BGA rework and reliability

Dear all,

What are your views on reworking BGA's? I mean, with the BGA going through reflow temperatures four times or more, and with a manual reballing process, wouldn't it raise up a few reliability issues? Would you consider not doing rework on BGA at all?

Hon

reply »

| Dear all, | | What are your views on reworking BGA's? I mean, with the BGA going through reflow temperatures four times or more, and with a manual reballing process, wouldn't it raise up a few reliability issues? Would you consider not doing rework on BGA at all? | | Hon | Like everything else, you need to do a quick cost-benifit analysis to justify BGA rework. BGA rework cycle time is very long, in the range of 20 minutes or longer even if you don't include BGA reballing process. Depending on type of package and reliability requirement, you may not want to reuse BGAs. In that case, it is quite possible that rework cost may outrun assembly cost.

For sure, reliability as well as quality of reworkd assembly can be seriously compromized depending on the quality of your process and reflow time in your in-line reflow process. If BGA site get heated above 183 C (for eutectic solder paste) for too long, you are compromising reliability of solder joint due to excessive growth of Cu-SN intermetallic layer. So, add all reflow time together and make sure that is not too long!

Excessive temperature of BGA during rework can compromize the reliability of BGA itself. Make sure you do not heat the package above the manufacturer's specification.

reply »