The nominal pad diameter on the device is 0.030" and this is the pad
diameter on the board. Solder mask definition reduces this diameter to
0.026". Thanks for the information on drop testing. After the initial
accidental drop I think that they "tested" 4 more assemblies from what they
way was 1 foot and the BGA popped off of all of them. One of the three
assemblies that we received looks questionable in that there is burnt flux
residue on and around the BGA area. I believe that they are screen printing
paste on the entire assembly. The assembly sees two stages, single sided
SMT (paste and reflow) and then wave soldering of some through hole
capacitors and a number headers and jacks.
John
_____
From: Dennis Fritz [mailto:[log in to unmask]]
Sent: Friday, November 19, 2010 1:16 PM
To: [log in to unmask]; [log in to unmask]
Subject: Re: [TN] BGA Falling off PCA
John,
I wish I could say I am surprised at your results, but only a little. BGAs
perform very poorly in drop shock testing. Both the original JCAA-JGPP
study at
http://www.acqp2.nasa.gov/LeadFreeSolderTestingForHighReliability_Proj1.html
And now the results from NASA/DoD phase two - papers by Polina Snugnovsky
and Tom Woodrow at SMTAI last month show that BGAs fail first of all the
compents on the test boards!! JGPP data shows that over 60% of the failure
cause on JGPP boards is (drum roll, please) - the kind of component!!! BGAs
fail first in most all kinds of stress - thermal cycle, drop shock,
vibration, combined environments, but by far the first in drop shock.
In the SMTAI report data from Polina Snugnovsky at Celestica - all but one
BGA failed to pass 20 successive drops from JEDEC test procedure
JESD22-B110A. On the same board, all 90 chip scale components passed the
same test.
My theory about the BGA failures is that the weight of the balls, and the
distance the chip is from the board creates tremendous shear on the ball
solder joints when dropped. BGAs do significantly better in thermal cycle
than the early failure demonstrated in drop testing.
That said, a one foot drop, one time sounds pretty mild and I would still
examine the board closely. Notoriously, ENIG finishes have suffered from
"black pad". You will need some EDAX elemental analysis of the board
surface to get a feel for how much phosphorus shows up.
I don't think the mask defined pad can be the only issue, but typically, the
mask defined pads are on the package, and solder is allowed to completly wet
the pad on the board. I trust the pad sizes are about the same on the
package and on the board? Finally, you may have access to a dye syringe, so
that dye-and-pry testing can be done on the assemblies to see if there is
already cracking of the BGA joints from thermal problems in assembly, or if
there is a "head-on-pillow" issue of poor wetting of balls to the paste
applied to the board for assembly. (presumes the assembler is using paste
stencilling, and not just tacky flux direct to the ENIG pads on your board.
Denny Fritz
SAIC, Inc.
-----Original Message-----
From: John Parsons <[log in to unmask]>
To: TechNet <[log in to unmask]>
Sent: Fri, Nov 19, 2010 3:30 pm
Subject: [TN] BGA Falling off PCA
I am not seeing a lot of the usual Friday quiz traffic so I know you
guys-n-gals are just itching for some real work :o).
We are a PCB fabricator and one of our customers has returned some
assemblies to us for feedback. What they noticed when someone accidentally
dropped an assembly is that the BGA (352 pin Cirrus Logic EP9312-CBZ) popped
off the board. They then proceeded to drop a few more from a height of
about a foot with the same results. All the other components on this
Pb-free assembly stay fast.
We have received three PCA's and 5 BGA's. On one of the BGA's a single ball
has remained on the PCB. The rest of them all the balls have broken cleanly
from the PCB leaving a very flat (under x10 magnification) dull gray
surface. I sure wish that I could take photo's but lighting is critical and
I don't think that I have the resources to get photo's that would do justice
to what the naked eye is seeing.
Also of note is that the BGA lands are mask defined with a 31mil circuit pad
and 26mil solder mask relief.
Does anyone have any ideas as to how we can determine the root cause of this
failure. This part has been built for some time but previously the surface
finish on the PCB was a solder finish. This is the first run of the boards
that has an ENIG finish. This doesn't look to me like a raw board issue but
that is where everyone looks first isn't it ;o).
Regards
John Parsons
______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please contact helpdesk at x2960 or [log in to unmask]
______________________________________________________________________
---------------------------------------------------
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
-----------------------------------------------------
______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please contact helpdesk at x2960 or [log in to unmask]
______________________________________________________________________
---------------------------------------------------
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
-----------------------------------------------------
|
