A daisy-chain reliability study
BY AVIROOT KONGCHAROEN, SUWAN TRONGJITWIKRAI, SOMMAI NETPHU
Plastic ball grid array (BGA) packages have gained popularity because of trends in product miniaturization and enhanced in-package electrical performance. The BGA package takes advantage of the area underneath the package for solder ball interconnections, which provides higher interconnect density than perimeter leaded packages. Solder ball attachment also gives lower parasitic inductance than leaded attachment. However, when a BGA package is mounted to a printed circuit board (PCB), the solder ball connections are no longer visible, in contrast to perimeter leaded package connections.
 Figure 1. 64-pin LFBGA used for the evaluation. |
null
Any mismatch in coefficient of thermal expansion (CTE) between the BGA package and the PCB will cause stress on the solder joint. In the extreme case, the solder joint may crack and result in an electrically open circuit.
BGA Test Structure
To understand the solder joint reliability, a daisy-chain structure was designed and built to evaluate the microstructure and thermal mechanical stress on the solder joint under thermal cycling (-65°C to 150°C). Resistance was measured at every 100 temperature cycles, and failure was defined as any solder joint resistance change of more than 20 percent of its value before any temperature cycles. A 64 I/O plastic low-profile, fine-pitch BGA package (LFBGA) was used for this evaluation. The package details were 8.0 x 8.0 mm body size, 0.8-mm ball pitch, 0.4-mm diameter 63Sn:37Pb solder balls, 1.4 mm package height, 0.27 mm ball height, and a 2-layer bismaleimide triazine (BT) resin substrate (Figure 1).
Solder Joint Inspection
As mentioned, standard visual inspection methods cannot be used to ensure the quality and reliability of the solder joints in BGAs. Interfaces between solder balls with solder ball pads on the substrate and the solder ball pads on circuit board cannot be seen visually.
There are at least two methods to use for inspecting solder joints that are not accessible with optical means. One is the acoustic scanning technique. Scanning acoustic microscope (SAM) equipment converts acoustic characteristics of ultrasonic waves (15, 25, 50,
 Figure 2a. LFBGA image by scanning acoustic microscope equipment. |
null
 Figure 2b. LFBGA image by real time X-ray equipment. |
null
100 MHz or higher) traveling through each material to create an image based on the different acoustic impedance of materials involved (Figure 2a). Another technique is based on X-rays. Real-time X-ray imaging equipment can be used to see solder joints that are not at the edge of the package (Figure 2b).
Both of these techniques can be used to inspect solder joint interface quality. However, both transmission X-ray and SAM techniques do have their limitations when inspecting for solder joint cracking. The effectiveness of SAM equipment is dependent on the properties and geometries of the test specimens, along with the structure and magnitude of the features of interest, such as a crack or void at the solder joint. The transducer selection, parameter setting and equipment capability also are key factors for the effectiveness of SAM equipment. To detect solder joint micro-cracking, a high-frequency transducer needs to be used to detect the air gap created by micro-cracking. With a high-frequency acoustic wave, however, the transmitted signal may be absorbed by the molding compound or the PCB. As a result, there may not be sufficient detectable signal magnitude to be reflected by the micro-crack. The performance of X-ray equipment in this application depends on equipment capability and properties and geometries of the specimens. Transmission X-ray detection depends on the material density variation within the test samples, so it may be difficult to detect solder joint micro-cracking.
Daisy-chain Design Concept
A solder joint reliability study was designed by taking into account the ball distribution pattern compared to the package and die structure. The die size in the 8 mm square LFBGA was 3.973 x 3.720 mm. Solder joints were categorized into three groups — inner die perimeter (I), die perimeter (DP) and outer die perimeter (O) — as shown in Figure 3a. Each solder joint group was connected in a daisy-chain form (Figure 3b).
 Figure 3a. The arrangement of solder balls on the test package. |
null
 Figure 3b. A daisy chain connects in each group of solder balls. |
null
The LFBGA test specimen was prepared by wire bonding between each couple of lead fingers. The daisy-chain test board core material was FR-4, with a 25.4 x 25.4 mm board size and a finished board thickness of 0.434 mm. (Recall that the LFBGA substrate core material is BT and the finished substrate thickness is 0.32 mm.) There were three test points, O, DP and I, plus a common for testing each solder joint connecting loop. The LFBGAs were assembled to the test board by a surface mount process.
Test Conditions and Results
The experiment was designed to study the stress imposed on the solder joints during thermal cycling, by measuring resistance changes between each test point (O, DP and I) with a common point on the test board after a certain number of thermal cycles.
Test specimens were measured for electrical resistance at ambient temperature before starting temperature cycling. The duration of each cycle was 30 minutes, with temperature varying from -65° to 150°C. Test specimens were pulled from the temperature cycle chamber and resistance measurements taken at an interval of 100 to 500 cycles. Also in this evaluation, two BT substrate suppliers were compared, each with 20 test specimens.
Summary
The daisy-chain test method is a simple but effective method for studying solder joint reliability. This approach is useful for product development because of the limitations of the various inspection methods. It can be applied to several package types, and the daisy-chain design should take into account all package structures, including I/O shape and pattern, substrate, package body, and internal package structure.
Daisy chain is one of several methods to study solder joint reliability. However, each method has its strengths and weaknesses, so one must understand the objectives of the evaluation to select the right approach for the experiment.
Suwan Trongjitwikrai, executive vice president of quality/reliability assurance and new product developments, Sommai Netphu, section manager of special products, and Aviroot Kongcharoen, senior engineer of special products, can be contacted at Circuit Electronic Industries Public Co. Ltd., 45 Moo 12 Rojana Industrial Park, Thambon Thanu, Amphur U-Thai, Ayutthaya 13210, Thailand; 6635-226280-9; Fax: 6635-226710; E-mail: [email protected], [email protected].