Hi Bob,
I want to add to the previous Werner's e-mail. To be exact, solder
joints fail because of crack growing. Obviously, there are shouldn't be
any cracks in as formed solder joints, so a crack in a joint has to be
formed to result in final failure.
Overall plastic deformation of solder joints (accumulating creep-fatigue
damage, as Werner said) will result in formation of a crack. However,
I'm not sure whether the overall plastic deformation or localized
processes around the crack tip will play a role in its further
propagation.
Regards,
Vladimir
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Robert Kondner
Sent: Tuesday, December 09, 2008 6:57 PM
To: [log in to unmask]
Subject: Re: [TN] Xbox360 BGAs
Hi,
Ok, thanks.
Dare I ask further but:
If the material creeps at normal/high temps that would tend to reduce
the
stresses, right? Why then would a joint tend to fail on repeated
temperature
changes, especially going hot?
That Xbox photo showed really squished solder balls. At elevated
temperature the creep was probably high and the balls squished due to
the
force of the bottom board warping. (An Al heat sink probably holds the
top
chip pretty flat.)
If balls got that squished to the thickness of paper I would almost
expect
shorts to form between squished balls. Do we know if the unit failed
with
opens or shorts.
Would placing a rigid metal stiffener under the bottom side (Bottom
side
heat sink) help reduce warp? Would under fill on a large BGA help do the
same thing?
Thanks,
Bob Kondner
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Werner engelmaier
Sent: Tuesday, December 09, 2008 6:30 PM
To: [log in to unmask]
Subject: Re: [TN] Xbox360 BGAs
Hi Bob,
Forget what you know about metals when talking solder---most of it does
not
apply; we do not use any other metal this close to ts melting
temperature.
At Th>=0.5 [Th=T[K]/Tm[K]] all metals will creep significantly. Thus,
any
elastic strains get converted to plastic strains in short [depending on
T]
order, thus reducing any stresses. So the pre-stressed concrete analogy
does
not apply.
At those T's stresses are not the issue, it is displacement strains.
Werner
-----Original Message-----
From: Robert Kondner <[log in to unmask]>
To: 'TechNet E-Mail Forum' <[log in to unmask]>; 'Werner engelmaier'
<[log in to unmask]>
Sent: Tue, 9 Dec 2008 8:57 am
Subject: RE: [TN] Xbox360 BGAs
Hi,
I have a general question about:
Solder joint strains cause by thermal expansion differences.
It would seem to me that cooling of a assembly after reflow would
introduce built in strains. It seems very much like a pre-stressed
concrete
beam. It would seem that exposing a device to cold temperature would
increase these stresses.
Question:
Do crystals in a solder ball "Flow" at room temperature?
Can a solder ball be annealed with a soak at some temperature?
Thanks,
Bob Kondner
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Werner engelmaier
Sent: Monday, December 08, 2008 6:36 PM
To: [log in to unmask]
Subject: Re: [TN] Xbox360 BGAs
Hi David,
You are somewhat off.
Nothing a SJ can do to stop being deformed---whatever the expansion
mismatch
is will wind up in the SJ.
HOWEVER, for the same thermal expansion mismatch, different solder joint
heights will result in different cyclic strain ranges
[ that is why IBM makes solder columns].
The corner balls seethe highest thermal expansion mismatch because of
the
largest DNP, and the strain is further increased because of the
compression.
Werner
-----Original Message-----
From: David Tremmel <[log in to unmask]>
To: [log in to unmask]
Sent: Sun, 7 Dec 2008 10:17 am
Subject: [TN] Xbox360 BGAs
Hello Technetters,
Since I have had some people wanting to see pictures, I have resubmitted
my
email with some corrections.
I had someone who got a used Xbox360 that failed and when we
disassembled
it, we found they had put an aftermarket heat sink on two of the largest
BGAs.
The original Microsoft heatsinks make contact with the dies (the BGAs
are
flipchip designs) and have four arms that extend past the corners of the
BGA
and then have posts which go through the board to a latching piece on
the
bottom of the PCB directly underneath the BGA which adds tension to make
good physical contact with the top of the die and the bottom of the heat
sink but does not add sufficient pressure to deform the PCB. The
aftermarket heatsinks have posts with spacers above the PCB and washers
below the PCB but the spacers above the PCB are not the proper length
and
after time, the PCB warps upwards and the spacing at the corner of the
BGAs
was severely compressed to the point where I could not put 2 pieces of
paper
between the corner of the BGA and the PCB. The solder spheres are 25
mil
and are probably being compressed to less than 10 mil.
Anywho, this is what I think has the failure mechanism is and I would
appreciate some feedback from the gurus:
While the chip operates, it generates heat which expands the device to
some
degree in the X,Y axis
Part of the job of the solder spheres is to sink the heat to the logic
board
so the thermal expansion of the BGA does not act like a sheer force at
the
solder ball interfaces. The logic board also expands due to the heat
and
any CTE differences between the BGA and the logic board deform the
solder
spheres to some degree. The BGA substrate size is 35mm x 35mm and the
solder spheres on the corner of the BGA are subject to the greatest
sheer
forces which, in this case, is detrimental.
Due to the compression of the solder spheres at the corners because of
the
poor heatsink design, they are unable to deform and any CTE difference
between the BGA and the logic board are turned into sheer forces and
cause
ball/interface failures.
I also think that the constant pressure of the lead free solder spheres
would greatly increase the chances of tin whiskers.
I have a picture of just how bad the solder sphere compression is if
anyone
is interested.
Am I way off base?
Thank you in advance for any correction in my theory. A confirmation
would
be better!!
Thank you,
David Tremmel
<http://valurecovery.com/> http://ValuRecovery.com
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0
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/contentpage.asp?Pageid=4.3.16
for
additional information, or contact Keach Sasamori at [log in to unmask] or
847-615-7100 ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0
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/contentpage.asp?Pageid=4.3.16
for additional information, or contact Keach Sasamori at [log in to unmask]
or
847-615-7100 ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0
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/contentpage.asp?Pageid=4.3.16
for additional information, or contact Keach Sasamori at [log in to unmask]
or
847-615-7100 ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0
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/contentpage.asp?Pageid=4.3.16 for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100
ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0
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/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------
|
