Authored By:
Kim Flanagan and Greg Wade
Indium Corporation
Clinton, NY
Summary
Voiding in bottom termination components (BTCs) like QFNs and LGAs have become quote the hot topic in the SMT industry. Surprisingly, one type of BTC component that is observed to have excessive amounts of voiding is the D-PAK. One would think that a component with leads located on only one side would mitigate flux entrapment and allow outgassing to escape more easily from beneath the component, as compared to other BTCs where the component is affixed to the PCB on two or more sides. However, the exact opposite has been observed in most cases.
This study looks at an analysis of why a D-PAK exhibits more voiding than other types of BTCs. Voiding results based on an analysis of several process variables, such as adding weights to the tops of D-PAKs and cutting off the leads of the D-PAKs, will give insight into the physics behind D-PAK voiding, and provide an answer to the root cause to provide more insight on how to remedy the phenomenon.
Conclusions
Cutting the leads off the component improved voiding for D-PAKs for both Part I and Part II of the experiment. This would support the initial theory that leads on only one side of the component contribute to higher voiding.
Another key point that should be noted is that voiding varies between the different types of D-PAKs. D-PAKs are designed in all different profiles. For example, reviewing the drawing specifications of the components, it was noticed that the two components that voided the least have a "step" on the underside of the component from the seating plane, while the component that voided the most does not have a "step". It is important to note that even though voiding is primarily the result of outgassing volatiles in fluxes during reflow becoming trapped in the solder joint, non-wetting and solder starvation are crucial factors that promote voiding, as well.
There could be other possible root causes for why component C voided the most and components A and B voided the least, for example, the component plating condition or oxidation levels on the board or component terminations may be different. In order to verify that, cross sections of the components would need to be taken and analyzed.
Another consideration that should be kept in mind is that the aperture design without window-panes showed much more component skewing during reflow and, of course, more voiding in Part II of the study than when the aperture design had window-panes in Part I of the study.
It was noticed that for components C with 5 leads, voiding was consistently greater than any of the other components. The initial theory was that it had something to do with these components having more leads than the rest; however, when the leads were all trimmed, the data still showed more voiding than the other components.
The weights actually increased overall voiding. This may be for two reasons. The first reason being that because of the size and mass of the weights, the reflow profile may have altered at the reflowed components. Also, adding the weights could have decreased the standoff height of the components which would promote more voiding.
Overall, cutting the leads for all of the components improved voiding in all cases of the study. Not only did it decrease overall voiding, but it also decreased the standard deviation of voiding.
Initially Published in the SMTA Proceedings
|
