fatigue life of solder joints in plastic ball grid array assemblies.

Computational analyses on the effects of irregular conditions during accelerated thermal cycling tests on board level solder joint reliability

This study is intended to investigate the effects of irregular conditions on the board level solder joint reliability during accelerated thermal cycling tests. In a previous research, the thermal fatigue lives of PBGA solder joints have been estimated using a 2-dimensional finite element model. The computational results agreed very well with the experimental data. In the present study, with the previously developed finite element model, three cases of parametric studies are performed. In actual thermal cycling tests, it is usually quite difficult to achieve the ideal temperature profile in four linear segments as ramp-up, high temperature dwell, ramp-down, and low temperature dwell. The first case of the present study is to investigate the effect of imperfect temperature profile on the thermal fatigue life of PBGA solder joints. The second case under investigation is the effect of service interruption of the thermal cycling machine during the test. A certain time period of constant room temperature is inserted in the middle of the original temperature profile to simulate the breakdown of thermal cycling machine. In the third case, different starting temperatures are specified in order to simulate the effect of different stress free temperatures. The results of various computational analyses are compared and discussed in details

Correlation Between the Strain on the Printed Circuit Board and the Stress in Chips for the Failure Prediction of Passive Components

Due to the demand for miniaturization of microelectronic devices, the density of packaging has become higher and higher. Also, the sizes of components have become smaller and smaller. In addition to advanced active components such as chip scale packages (CSPs) and flip chips (FCs), mini sized passive components such as chip capacitors and resistors are also important elements for high density packaging. It is quite common to see dozens up to hundreds of passive components on printed circuit boards (PCBs). Both active and passive components contribute to the function (and also malfunction) of electronic systems. However, the reliability issues of passive components are often overlooked because they are relatively small in size and cheap in cost. In view of the fact that “small components could lead to big problems”, the present study is conducted to evaluate the threat to passive components assembled on PCBs under a specific type of mechanical loading. Because of the nature of mass production, microelectronic devices are always manufactured in a batch mode. It is quite often that several PCBs are linked together during the surface mount assembly process. Even if the PCB is a stand-alone unit, extra peripheral frames or tie bars are needed for tooling and fixture. After the board level assembly, a depaneling process is usually required to singulate individual PCBs or to remove the tooling frames for the system level assembly. Some depaneling processes may be automated with precision control. However, it is not unusual for operators in the factory to perform manual depaneling. During this process, the PCB is subjected to mechanical bending and the curvature of the bent PCB may be big enough to damage small passive components. The present study is intended to establish a model for the failure prediction of passive components under depaneling load condition. Computational stress analysis is performed with a 3D finite element model. The emphasis is placed on finding the correlation between the bending strain on the PCB (which is an index of the local curvature of the bent PCB) and the bending stress in the passive components (which is the reason to crack capacitors/resistors). It is observed that such a relationship can be established. With this model, the cracking of passive components may be predicted under the depaneling load condition. The understanding of this potential threat can be turned into a design rule to avoid mounting passive components in the “high risk” area on the PCB. As a result, the objective of “design for reliability” (DFR) can be achieved. The details of the aforementioned model and the results of stress analysis will be presented in this paper.Copyright

A SYSTEMATIC APPROACH FOR DETERMINING THE THERMAL FATIGUE-LIFE OF PLASTIC BALL GRID ARRAY (PBGA) LEAD-FREE SOLDER JOINTS

Reliability of plastic ball grid array (PBGA) SnAgCu lead-free solder joints is investigated. Emphasis is placed on the design for reliability (DFR) of lead-free solder joints. In particular, the thermal-fatigue life of the lead-free solder joints of a PBGA package assembly is predicted and compared with thermal cycling test results.Copyright

Thermal-Fatigue Life Prediction Equation for Plastic Ball Grid Array (PBGA) SnAgCu Lead-Free Solder Joints

A new thermal-fatigue life prediction equation for a class of lead-free solder alloys, 95.5wt%Sn4.0wt%Ag0.5wt%Cu, is proposed in this investigation. The test vehicle consists of a lead-free solder plastic ball grid array (PBGA) package, a lead-free PCB, and lead-free solder paste (95.5wt%Sn3.9wt%Ag0.6wt%Cu). The coefficients of the fatigue equation presented herein are determined by best fit of the test vehicle’s isothermal fatigue data. Failure modes and locations of the failed samples are discussed.© 2005 ASME

Isothermal fatigue tests of plastic ball grid array (PBGA) SnAgCu lead-free solder joints at 60/spl deg/C

A new thermal-fatigue life prediction equation for a plastic ball grid array (PBGA) package with 95.5wt%Sn4.0wt%Ag0.5wt%Cu lead-free solder balls is proposed in this investigation. The test vehicle consists of the PBGA package, a lead-free PCB, and a lead-free solder paste (95.5wt%Sn3.9wt%Ag0.6wt%Cu). The coefficients of the fatigue equation presented herein are determined by best fit of the test vehicles isothermal fatigue data tested at 60degC. Failure modes and locations of the failed samples are discussed