SYN@PTICS®
Data enfolded in database, any time that the system detects misregistration during the test-coupons drilling, are available to be explored selecting:
1) PANEL SIZE
2) LAMINATE
THIKNESS
3) TYPE (GROUND/SIGNAL/MIXED)
4) SUPPLIER
5) RANGE OF
HISTORY TIME
DATA EXPLORERS 
Statistical
data can help the choice of materials and suppliers, enabling more homogeneous
lay-ups building.
Really, I
found that it's not sufficient to have statistical data in production environments
where a lot of new lay-ups are produced everytime.
In any case
I confirm the usefulness of them in order to perform a first rough correction
on the films, specially when working on a known lay-up.
The goal of
this project is to give a solution that behaves in "REAL-TIME" and in "LONG-TIMES",
with a very little price compared with x-ray machines.
This system is intended more useful in environments where there is consistency of data on innerlayers misregistration inside a lot. Some people working in pcb fabrication tell me that this condition is not present in their process. Really I think that a random error of registration inside a lot can exist, but it must be a fraction of the error caused by material shrinkage. Infact a random error can be generated by flaw in some mechanical registration systems or in wrong procedures, and nobody can accept 0.0005In/In for this cause. I think that a mechanical registration error (random) can be reduced to be:
1)Inside a lot all the laminates of the same type must be supplied by the same firm and must have the same characteristics(resin content is the most important).
2)The artworks must be stabilized before use(indipendently from the fact that one can make here a first compensation of laminates shrinkage).
3)Any mechanical
registration system involved in the process must be adequate to the required
precision.