TECHNET Archives

September 2002

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:
Earl Moon <[log in to unmask]>
Reply To:
TechNet E-Mail Forum.
Date:
Mon, 9 Sep 2002 15:48:24 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (95 lines)
Peter,

Here is something I sent to another concerned person off line:

Thank you first for the kind words. I have been doing BGA's since 1994. I my
earliest process development days, one of the first things I learned, and
very fast, was the coplanarity issues with BGA's was very important and
everyone realizes that now. Though I have done much development work over
the years, I have only written on article on the subject concerning BGA
resork as all other processes leading up to it only damanded my attention in
terms of many pages of procedures concerning production requirements so I
left it at that.

What I learned about rework, as part of any new process or product
development using BGA's, is that no acceptable solder joints may be formed
if no compliance between any or all balls on the BGA is not effected.
Simply, without some intimate contact between balls and paste or flux, there
will be no solder joint formation.

As most all BGA production assemblies requires using solder paste, solder
joint formation usually is effected even if a moderate degree of bow and
twist is present on either of the two substrates, meaning the device
substrate and the PCB substrate. That part is easy enought to understand and
the percentage of defects based on that fact, is very low as the solder
paste print height with a six mil thick stencil usually is about five mils
or so. Therfore, but for out of spec warpage on the device and board would
not allow all balls to make contact with the paste "columns."

Where this got really interesting was during rework, modification, or
repair. Early on I believed not solder paste was required during either
production, but for the obvious need to print solder paste for all other
parts forcing its use on BGA's, so that was acceptable. During rework,
modification, or repair - requiring BGA replacement, this fact became very
evident.

I started using only liquid flux in an experiment to determine if acceptable
solder wetting and solder joint formation was possible with this flux type.
When coplanarity issues were not an issue or factor (about the .75% rule by
physical measurement of boards and parts), solder wetting and joint
formation nearly was perfect in all cases. When out of the specified range,
some balls would wet and others would not. This was/is especially evident
using super BGA's known to have warpage problems because of their asymetric
constructions - as their metal tops as and example.

The next thing I found, because of production processes requiring solder
paste, was the coplanarity factor relaxed from that specified as .75% to the
more traditional 1.5% range for boards not requiring SMD's of any kind. In
other words, parts and boards really had to be our of spec to not solder
during rework. Then, using a rather viscous paste flux (much the consistency
of solder paste and capable of being micro-stencil printed), the same
wetting and solder joint formation was effected as acceptable. Actually, the
solder joints were near perfect as the BGA balls, when in contact with the
paste during reflow, attained their traditional "pumpkin" shape proving
solder paste was not required and the coplanarity specification could be
relaxed a bit. However, having said that, because of the vast variety of
different ball grid array devices, the tight coplanarity requirements must
be observed.

I also did much work with ceramic column grid arrays, and continue doing so,
and found the same things to be true. I must say that working with CCGA's is
a pleasant experience as it is nearly impossible to effect poor solder joint
quality using these devices. The caveat is that CCGA;s nearly always have
coplanar features that are perfect regarding their columns but this requires
very "flat" boards. This is one of the reasons I so adamently promote "good"
MLB constructions to ensure "flatness" is not an issue. I only wish BGA
device suppliers took the same approach to fabricating their parts'
substrates notwithstanding their overall constructions as die attachment
geometries in the z axis.

As all this is rather old news, there is no need to turn this information
into another article about BGA related issues with respect to coplanarity
requirements well known throughout industry. I know Peter well through the
forum and personal email conversations. I regard him most highly. I don't
know John, but he must be of the same type and has a handle, as does Peter,
on this subject.

I hope this little bit of information helps. Again, BGA's are no big deal.
They are about the easiest devices with which I've ever worked but boards,
on the other hand can be a handful and are often screwed up.

I hope you have a nice whatever time it is where you are,

Earl

---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/html/forum.htm for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------

ATOM RSS1 RSS2