Check out IPC-TM-650 for a test method to test the cure of solder mask.
Also if you are manufacturing other boards using the same glue and process
that are not having this problem then that even futher points to a mask
issue.  I have seen this before and in that case even the mask vendor could
not explain what was wrong but a new batch of boards with the same mask
cured the problem.  Either they were not telling or "go figure"

Ken Kirby
Process Specialist
Speedline Tenologies





 (Embedded
 image moved   Doug Pauls <[log in to unmask]>
 to file:      07/06/99 11:32 AM
 pic17937.pcx)




To:   [log in to unmask]
cc:    (bcc: Kenneth Kirby/ElectrovertUS/Cookson)
Subject:  Re: [TN] adhesive to solder mask adherence issue




In a message dated 07/03/1999 11:45:15 AM US Eastern Standard Time,
[log in to unmask] writes:

> We cleaned the boards with IPA which should have removed any flux residue
>  from the HASL process.  The cleaning had no effect on the adhesive
bonding
>  to the PCB.

I wouldn't bet the farm on the first statement.  IPA cleaning is fairly
benign and won't get absorbed or tenacious residues.  The fact that you see
no difference in the problem can be taken as support for this position.

>
>  Interestingly enough we are seeing good adhesion and very poor adhesion
on
>  the same PCB.  The PCB is about 8"x8" and has 120 part glued down.  The
>  parts are SOIC-8, SOT23, and 1206 resistors and caps.  The problem is
seen
>  with every part type.
>
>  Our thought right now is that the solder mask (SR1010G) is not fully
cured.
>   However the PCB fabricator insists this is not the problem because
>  non-cured mask would be evident by mask defects during HASL process.

Maybe, maybe not.  Are the problem areas characterized by large expanses of
unbroken solder mask or lots of developed edges.  Boards can have solder
mask
that are not completely cured and not have the defects that the fabricator
is
indicating.  Boards in such an uncured condition can still grab onto HASL
residues and not let them go without significant energy (chemical or
mechanical) input.  I would be curious as to what defects the fabricator
would expect to see from incompletely cured mask.  You might also have
residual mask development chemicals in or on the mask as well, although my
guess would be residual HASL flux.

>
>  What is polyglycol and where does it come from?

This is a general chemical class.  Ethylene glycol, one of the primary
parts
of automotive anti-freeze is a common one.  As Jack Brous (Alpha Metals -
retired) showed in the past, there are detrimental polyglycols and some
which
are not detrimental.  As a general class, these chemicals are added to
fluxes, cleaning agents, etc., as wetting agents which break down surface
tension, allowing greater flow and greater water breaks.

Doug Pauls
Technical Director
CSL