Troubleshooting Flex Applications

I imagine that everyone has been in this position at one time or another, despite everyone’s best attempt at creating the perfect design, PCB fabrication and assembly, something goes wrong and the trouble shooting begins.  I had the opportunity to sit down with Ed Knutson, the President/Founder of Dimation, to swap some of our best war stories.  Ed specializes in quick turn assembly and design and I bring the fabrication piece to the discussion.  Our banter back and forth was primarily focused on flexible circuit applications for aircraft and Mil/Aero projects.  I am not sure if that is because of the more stringent requirements for those applications or more likely because that is an industry segment that we both work with regularly.   At the end of our discussion, we concluded that most of the war stories could be traced back to a break-down in communication and often times simply not fully understanding how each piece of the design-fabrication- assembly puzzle fit together.  We want to share a few of our stories and lessons learned.

UL Materials

Aircraft applications typically require materials rated to UL94V-0.  The assembly is complete and the burn test fails.  What happened?  The perfect storm.  When the design files were created for the PCB fabrication and assembly, the UL requirements were noted in the assembly files only and called out by test requirements, not UL 94V-0.  This was an ITAR application, so the PCB fab files were separated from the assembly files and forwarded the flex manufacturer.  Because there were no UL requirements listed on the fabrication notes, the supplier defaulted to their standard materials and the flex was not built with flame retardant materials.  That explains why the final assembly failed the burn test.  Lesson Learned:  Always clearly communicate UL requirements and include the requirement in both the PCB fabrication notes and the assembly notes.

Coverlay

There were many stories along this line, but this one is classic, we have both seen this more than once.  A particular application, on a tiny flex circuit, requires a very tight pad pattern. Standard, adhesive based coverlay, was called out in the stack up.  As the flex manufacturer was setting up the tooling, they asked if that area could be “gang opened” because the tight features would cause fabrication issues when aligning the drilled coverlay.   That is a very common question that I have seen asked and approved hundreds of times.  The designer agreed that would be fine and that pad location was left free of coverlay.  But, once the parts arrived at the assembler and they went to screen print the paste, the area shorted out.  The problem was ultimately solved by using photimageble coverlay to accommodate the tight feature pattern.  Lesson Learned:  Review even the “standard” requests with a critical eye for the next processing steps the flex will see after fabrication.   The size of this particular flex combined with the tight features was the perfect combination to cause an issue with something that is routinely done.

Bend radius

By definition, flexible circuits are designed to bend, fold, and flex during installation and/or use.  That doesn’t mean that the copper will not crack or break when it is overly stressed.  There are two very important things to be aware of.  First, RA (rolled annealed) copper versus ED (electodeposited) copper.  There really is a significant difference in ductility.  With a tight bend radius, or for a dynamically flexing application, specify RA copper.  Second, involve your fabricator.  The flex manufacturer is only going to see the design in a two dimensional view.  They will not know exactly how this is going to be used in your final assembly.  If you are concerned about bend radius or otherwise stressing the copper, ask for their advice.  There are many different “tricks of the trade” that a flex fabricator can recommend to ease the stress on the copper and improve performance.  Use their knowledge!

Array configuration for assembly

It is common knowledge that assembling flex can create challenges.  A lot of trial and error is done to find the best way to handle it.  Flex circuit size, array configuration, component placement and stiffener requirements all play into the decision, which just may be equal parts art and science.  The first decision is whether the assembly will be done by hand or machine.  If the assembly is not done by hand, whether to use a stiffened array or machined pallet needs to be determined.  Here are a few examples:

For a small flex, with a few components on just one side and no stiffener requirements, consider creating a FR4 stiffener pallet with adhesive on the outside perimeter only.  After assembly, the flex can easily be peeled away from the stiffener pallet.  Caution:  a stiffener pallet with adhesive in selected areas only can easily be misunderstood during fabrication.  Make the objective very clear in the fabrication files.

For a long flex with stiffeners, we suggest cross hatching the copper, or adding in a copper pattern to maintain as much of the copper in the array as possible to add stability.  The flex can be pre-routed with tabs left to hold this into the array during assembly.  Once parts have been assembled, simply cut the tabs to release the flex from the array.  Caution:  stencil tolerance over this long length is an issue to be aware of.

A custom pallet is another common choice for assembly, especially when you are running more than a few panels.  Most often this is designed with FR4 material. The benefit to this is stability and flatness during assembly and also the ability to nest the flexible circuits in the tightest configuration possible to reduce the cost of the fabrication.  There is no need to add in extra copper area in the array for stability.

These are just a few of the lessons learned that we have accumulated over time.  I hope that these provide insight and suggestion that will help with your future flex designs, or at the very least, let you commiserate and know that you are not the only one challenged with these types of issues.  Feel free to get in touch and share your stories with us!

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s