Issue link: https://iconnect007.uberflip.com/i/1424540
NOVEMBER 2021 I SMT007 MAGAZINE 23 and/or ionic residues. is is a way to meet the standard without even doing any new analysis. is falls under the "if it ain't broke, don't fix it" category. At no time has anyone empirically said that 1.56 µg/cm 2 isn't a valid number for your prod- uct and process, it just needs to be proven that it works at weeding out dirty boards. e third note is to perform electrical testing on live product under elevated heat and humidi- ty conditions, using normal operating power. In my opinion this is the most important test when determining the impact of ionic cleanli- ness on field operation. e first note is what is throwing a lot of assemblers, as it requires sur- face insulation resistance (SIR) analysis that uses some sort of test board, most oen the IPC B-52 CRET board, using the actual prod- uct equipment and material set. at includes bare board solder mask, metallization, SMT paste flux, and any fluxes used for PTH solder- ing. e idea is to mimic the final product as closely as possible and see if the assembly pro- cess yields a product that is not adversely im- pacted when subjected to elevated heat and humidity. is is oen coupled with ion chro- matography analysis to determine the accept- able amount of residue aer assembly. You can also use ROSE in combination with the SIR analysis once you have a proven process. It is very important to use your in-house ROSE tester to determine what your acceptance criteria is if the plan is to use that for process monitoring. SIR is the requirement for new product ac- ceptance since obviously you won't have his- torical evidence at that point. ere is a fan- tastic document, IPC WP-019B, that goes into great detail on how we got to this point and even more importantly, how to go forward. Full disclosure, I am referenced as a subject matter expect (SME) in the document but can assure you I had only the smallest input. I'm sure the real contributing SMEs would agree with that as well. Honestly this article could have been one sentence, "Re: Cleanliness, see WP-019B, the end." Not sure why I am just now thinking of that, actually. Could have saved us all some time. It is very telling that the WP-019B uses 28 pages to explain three pages in the J-STD. Cleanliness is that important. Within the WP- 019 documents there are multiple examples of how to create objective evidence, and when you need to requalify your process, based on minor and major changes, with many examples based on the size of the CM. Here in our lab, we have been recommending something like this for many years. If you plan to use ROSE testing for process monitoring, at a minimum you need to do a rough correlation study. I say rough correlation because there isn't a 1:1 comparison between ROSE and IC testing. ROSE testing measures the amount of material that can become sol- uble during the test and gives you a resistivi- ty measurement; IC tells you exactly what the ions are and in what amount. What you can do is build a lot of 50 boards (or some statistical- ly relevant number) and then test 25 with IC and the remaining 25 with ROSE and see what the average is. If the ROSE average falls under the 1.56 µg/cm 2 great; if not, also great. If you have historical data that says you don't have is- sues with contamination and your ROSE tester gives you an average of 20 or 200 µg/cm 2 that ROSE testing measures the amount of material that can become soluble during the test and gives you a resistivity measure- ment; IC tells you exactly what the ions are and in what amount.