We're experiencing increased flux residues after wave solder cleaning.  The
flux residue is causing false failures at bed of nails testing, is visible
in larger quantities, and measures higher on bulk ionic contamination tests.
Our experiments found two primary problem contributors.  Both 1)one specific
wave soldering flux and 2)some lots of circuit boards must be present.
The wave flux is a low solids, no-clean(yes, we are water cleaning).
The boards are FR-4 with liquid photoimageable solder mask, mixed surface
mount & thru hole, and with routing & holes.
Both the flux and the board fabricator have been used successfully for years.
The ionic contamination levels range up to 25 times higher(0.1 ug vs 3 ug
equivalent NaCl) on recent controlled experiments and 100 times higher than
on past process performance.
A band-aid fix has been to add saponifier to our hot water wash.
What's the mechanism of why some board lots have much higher flux residues
after wave cleaning compared to other board lots in a long established process?
Any advice would be appreciated.