The method you suggest is about the only way to �clean-up� this terrible situation. Realistically, this is a "memory" type condition induced by design, build symmetry, "cross ply construction", excessive cool down rates during lamination or reflow at assembly. [Read � it may not go away / it may go away, but later return / it may go away.]
Comments on your dewarping plan are: * Our board fab and we �dewarp� multilayer boards at 300�F. Of course, we do this on bare boards in a �pressing� action that wouldn�t work on a stuffed board. After our fab or we dewarp boards, the dewarper uses a electric engraving pencil to etch a 1/4� �W� within 1/2� of a global fiducial on side 1. * We�d be uneasy about dewarping an assembled board at Tg held in a clamped or pinned position, because of the potential of the center of the board to sag in the center worse than the original warping. Do not assume the Tg of your board is 135�C. There is no standard Tg for FR-4. * Appreciating that you built these boards for a customer, you built the boards out of specification, and your customer probably will FREAK when they find-out what you�ve done; consider talking your customer into waiving the defect on these boards. * Yano in real life, even without requesting a waiver, you probably should notify your customer that you�ve dewarped their boards, because dewarping is so radically different from your routine production process. This suggestion carries more importance as the number of layers of the bare board increases.
reply »