Unusual solderability issue

Sounds like you and your board suppliers have done alot of work with this issue trying to solve the problem. When I got into this situation previously, I relied on an outside lab to help me. We didn't have SEM or any testing capability at my shop either. Contact Larry Bross or Ron Sweezo at Specialty Labs (http://www.thespecialtylab.com/html/contact.html). Larry would give me alot of information over the phone and help me to understand potential causes. Their test results were impeccable and affordable. It would help me to save time if an issue was dragging out too long and the problems were still occuring. Good Luck!

reply »

Hi, if you want to verify the gold contamination, is mandatory to use the XPS and FTIR analysis, but is difficult think about a gold contamination from 3 different manufacturers. Do you have cleaned the PCBs after an erroneous screen process ? If not, I suggest you to verify better the reflow profile and the chemical of solder paste. Many lead-free solder pastes lost most of chemical activity before reaching the peak zone. So the solder pastes show too much activity in the first part of the soak zones, resulting in poor performance in the higher temperature areas of the solder process. In any case, the answer can be found only after a correct tests program. There are a lot of thing to verify, but in order to understand better all the phenomenons that occur during the production, is important to study a good handbook before, to avoid to loose a lot of money in erroneous failure analysis... Good Luck, www.analysispcb.com

reply »

We have had some unusual reflow conditions with BGA's as well. We did not come to the conclusion that something was covering the lands though. It typically occured with only the BGA lands (not the decoupling caps on the back side). We found that the mass from the lands,PCB substrate, and part, took more heat through the reflow process than expected. We literally drilled a hole through a board underneath the BGA and afixed our thermocoupler from our heat profiler to the PCB between the BGA and the PCB. The actual temperature between the component and the PCB was much to low to get a good relfow/wetting. Since we increased our oven profile temp we have not had any of these issues.

reply »

Thanks to all for your input.

We did have a third party do an analysis on boards from the first shop. They were shipped directly from the board shop to eliminate the possibility that we could contaminate them. The third party (Forsite) did find Silica along with traces of chloride, sodium and sulfate. They stated that the level of silica was high enough to cause solderability issues and recommended a cleaning process. The cleaned boards were perfect. Unfortunately by the time we received these results we had moved all our high end boards away from this shop so we still can't confirm that the shop was able to produce good boards ongoing.

We've gone through our profiles many times. What makes us relatively sure it's not profile related (I'm not naive enough to say it's imposible!) is that the pads won't solder in our hot air test. We simply place solder paste on the pad and use hot air to bring it to reflow right in front of our eyes. On any other pads, even fiducials the solder wets across the pad as soon as the solder reaches liquidous but on the suspect pads it balls up and will actually blow off the pad if you're not careful. Also note that the defect shows up on the bottom of the board as well but only on parts directly under the BGA (decoupling caps etc). This rules out the mass of the BGA as a cause, at least for the bottom side. I've attached pics of bottom and top side defects. The bottom side clearly shows the solder wetted to the parts but not to the PCB. The BGA pad is totally devoid of solder and I was unable to solder to the pads unless I mechanically scraped the pad with the soldering iron.

I'm expecting an update from the third shop in the next two days. I'll keep everyone updated. Thanks again for your input.

Attachments:

reply »