TECHNET Archives

1995

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:
[log in to unmask] (Jerry Cupples)
Date:
Mon, 9 Oct 1995 21:25:27 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (128 lines)
(weird header junk deleted)

Mike Cussen commented:

>I ran into a delamination problem, with an individual supplier on two
>different multilayer boards (FR4) after the IR process in our SMT
>assembly.

He then asked:

> My question is, is it nessesary to bake these blank boards prior to the SMT
>process and if so, what are some guide lines I may use.
> 1. Tempurature 230 degress F. for 2 hours
> 2. Stacking Requirements in the oven?
>    a. Separation between layers.

We bake 'em 4 hrs min (usually overnight) in a convection oven, unwrapped
and stacked 10-20 up (not separated) at 110 C. make sure you cool in a dry
spot for several hours prior to beginning process. IMO this is 99%
_unecessary_ but we do it because the guy who had the job before me wrote a
spec. At least it does not cost much, and does little harm. Impresses the
ISO auditors, otherwise useless. What really matters is that you keep them
in a cool, dry place away sulfur bearing kraft paper or corrugated. Always
unwrap the boxes, detrash, and keep the boards in their poly bags as the
fab vendor provides. Most places heat shrink poly in stacks 1-2 inches high
anymore.

and then, an extensive quotation from the gospel according to IPC:

>                         Delamination & Blistering
>                                IPC-A-600D
>
>Although it may occasionally occur because of a sever process
>breakdown, delamination and blistering occurs as a result of an
>inherent weakness of the material. Either condition represents a
>breakdown in the cohesive strength of the composite laminate and
>the internal openings that develop can mechanically weaken in the
>final product and also provide undesireable conductive paths
>between circuit elements.
>
>Delamination or blistering may occur in clad laminate materials
>either between plies of laminate, or between the laminate and the
>foil cladding.
>Any occurrence, either in the as-received condition or after
>routine thermal stress testing at receiving inspection is
>evidence of a major flaw, and the material should be rejected,
>since the defect will almost always propagate during the normal
>stresses of manufacturing.

Amen, brothers and sisters! You listen to that part, Mike, and send those
boards back where they come from. Tell the PWB house they bought the
components on those boards, too. This is a serious problem, and they won't
quibble, I bet unless there's something else you're not specifying.

If you have any more bare boards from the same date code, hold some back,
confirm whether they have any at the fab house, and tell them you strongly
suspect lamination or materials problems there.

You could put some of those holdback boards through a bake out and then
into a fluidized bed thermal shock test to confirm a suspected tendency to
delaminate. Split your sample and do the smae test on boards as-is.

If they ask about how you baked 'em, get a righteous attitude and tell them
it does not matter.

Further verses were quoted:

>
>                          Causes of Delamination
>              "Quality Assessment of Printed Circuit Boards"
>
>The reason for delamination is epoxy starvation in the glass
>cloth layers or an incomplete curing of the base material. It has
>been reported that some manufacturers of laminates have reduced
>the lamination press cycle a little in order to increase
>production.
>In multilayer boards, visible delamination may be due to poor
>lamination of layers or insufficient bonding of prepregs to the
>inner layer copper circuits. Here, the PCB manufacturer is in a
>way also the "manufacturer" of the base material.

I believe every word of that, too.

>Delamination may develop, however, in a board after assembly and
              ^^^
>soldering. There are several possible reasons for this:
>1. Asymmetrical buildup of multilayer boards.
>2. Faulty soldering techniques, such as skipping the baking which
>   removes retained moisture from the board.
>   (A 1 to 2 hour baking at 230 degress Fahrenheit immediately
>   before assembly and soldering is highly recommended.
>3. Lack of preheating as the first step of mass soldering.

Naaaah. And it _might_ be due to some rare bacterium munching on the
adhesion promoters, so you should soak in disinfectant? The above was
inserted in later versions from the original by Greek translators who had
been schooled in the doctrine of PWB fab house sympathy, and had been paid
to offer this escape avenue for sinful PWB fab operators. It should have
been expurgated.

The first two sections apply just as well to boards which have been
soldered. If you are doing anything near normal at reflow, no prebaking or
weak preheat would not matter. I also contend that a 1-2 hour bake at that
temp would not dessicate any moisture between internal layers, fer shure.
You _can't_ do it "immediately before assembly", anyhow. And within about
24 hours, the glass epoxy will have gone back to equilibrium with the
storage atmosphere. You didn't store these boards outside in a wet
cardboard box or anything, did you?

Seriously, I have seen this only a very few times, and cannot accept any
explanation other than that it was resin starved or old material, not kept
in the laminating press long enough, or maybe some bizarre too-thick black
oxide. You should be able to put test coupons through a ferocious thermal
shock in a fluidized bed for repeated cycles and see _no_ delamination.

Mike then went on to quote the whole thing for clarity once more, which I
will not repeat.


cheers,

Jerry Cupples
Interphase Corporation
Dallas, TX




ATOM RSS1 RSS2