TECHNET Archives

August 1997

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Paul Gould <[log in to unmask]>
Reply To:
TechNet Mail Forum.
Date:
Thu, 28 Aug 1997 23:14:30 +0100
Content-Type:
text/plain
Parts/Attachments:
text/plain (58 lines)
Richard MacCutcheon wrote:-
>We seem to have developed a gremlin in our registration dept.
>A very large portion of scrap is attributed to NO Annular Ring (<.002").
>
>Over 90% of our product is 8-layer GIL with 2 and 7 as ground planes.
>Cores are .005 constructed of 1-106 and 1-2313.  Prepreg is 2313.
>
>Layer to layer misregistration is around .005" on average.  Artwork pad
>sizes are .022" over the max acceptable hole size.  Tooling seems to be as
>tight as ever but in the last few weeks we have had an inordinate amount of
>scrap due to NO Annular Ring.
>
>What is the best systematic approach to take in identifying the problem(s)?

Hello Richard,

The cause of the misregistration could be:-

1 Layer to Layer misalignment on the same core
2 Core to Core misalignment
3 Drill target accuracy for CNC drill
4 Image stretch or shrink
5 Random drill variation

Sectioning can determine (1) or (2). (1) could be due to photographic
tooling for imaging inner layers. Carry out checks on inner layers
before pressing.
(2) could be caused by inaccurate tooling for pressing. This can also be
checked by pinning panels together before pressing and looking for
misalignment between the cores. It should be easy to see with thin
cores.

(3) is critical and if you are seeing good registration between all
layers on the micro-section, but the drilling does not match the inner
layers, this could be due to inaccurate target scoping. If the image is
rotated relative to the drilling, this will almost certainly be your
target scoping or pinning. Pins must be absolutely vertical as you will
get around 3 mil error for every degree off the vertical on a 3 high
stack of 0.062" panels. The top panel is the worst.

(4) is due to photography or laminate dimensional changes. You can
overlay the inner layer image on the drill pattern to determine whether
the right stretch factor has been applied. If the material is moving
more than predicted, then new photoplots will be required.

(5) is least likely as you would have identified that quite easily but
is due to spindle run out, lead screw wear, or stack movement.

Once you are certain of the cause, the corrective action should be easy
to work out although maybe not so easy to implement.

Good luck
Paul Gould
Teknacron Circuits Ltd
[log in to unmask]
Isle of Wight,UK


ATOM RSS1 RSS2