Operation Chip Shape

Print

Size Matters

Figure 1: Line of single pads. Each pad should be two pads and able to fit component shown in Figure 2.

Please forgive me, it's football season and researching the infinitely detailed IPC and Joint Industry Standards can make one cross-eyed and promote day-dreams.

The quarterback looks over the landscape, sees something he doesn't like and calls out an audible, "J Standard". He alerts the members of the team.  "001".  He narrows the formation, "Rev D".  The tight-end goes in motion.  "7.6.4".  The whole team is ready to move. "Hut!"

What would cause such ruminations?  A customer was wondering what industry standard applied to a pad / part size mismatch.  Well, how's this for a mouthful, "IPC-J-STD-001D, Requirements for Soldered Electrical and Electronic Assemblies, Table 7-4, Dimensional Criteria- Chip Components - Rectangular or Square End Components - 1, 3 or 5".  Specifically, minimum end solder joint width on a rectangular chip cap.   Of course that's not to mention "IPC-A-610D, 8.2.2.9 ..."  Which basically deals with the same subject.

Figure 2: Component footprint on top of mis-sized pads.

It's a wonder pad to part size mismatch doesn't happen more often considering the thousands and thousands of opportunities for failure on many assemblies.  No matter, when it does happen it can cause a board to be un-usable, and try as hard as we might to read a solution into the IPC specs, it's not easy to "audible out" of that situation.  In Figures 1 and 2 you can see the results of mistaken communication. 

Through the artwork design phase of this circuit board there was an error in the pad layout. What should have been two individual pads to solder a 1206 chip component in-place, ended up being a single elongated pad. Now, if the pad was the proper "width", the fix would have been relatively simple; just create a small separation in the middle of the two ends.  Unfortunately the single pad was too thin.  The applied class three criteria required a 75% end width connection, and this component was just a little too wide to satisfy that standard.  "Bill boarding" the component, that is setting it on its side, was briefly considered inasmuch as that would have allowed the component to meet the end width requirements, but the vulnerability of the components when placed at multiple locations was a matter of concern.  Well, after multiple measurements, component placements, comparisons and conversations it was decided that these pads would not do as they were.  Over thirty locations were not acceptable.

Figure 3 : Line of un-repaired pads.

What could be done? Replace the pads with the proper size pads that's what!  Not every assembly justifies that amount of effort,  but when an assembly has that kind of value, and time is of the essence, it is worthwhile to repair every location to get this board functioning and into service.  See 4.7.2 Surface Mount Pad Repair, Film Adhesive Method outlines the basic process for conducting this repair.  This is very useful to start this repair, though a basic procedure like this can't cover all of the variables that challenge technicians on every board. In this case the connections to each pad presented angles that were not standard to every pad,  so many of the pads had to be custom cut in order to fit every location.  It was quite a piece of work to cut out over 60 pads and bond them securely to ensure proper component placement once the pads were adhered.  Figure 3 shows a row of un-repaired pads. 

Figure 4: Line of same pads - repaired.

Figure 4 shows the same row with a new set of pads and Figure 5 shows a close up of a component placed on new pads.

The customer was able to get the board up and functioning and though we didn't get to hear 50,000 fans yell in glorious delight as if we scored a touchdown we felt pretty good to know we did our job.

Now, "Go Patriots!"

Several members of the Circuit Technology Center team contributed to this feature story.