Common SMT Process Defects to Avoid During Soldering
As more and more designs use smaller components with surface mount pads, surface mount technology (SMT) process defects can cause mounting and impact yield due to various design and manufacturing issues. These problems may have been small in the past with a negligible impact on yield, but they can lead to rapidly increasing rework costs and should be avoided. Here are some basic SMT process flaws that can be avoided during design and assembly.
Bridging between pads
Bridging is common in low viscosity solder and can cause shorts between adjacent pads. It can also occur when the temperature is outside the ideal welding range, resulting in poor wetting or excessive wicking. The key to solving this problem is to place solder mask relief around the pads (i.e. non-pad defined SMD pads or NSMD). This relief provides space for excess solder to suck in, effectively preventing solder from flowing between two adjacent pads. This is similar to a solder mask dam between a ball grid array (or BGA) component and its dog-shaped sector vias.
Side view of the NSMD pad and its solder mask relief used to prevent bridging. The gap between the NSDM pad and the solder mask provides room for excess solder.
Dehumidification
Dehumidification is a matter of choosing a solder paste. This problem rarely occurs with water-soluble lead-free solder pastes, although it can occur with halide-based solder pastes when soldering to HASL finishes. This can also happen if the conductor surface is heavily oxidized or if the solder paste has failed (i.e. the flux has failed). Using a highly activated solder paste will allow your solder to form a strong bond with the pads during assembly. You should also make sure to remove as much as possible any oxide from the metals you are connecting to. This will prevent surface tension from pulling the solder over the pads and into the balls during curing.
Another part of the process that helps prevent wetting is flowing nitrogen through the reflow oven during soldering. This helps prevent oxide formation in high temperature ovens. You should also check that the plating thickness is sufficient (at least 5 microns). Both measures help prevent oxide formation and diffusion into the plating during soldering.
Poor wetting of lead-free solder
Lead-free tin-silver-copper solder is important to maintain RoHS compliance, but has poor wetting properties when soldered on bare copper. This is one of the many reasons to use surface treatments on exposed conductors. The surface finish of tin, silver and ENIG provides better wetting.
The peak temperature during soldering should also be within the correct range. The optimum operating temperature for lead-free tin-silver-copper solder is approximately 240°C, and soldering outside this range can cause wetting problems. The image below shows an example with a BGA ball. Due to the higher surface tension of these lead-free solders on BGAs, they may require larger spacing and the spacing/temperature profile should be checked with test coupons prior to full production.
Component moves and tombstones
Tombstones are also a wettability related issue. In an ideal soldering process, molten solder wets all pads of the SMT component at the same time. If the pads on one side wet (ie reach a sufficiently high temperature) before the pads on the other side of the part, the solder will pull on the part as it solidifies. A force mismatch on the two sides of the assembly can cause one side of the assembly to lift off the pad slightly or deviate from its ideal position on the pad. In some cases, the component will remain attached to the pad by the solder, although the contact resistance may be high and the bond weak.
In extreme temperature mismatch situations, such as one end of a component not getting wet at all, this can cause the component to stand on one end, which is called a tombstone. This is a common problem in SMT resistors and capacitors. Tombstoning during reflow soldering corresponds to a number of possible causes. The most common cause is uneven temperature in the reflow oven, which can cause solder to wet earlier in different areas of the PCB than other areas. Uneven application of solder paste during assembly can also cause wetting variations across the board.
In terms of design, component groups with unevenly arranged pads are susceptible to wetting and tombstones during reflow soldering. The size of the pad also affects its temperature differential during reflow; larger pads require more heat to reach the defined temperature, therefore, equal-sized pads should be used on each side of the component. When defining the pad size, it is easy to make the pad too large and the excess copper will dissipate during the soldering process.
When building the layout, be sure to check the pad size and clearance to ensure solderability during assembly. The use of heat sinks on the cooler side of the assembly is known to help prevent tombstones during reflow soldering. Note that pads and vias with heat sinks have complex impedance structures similar to real capacitors, creating some signal integrity issues in high speed/high frequency designs.
