William O. Alger, Pedro J. Martinez, Weston C. Roth
Hillsboro, Oregon, USA
Cavity technology in a Printed Circuit Board (PCB) has existed for many years. The methodology to create the cavity in the PCB has evolved over time as technologies have advanced and the manufacturing process varies by the individual PCB fabricator as well as the reasons for using the cavity technology.
For the purpose of this paper, a cavity will be defined as a hole in the PCB going from the outer copper layer to an inner copper layer, but not completely through the PCB. The cavity design and assembly issues identified during the design of experiments (DOEs), the findings, reliability results, and conclusions will be discussed in this paper.
The use of cavities in a PCB as a method to reduce the component height or to increase component clearances is a viable technology. SMT testing of the cavity boards assembled with the 0.4mm pitch BGA packages at multiple cavity depths has demonstrated that PCBs can be successfully paste printed, assembled, and reflowed in the same multi-level processes with the existing equipment capability.
Single pass, multi-level solder paste printing is the most challenging aspect to implementing cavity technology into a high volume manufacturing environment. The key to a successful paste print is optimizing the critical print parameters; 1.) Solder pad to cavity wall air gap, 2.) Squeegee flap length and thickness, 3.) Print fixture support, and 4.) Pocket stencil alignment.
Optimizing the print parameters helps insure multi-level paste print quality but the parameters may need to change as the cavity design changes. The testing demonstrated a relationship between the cavity depth, air gap, and flap length and flap thickness that impacts the paste printing results. Test prints would be necessary to reset the critical parameters to accommodate the design differences. Other print parameters not mentioned in the paper that are also critical to achieving consistent solder paste volume with single level paste printing are important and also apply to multi-level printing.
The drop testing showed that the solder joints on the cavity components appear to be as reliable as the same components assembled to the surface of the PCB. The drop test results also showed a statistical reliability improvement for the 2-layer cavity over the 4-layer cavity.
An alternative methodology to paste printing the BGA style packages, is dipping of cavity components in paste or flux prior to placement. The dipping method does not require solder paste stencils to be pocketed. Flux dipping was not investigated during this study.
The conclusions are based upon a limited data set and additional testing is encouraged.
Initially Published in the IPC Proceedings