For best results, know where to put solder mask, final finish and components.
Fine-pitch devices and automation levels are resulting in greater board density. There are mechanical and logistical limits to how far these trends can go, however, and the following best practices will help ensure that an SMT device is manufacturable.
For fine-pitch devices, designers must be careful to ensure that no natural bridges occur. A natural bridge is a clear, straight track between leads. This creates a trace over which solder can (and likely will) run when soldered (Figure 1). Although the result is not technically a defect, it appears as one to automated and human testers. This causes delays and additional manual (i.e., expensive) intervention to check. To avoid natural bridges, traces should be designed to route from the ends of pads, rather than using direct, straight connections.
Vias are holes drilled through layers of a PCB and plated to connect the layers of circuitry. Two design practices should be observed regarding vias and ball grid arrays:
- Always mask vias under BGAs. Solder from the paste or balls on the BGA may come into contact with the via, creating a defect. Rework is problematic, as more heat is required to melt the solder. To avoid this situation, mask vias under BGAs with solder resist.
- Avoid placing vias in pads. If a via is placed where a device is to be mounted, solder can run down the via, making rework difficult and reducing control over the solder joint volume.
- Always ensure that copper pads define the solderable area, rather than using mask-defined pads. The mask application process is not as well controlled as the copper definition process, and defects are more likely to result.
Balance the thermal mass on leads of smaller devices, especially 0402s. Thermal imbalances can cause a defect known as tombstoning. In such cases, as one end of the device reaches temperature faster than the other end, surface tension on the molten lead causes the device to stand up on end.
Also, to achieve a flat PCB finish, organic surface protectants (OSP), electroless gold plating, or immersion tin finishes should be used on fine-pitch and BGA devices. This helps prevent defects such as open circuits on devices that contain many coplanar leads.
Solder resist dams of 0.003" are required, particularly on fine-pitch SMT devices. Omitting these dams increases the risk of short circuits through bridging between leads. The dams between these leads must be a minimum of 0.003" to be manufacturable by modern PCB fabrication processes.
The most economical method of soldering plated through-holes on SMT boards is with a solder wave machine (rather than manually). To help ensure that a board can be successfully passed through the wave, the following best practices should be observed.
Problems in manufacturing arise when a mixed board is heavy. Proper support and clear edges for stiffeners help avoid such issues. “Submarining” is a term used to describe the difficulties encountered when a double-sided board with inadequate support is put through wave solder. The board may sag, permitting solder to run across the top of the board and ruin it.
If a double-sided mixed-technology board is so dense that the above guidelines cannot be observed, a special fixture called a selective wave fixture into which the board is snapped can be used. This masks the bottom-side surface mount from the wave. To do this, there must be sufficient clearance (0.1") between the PTH pins and solder-side SMT so that the SMT can be masked by the fixture.
Keep in mind that fixtures add cost to the product, and it is best to avoid designs that require these. Several fixtures are required (at several hundreds of dollars each), and they have to be cleaned, loaded, unloaded and occasionally replaced.
Careful SMT pad design should be employed on the solder-side of a board. This includes thieving pads and toe pads, both necessary for successful wave soldering. In addition, components on the board should be oriented so that the solder joints are not shadowed as the board moves through the wave solder.
In all designs, snap-off materials should be designed on the edges of the board (Figure 2). These help support the board through the wave machine and can be removed after soldering. They also permit conveyors and clamps to function properly.
In some cases, titanium stiffeners can be used (provided there is sufficient clearance on the board) to help make the board more rigid. Both techniques require additional labor and therefore, add cost to the final product.