Guenter,
       With entrapment of the water solubles, you will also find that if 
there is any moisture still present, when power is applied, the corrosive 
effects are greatly accelerated.  Also, depending on the level of other 
contaminates present, dendrite growth can occur rapidly resulting in shorted 
conductors.  This effect will occur at any time in the future if moisture 
becomes present under that package while power is applied.
        In place of a HASL coating, gold over nickel plating may provide 
more consistent results if the OSP coating does not prove to be acceptable 
throughout the soldering process.

       Norm Dill
       [log in to unmask]
 ----------
From: TechNet-request
To: Technet
Subject: Re:Preferred PWB Surface for BGA
Date: Tuesday, May 07, 1996 8:37AM

Dear Gita

 -Due to the high coplanarity of the solderballs of BGAs and the variing
thickness of HAL surfaces theoretically OSP should be better. However, I
made several tests with BGAs on HAL pads and didn't face any problems so
far in this respect.

 - I recomment NoClean flux. Tests, where we measured the surface resistance
underneath  QFP packages, suggest that flux residues are washed underneath
the component and trapped there. I belive that this effect is even worse
with QFPs. This trapping is disastreous with water soluble fluxes since
their residues are highly corrosive.

Best regards

Guenter Grossmann
Federal Institute of Technology
Reliability Lab.
Zuerich, Switzerland