End-use product failures represent the highest risk and highest cost to an original equipment manufacturer (OEM), particularly when a failure occurs at the printed circuit board (PCB) level. This risk has increased significantly in recent years for two main reasons:
- Complex Structures: Board designs are more complex than ever before, with smaller attributes and geometries. The use of sequential lamination, multi-level microvias, planarization of layers, and buried, blind, filled, and capped vias is becoming more prevalent, resulting in complex structures. Simply put, there’s more that can go wrong.
- Mass Production: Driven by cost reduction, a shift has occurred in PCB production from engineered, industry-specialized production done in Europe and North America to low-touch, high-volume done in Asia. With substrate manufacturers now producing everything from the simplest to the most complex products, they must use the widest possible latitude in equipment, chemistry, and materials.
Board quality assurance, then, is more critical than ever before. But, lead-free assembly makes reliability testing on its own insufficient for today’s PCBs. During lead-free assembly a PCB is exposed to a number of thermal processes that weaken copper interconnects and dielectric materials. While these assemblies may pass reliability tests with flying colours, this is no guarantee that they won’t fail in the field.
Survivability testing is now necessary to prevent the risk of in-field failures and the associated costs. Even before long-term reliability is established, it must be confirmed that via structures and base materials can survive assembly, rework, burn-in, and end use.
How Lead-Free Changes the Game
In the past, electrical characteristics dominated designers’ substrate selection processes. Today, however, three critical base properties must be considered when selecting the most optimal base material for a PCB substrate for lead-free manufacturing. These are:
- Decomposition temperature (Td)—the temperature at which the material starts to decompose
- Coefficient of thermal expansion (CTE)—specifies the rate of x, y, and z expansion per degree Celsius, which creates the load, or strain, on the plated structures.
- Moisture absorption—determines the material’s relative moisture capacity (not to be confused with the moisture remaining/entrapped from the manufacturing process).
Interconnect stress testing (IST) reliability studies have shown that well-made, high-mass test boards may experience a performance reduction of 50% following exposure to between four and six lead-free assembly and/or rework cycles. If fabrication problems are present (e.g. minimum plating thickness, poor copper distribution, or low-grade materials) cycles to failure are reduced to two or three.
The effects of higher temperatures in lead-free assembly and rework increase these risks. For example, delamination is becoming more common. With well-produced substrates, the common reduction in cycles to failure is 20%-40%; for a poorly fabricated PCB, this is reduced by an additional 40%. Three or six thermal excursions at lead-free temperatures (3x or 6x 260°C) reduce the thermal cycles to failure between 30% and 50%, and up to 50% for a poorly fabricated PCB.
Plated structures are the dominant failure mode, with quality of base materials also influencing reductions in performance. Additional material damage complicates matters because stress-relieving effects on the plated structures will actually increase their perceived performance. Put another way, a poorly fabricated PCB can demonstrate increased performance after assembly and rework despite containing material degradation.
Copper quality, material robustness, and design must all be in balance to produce reliable product. For lead-free applications, the primary consideration is to achieve a balance between copper plating, materials, and design to assure reliability.
Survivability Testing: How It Works
The PCB substrate, or bare board, is the foundation of a finished product’s reliability. Typically, reliability data is provided in the form of micro-section reports, process control charts, impedance data, and/or thermal oven tests. But none of these can demonstrate that the product will survive assembly or in-field usage.
Today, high-temperature testing is required to demonstrate not only the reliability of copper interconnections but also the robustness of dielectric materials. Survivability testing uses maximum temperatures in the range of 220°C to 260°C to represent the thermal stresses experienced in the elevated temperatures of lead-free assembly and/or rework cycles. (Leaded assembly is commonly heated to 220°C to 230°C).
The objective is to quantify whether the plated structures, internal interconnect structures, and/or base materials can withstand the anticipated number of required cycles, both during and after the elevated temperatures. The number of cycles is determined by the complexity and maturity of the product—commonly between two and six cycles.
Ideally, reliability testing begins prior to exposure to simulated assembly to establish a baseline or reference for comparison purposes. The test vehicle should contain multi-purpose “intelligent” circuits that deliver important product construction and process control information. The data can be collected electrically for analysis of control for both product and process.
The responsibility for the survivability testing protocols belongs solely at the PCB manufacturing level; the assembler “assumes” the product has been sufficiently engineered and tested, to confirm the product contains the right material to withstand multiple passes through their equipment and potential rework operations.
Increasing test temperatures changes the fundamental understanding of measuring overall product performance. Higher temperatures are suitable for determining if vias and materials can survive multiple exposures through assembly level temperatures, but are not sufficient to assess the long-term reliability (fatigue, wear-out, accelerated life, etc.) of a product.
In traditional tin/lead assembly, the influences in failure are recognized to be copper quality first, then material robustness, then design. But the temperatures associated with lead-free assembly shift this hierarchy and require a balance between copper quality and material robustness. This in turn requires effective testing protocol to quantify the impact on product performance of high temperatures: survivability testing.