I'd offer one other suggestion.

We've always been told that the OSP coating thickness cannot exceed 0.5
um, but we've seen instances where it greatly exceeds this value - up to 3x
in some situations.  The OSP suppliers have argued that this was
impossible, but the data did come from a very reputable lab and I don't doubt the
results.  The difference (in my opinion) comes from the analytical method.
 The OSP suppliers measure thickness on copper coupons by chemically
reducing the coating and calculating an average for the entire coupon.  (At
least, this was how they did this 3-4 years ago when we had this
discussion!)  The significantly thicker coatings were found in localized regions,
such as on BGA SMT pads.  The pad geometries, coupled with surface tension of
the coatings, may have contributed to the thicker readings.

Greg Bartlett
Mercury Computer Systems
Chelmsford, MA
[log in to unmask]

>
>     Hello Technetters,
>     A colleague and I have been facing some problems  in Circuit
testing
>     with 2 types of  boards that have OSP coating. There has been a low
>     First time pass yield at approx 40% due to the thickness of the OSP
>     coating, which we are specifying to the PCB manufacturers to be in
the
>     range of 0.2-0.25 um.
>     The  OSP coating is Entek Plus type 106 and the type of pin
material
>     used on these OSP boards are High conductivity steel made by QA
>     technologies.
>     The Incircuit tester used is HP 3070 series II with vaccum
fixturing
>     and the total pin count is 2004 pins on one board type and 1443 on
>     another. Test time is approximately 1-2 minutes per board, and the
>     failure if faused by poor pin contact will occur within the first
30
>     seconds of testing.
>     We have been trying some alternative pin types ie. rotating flux
>     buster type pins. Has any one out there used this pin type or any
>     other that has increased the success rate on the FTPY?
>     Does anyone have any suggestions on
>     - alternative pin types and manufacturers
>     - changes to the test fixturing to help resolve this problem
>     - information on how to ensure while manufacturing these boards,
osp
>     coating proceses are controlled to get the right coating thickness.
>
>     Thanks and regards
>     Arjunan Murthy
>     [log in to unmask]
>