One thing I have experienced with companies on the commercial side of the industry is that with the advent of lead-free processes and higher reflow temperatures there are more "baked-on" flux residues (both water soluble and no-clean) and other hardened residues that the cleaning processes (if used) cannot clean, and because those residues are not only entrapped but baked on, the ROSE solvents do not pull them out very well for measurement. That is the basic fault of the ROSE system, it can be a pretty good "indicator" but you cannot judge cleanliness by that process alone. In the instances where things such as corona or power leakage took place, when I did the failure analysis and carefully removed component(s), I often found hardened flux high water marks around the circumference of the inner area under the component, and within that area both water-soluble AND no-clean sticky flux residues were present that had contributed to both power leaks which turned into carbonized deposits leading to catastrophic failure, and also on large processor components where the temporarily or partially entrapped sticky flux led to signal crosstalk and other similar issues. The fallacy that no-clean flux residues are harmless because "100% of no-clean flux is inerted by the reflow temperatures" and are thus harmless is a very dangerous assumption. What happens is that the very minor flux residues (from solder paste during reflow) around the perimeter of large flat powerpak components, D-paks, large 1206 caps, etc,  will harden, but near the center the flux is still an inactivated OR still active residue, especially with hot gas or hot air reworked components. Even with oven profiles that reach lead-free temperatures above 245 C there are places where partially entrapped flux may never reach temperatures above 200 C, and for much shorter periods of time such as only a few seconds. 1206 and larger caps especially are problematic simply because the terminations are right at the ends of the component body and the pads are often partially under the component body at each end. When paste is deposited the flux residue is already entrapped when the component is placed. Proper pad design where the inner edge of the large flat component pads do not go under the component body helps immensely but does not completely eliminate the problem. D-Pak power transistors/diodes acting as a rectifier circuit quite often have a belly pad where a deposit of paste leaves an entrapped flux reside deposit that is difficult to remove after two lead-free reflows for a double-sided CCA, and the downward spiral begins there. Non-activated flux is trapped not only between the PWB surface and the component surface but is also dammed in by the hardened flux around the perimeter, but not completely around it, and later if the dam breaks or is not fully sealed, a path is provided for electrical leakage and moisture exposure and this leads to all kinds of bad news. For CCAs like this, I like to see ROSE measurements at the 10 minute interval show 0 or less than .1 equivalent units. ROSE testing has to be accompanied by periodic ion chromatography testing to ensure the CCA is clean enough or at least the residues are not allowing a path for high power leakage or signal crosstalk, let alone active enough to lead to corrosion or other reactive issues. Even inactive sticky residues can present a path for signals or power that defeat the designed performance of the circuit.



-----Original Message-----

From: TechNet [mailto:[log in to unmask]] On Behalf Of Steve

Sent: Friday, June 23, 2017 6:50 AM

To: [log in to unmask]

Subject: [TN] PCBA Cleanliness



I realize that this is an old and continuing subject of debate in the industry, but I am interested in the current updated consensus on the topic. Specifically, what kind of PCBA cleanliness level should be reasonably expected to achieve? 



Of course the answer is always "it depends," because that is the reality. And the actual acceptance or control level needs to be set according to product requirements. But what I am after is slightly different: What level of cleanliness would you expect to be readily achieved by a good quality manufacturing process under normal circumstances?



Some general parameters to narrow down the question:



- Measurement technique bulk solvent (ROSE per IPC TM-650 2.2.25)

- Pretty normal PCBA design - 0403 components, some through hole, nothing really unusual

- Pretty normal process, RoHS, no-clean.

- Class 2 PCBA, but high impedance low current battery powered application

- High production volume (>1 million units annually)



I am not aware of anything that has replaced the 1.56/cm^2 "standard." but this has been around forever, it seems. Common sense tells me that even if a formal standard has not changed general performance might have improved, on account of smaller components, tighter board spacing and general technological improvements. 



What has been your actual experience in this regard?



Thanks in advance,

Steve