Steve,
Went through similar steps about seven years ago, so details are lost
but
some theory remains. Our findings were that the torque test was invalid
because we were testing cured adhesive at room temperature. The failure
mode is a preheated board hitting reflow or wave pot temperature. So,
first,
are you testing for the correct failure?
In the end, we finally found an adhesive that did not degrade as bad
when
heated (never found what we wanted) and started controlling our adhesive
application process closer using positive displacement pumps. On some
heavier components that required more adhesion, we even placed adhesive
at the outline edges of the part so when placed and pressed into the
adhesive
it would wrap around the outside edge and virtually clamp the part.
Agree with the matte finish mask, we even roughed part surfaces, but in
the
end, the adhesives just didn't like getting heated.
I would hope the adhesive mfgrs would have better products available
after
this much time though. Thats where I'd spend my time.
Good luck, Jeff
 ----------
From: Steve Collins
To: [log in to unmask]
Subject: [TN] Gen: Soldermask Adhesion