The pink pencil eraser and other solutions.

The past few months, I’ve been reviewing problems that PWB fabrication issues can inflict on automated assembly and mass soldering operations. I’ve reviewed general assembly complications, those specific to Pb-free, and prevention strategies. Over the next two months, I’ll close the series with methods of dealing with the problems when prevention is no longer an option.

Don’t get me wrong: I’m all for doing things right the first time, actively controlling processes and resolving root causes as opposed to treating symptoms. In principle, this is obviously the optimal approach, but practically, we need assembly lines running. It would be great if we could return all substandard materials and wait for replacement stock. Unfortunately, if we did that, we might find ourselves with lines down, idle staffs, and dissatisfied customers rapidly seeking alternate suppliers. Although we might be completely justified in returning defective materials, sometimes the most sensible approach is to simply identify a workaround and implement it as quickly as possible.

In structured approaches to problem solving, we learn how to “identify short-term corrective actions.” (Gosh, how I love that phrase! It is such a nice euphemism, and it looks great in a monthly report to management.) At its essence, this really is about bandaging the problem to get product out the door. Here are some of my favorites:

Surface finish problems. On HASL boards, dross can build in the HASL pot and be included in the pad coating. Too much dross, or even regular oxidation from prolonged or poor storage, can dull fiducials so much that machine vision systems cannot read them. When the machine can’t read the fiducial, the line stops, awaiting operator intervention. Repeated line stops are costly, even if operator response time is relatively short. The fix is one of the simplest in the world: a pink pencil eraser. A few strokes with the eraser and voila – shiny, readable fids! I hate to admit it, but I once had a batch of boards so bad that I bought the operator who loaded the PWBs one of those automatic erasers that draftsmen kept by drawing boards (before design software rendered drawing boards, pencils and erasers obsolete). It helped keep his production rate up and prevented repetitive motion injury.

OSP coatings can be compromised by a number of factors, including poor application, oxidation prior to coating and prolonged storage. Fortunately, OSP issues can be relatively simple to fix if identified before the first SMT assembly pass. A fabricator or OSP supplier can strip the coating with solvent, perform a fresh microetch to clean the copper, and reapply the coating. I’ve had several lots of boards reworked for these reasons. But like any rework, it’s not free. If rework is required because of poor application or an unauthorized substitution of a specified product, the fabricator should bear the financial burden. If it is due to poor storage, handling, or specification, then the repair cost is the responsibility of the assembler.

ImAg finishes can experience application and shelf-life issues. They can be compromised by sulfur-rich environments, which means they should not be stored near kitchen or cafeteria areas, as many foods such as onions or garlic can emit sulfur fumes. I see the humor, but it is a lesson learned by one of my customers. The fabricator or supplier can rework ImAg finishes if necessary, but it is far more difficult and expensive than OSP rework and can hamper reliability.

Electroless nickel-immersion gold (ENIG) finishes cannot be reworked. If solderability or black pad issues rear their ugly heads, there is no alternative: Scrap the boards. Because ENIG is typically used on expensive, high performance product and black pad can cause field failures, scrapping is often the only option.

Dimensional inaccuracy. It is not uncommon for PWB pads to be offset from their CAD locations by a few mils or more. I have an extreme case study from a few years ago where the component centroids drifted by up to 0.008" over 13", and I just heard about another one last week. Although last week’s deviation was not as drastic, the assembler still could not get satisfactory paste deposits. In such situations, the board cannot be reworked or modified to fit the process, so the process must be modified to fit the board.

The remedy is straightforward: Measure and map the boards and scale the stencil image to match the PWB. Note the centroids of critical devices and modify the placement program locations to match the board. This is a prime example of foregoing root cause resolution for the sake of fulfilling business needs. It’s certainly not a preferred solution, but it does maintain assembly quality and aid timely delivery.

Overetched (undersized) pads can cause bridging issues because they prevent good gasketing between the stencil and PWB pads during printing. This situation also can be resolved by customizing a stencil, this time with smaller apertures to match the smaller pads. This solution should be approached with caution, however. Before it is implemented, the downsized apertures should be checked for aspect ratio and area ratio violations. If these key metrics dip below critical levels, the paste volumes can be reduced and repeatability affected, creating a whole new set of quality problems. It’s one thing to digress from standard processes for the sake of meeting production; it’s entirely another if the digression creates more issues than it fixes.

We’ll wrap up other situations next month.

Chrys Shea is an R&D applications engineering manager at Cookson Electronics (cooksonelectronics.com); chrysshea@cooksonelectronics.com. Her column appears monthly.