Hi Mark,
     At any given point in time, we have at least a half-dozen different
     part numbers going through the shop that employ blind/buried via
     technology, including sequential laminations, resin filled, CD-100
     filled, controlled depth drill and laser-drilled micro-vias.  I can
     tell you that there are no simple answers to your questions.  I
     wholeheartedly agree with the previous responses concerning previous
     track record, proper controls, yield data, etc.  However, the fact is
     that every design is different and presents their own unique set of
     problems.

     As you are designing sequential lamination blind vias, I'll just touch
     on a couple of things with this type of product that can be of
     concern.

     1) Thickness of the sub-assembly - Thin (below .020") multi-layered
     sub-assemblies become very tricky to control from a dimensional
     stability standpoint.  Materials move during the first lamination, and
     then again, to a less predictable degree, during the second.  It can
     be a nightmare geting both the blind vias and through vias to end up
     at any reasonable true position location.  Different tooling
     techniques can help improve this, but it is difficult, if not
     impossible to resolve.  Increased outer layer capture pads for the
     blinds can help minimize failures, but most likely you don't have room
     to grow these any larger than they already are.

     2) Excessively thick sub-assemblies can pose a problem from the
     standpoint of adequate filling of the vias through the lamination
     process.  In most cases, resin will flow up the vias, and, most of
     time, one is more concerned with resin removal from the surrounding
     surfaces.  However, if the resin doesn't make it all the way to the
     surface layer, the vias take on the characteristics of controlled
     depth vias, and it becomes vital that the shop is able to protect the
     copper that is exposed with whatever finish metal is being plated, at
     least through the outer etching process.

     3) Controlled impedance on the inner layer of sub-assemblies can be a
     problem, depending on the technique used to metalize the vias in the
     sub.  Some folks use a "dot pattern" only plating approach, then use a
     print and etch process to define the inner layer patterns.  This keeps
     the trace width and height controllable but "off-contact" can be a
     problem with this approach depending on how much of a bump is left
     around the hole.  Otherwise, a full panel plate, tent and etch or a
     inner pattern/outer panel plate approach can be used.  With these
     approaches, the inner layer is being plated and, depending on current
     density, trace height and width become more variable.

     4) Multiple cycling of any materials through the lamination process is
     not a great thing for the material.  Your design should not require
     that any individual layer or sub-assembly sees the more than 3
     laminations, and preferably no more than 2.  Lower temperature curing
     materials must be employed for further laminations.

     5) The cost of producing this type of blind via can be significant,
     when compared to controlled depth blinds.  Understand that to produce
     say an 8-layer board with blinds from 1 to 4 and 5 to 8, the
     fabricator is actually building 2 4-layer boards, then putting them
     together as another 4-layer board.  3X the processing of a straight
     8-layer.  If your design can maintain an aspect ratio of less than 1:1
     (laminate thickness:drill size) to the target layer, then controlled
     depth drilling may be a better (less costly, easier) approach.  There
     are certainly things to watch out for in this approach as well, that I
     won't mention here.

     As George Franck mentioned, plating over the resin in the blind can be
     troublesome.  There are ways to "roughen" the resin surface that makes
     this less of an issue, and, for the most, it can be controlled.
     George also mentioned the CB-100 via-fill process.  This is a great
     method for insuring an almost uniform surface over the blind vias.
     For "via-in-pad" designs, whether standard SMT or BGA, it can help
     with routing constraints.  There are limitations and you would be best
     to consult with a fabricator familiar with the process before
     designing it in.

     Hope this helps.

     Andy Slade
     Sr. Product Engineer
     Hadco Tech Center East


______________________________ Reply Separator _________________________________
Subject: [TN] Overplated blind via's
Author:  [log in to unmask] at SMTPLink-Hadco
Date:    06/16/98 7:22 PM


Hello Technet,

Our company is busy designing some printed circuit boards with
"overplated blind via's" (I mean blind via's, made in a sequential
plating process, where the via is filled with epoxy, after which it is
covered with copper and a solderable plating).
All our vendors claim they can build such boards, but as far as I
known they do not use this technology very often,  and I don't want to
be their "guinea pig".
The question :
What kind of questions should I ask our bare board vendors in order to
check wether they are capable of building these boards with good
relaibility in a reproducible way ? What are the potential problems
that the board vendors should be aware of, and should have solved ?
Any ideas ?

Mark Roach

################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text
in the body:
To subscribe:   SUBSCRIBE TechNet <your full name>
To unsubscribe:   SIGNOFF TechNet
################################################################
Please visit IPC web site (http://jefry.ipc.org/forum.htm) for additional
information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700
ext.312
################################################################

################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe:   SUBSCRIBE TechNet <your full name>
To unsubscribe:   SIGNOFF TechNet 
################################################################
Please visit IPC web site (http://jefry.ipc.org/forum.htm) for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################