Behind the Scenes

Meeting drawing requirements may be more complex than you think.

As a design is completed, the CAD data and drawing define all the important dimensions and requirements. It is the expectation that when the first shipment is received – and every one after that – the parts will meet all dimensional requirements. But has thought been given to how the manufacturer makes that happen?

It all starts with the CAD data the designer supplies. That data sets the standard. Now it is up to the manufacturer to create the tools to build the part. As it does, keep in mind that the manufacturer does not build a part, it builds a panel of parts. So unless the part is very large, the production panel will hold multiple parts. To accomplish this, the fabricators need to ensure all the features are aligned on all the layers across the entire panel.

As the number of circuit layers increases, so does the challenge of getting everything to align and finish at nominal. The first thing to consider is the materials used to make a circuit are not made of granite. They move in all dimensions, and different materials move differently. Manufacturers need to compensate for the movement. This is done with two techniques: alignment tooling and scaling.

Tooling pin systems help stack together all the pieces of material without movement during lamination, but they only do so much. If the materials are not the same size, tooling holes in the material will be prone to distortion and the result will be layer misalignment.

The key is to make sure all materials are as close to the same size as possible at the lamination step(s). All materials have residual stress from their creation. The pressure of lamination imparts a slight stretch in the copper-clad laminate as it is made. Once it is etched, some of that stress will be released, and the layer will shrink. Of course, the layers never seem to move the same; fabricators need to compensate for this movement.

This is done by scaling the artwork on each layer at the image and etch stage. Each manufacturer has rules of thumb for scaling. The factors include copper thickness, copper coverage, laminate thickness and laminate type and laminate manufacturer. While the rigid material moves, the glass weave does constrain it in the x and y directions to some degree. Non-reinforced materials such as flex laminates move more, however, and not necessarily linearly. This can create some real mismatches as layers are stacked in rigid-flex constructions.

So how do fabricators scale a part for the first time? Well, designers get the benefit of all the parts that went before. Scales are applied based on experience with the material set being used. Each layer in the stackup will receive a different scale factor based on material specifics and the processes the layer will see prior to being laminated to other layers.

When the first panels reach the drilling stage, the fabricator can analyze the panel. We can see how well the layers align to each other. We can see how square or skewed the panel is. We can also see how close we are to a nominal scale across the entire panel.

We can now determine our drilling strategy – creating a best fit adjustment of the drilling for maximum annular ring. The systems will adjust the drill pattern to hit all layers in the stack across the full panel. In fact, the analysis can even determine no drill solution will achieve minimum annular ring and set that panel aside. Now we can move forward with the panels with drilling and complete processing.

Just because we were able to drill, however, we may not be well optimized. We may see layers misaligned to some degree. We may also see that we are barely achieving annular ring. In the spirit of continuous improvement, we want to make adjustments for future builds. So we review the collected data to determine which layers are out of place and how close to nominal the panel is. With enough panels to create meaningful data, we can adjust the scales on just one layer, or up to all layers if needed.

Often, certain layers will be an outlier from the rest of the pack. For example, a pair of layers on a flex core is likely to move much differently than the rigid layers. Or a layer set with a low percentage of remaining copper may contract substantially, especially if it is a single-layer core.

The system reports on targets placed in the four corners of the panel. Essentially, it looks like a bullseye target, and we can see how close each layer is to the bullseye in all four corners. With this information, adjustments in scale can be made on the x and y axes. With updated scales, we can evaluate the impact on the next production lot. Depending on the results, we can adjust further, or lock in a workable solution.

These data are added to preexisting data for the particular material and board construction. In this way, initial scales become more accurate over time, resulting in higher first-pass yields.

Using a new material set or combination introduces a wild card. As a manufacturer, we won’t have history on how the new material performs. First-pass scales with a new material set will have a higher risk of failure to achieve annular ring. We may process a “scout” lot to confirm alignment before committing a full production volume to the shop floor. Over time, we’ll build up a new data set to support the new material.

Hopefully, this helps you understand that behind the scenes, a lot of science is behind producing a part that meets the drawing requirements.

Nick Koop is director of flex technology at TTM Technologies (ttm.com), vice chairman of the IPC Flexible Circuits Committee and co-chair of the IPC-6013 Qualification and Performance Specification for Flexible Printed Boards Subcommittee; nick.koop@ttmtech.com. He and co-“Flexpert” Mark Finstad (mark.finstad@flexiblecircuit.com) welcome your suggestions. They are speaking at PCB West at the Santa Clara (CA) Convention Center in October.