Mr. Franck, Mr. Paul, and others:

Mr. Paul's last comment on the utility of the existing bow and twist
measurement guidelines is an important issue.  Mr. Paul noted (in general)
that the test method should be used as a base line measurement for analyzing
the effectiveness of process/material improvements.

It is apparent that the existing guidelines and procedures surrounding bow
and twist have two major shortcomings.  First, the technique for measuring
bow and twist is not exactly optimal.  There is a large amount of operator
dependency, measurements are made with mechanical/contact devices, and
measurements are not made in a 'free state' (e.g. one must hold corners down,
etc.).

Second, information and know-how on bow and twist does not exist in a readily
transferable knowledge base.  It appears that most manufacturers are
dependent upon certain key individuals with years and years of experience to
identify causes of excessive warpage and to correct it.  The problem lies in
the fact that the decisions these key individuals make to solve problems are
based upon something they have seen or done before.  As effective as they may
be in solving the problem, once the person is gone it is tough to replace
them.  A quantifiable knowledge base is not left behind for others to work
with/from.

A new technology developed at the Georgia Institute of Technology (Georgia
Tech), Atlanta, Georgia, may help address the two issues outlined in the
above paragraphs.  A technique was developed which allows assemblers to
monitor, in real-time, thermally induced warpage during processing (e.g. IR
reflow, wave solder, convective reflow, UV cure, etc.).  The technology was
licensed to a spin-off, start-up company which has commercialized the
measurement service under the tradename TherMoire.

TherMoire incorporates a well known, and proven technique called shadow
moire.  Shadow moire is an optical technique which is capable of
characterizing surfaces in the z-axis dimension.  Basically, it provides a
topographic map of large area surfaces, much like the contour maps produced
by the US Geological Survey.

Shadow moire has been used by the industry with success to study the effects
of process/material variations on bow and twist in printed circuit
boards/assemblies.  The technique is superior to the existing IPC measurement
technique for bow and twist for many reasons.  Most notably, shadow moire is
optical, non-contact, operator independent, and it allows for boards to be
measured in a free state (e.g. just resting on a table top - or a calibrated
slab of granite).  

Electronic Packaging Services, the company that spun off from Geogia Tech,
has developed a modified version of TherMoire for production level usage.

During the 1996 IPC Printed Circuit Expo in San Jose, CA, a prototype of a
new system called LineMoire will be unveiled.  LineMoire will be a production
level, in-line, warpage measurement piece of equipment.  LineMoire will be
fully automated and capable of 100% inspection.  LineMoire will not only
accept/reject each sample based upon a pre-set parameter for bow and twist,
but it will produce statistical analyses on frequency, magnitude, and
LOCATION of bow and twist (sounds like a bow and twist SPC system to me).
 LineMoire can make local and global measurements on a surface.  Local
warpage information/data will only become more critical in the future.  A
sample application:  LineMoire can analyze a large QPF interconnect area or
say a BGA area for planarity prior to solder paste screening and/or automated
pick and place.

I encourage all to come by the booth at the IPC Printed Circuit Expo in
March.  As excited as we are about LineMoire, we need feedback from the
technical community on how shadow moire techniques could potentially be used
in a revision to IPC 2.4.22.  Understandably, all IPC standards and approved
techniques must be readily available to all.  Challenges exist on how to make
the data generated by shadow moire easily readable and techniques for
interpretation more standardized.

I look forward to meeting and working with many of you in the months ahead.

Best Regards,
Patrick Hassell
Electronic Packaging Services
(404) 881-1114
[log in to unmask]