I have seen this a few times. Here is what is what may be happening:
Many BGAs are made with a temperature sensor circuit that is supposed to
shut down the component if the operating temperature range is exceeded.
Often this portion of the circuit is not designed properly and/or not
manufactured properly. A certain quantity of the chips do not work, but
they get put into more than one lot of packaged BGAs, so there is
sometimes not a "defective lot" per se.
Many times the test techs think it is a bad solder joint because they
can put their finger on the part and it either starts working or stops
working. This may not be because of a intermittent mechanical solder
joint connection, but because the BGA senses the temperature difference.
Upon investigation, it turns out not to be the solder joint (looks great
in the X-rays, can't see anything visibly wrong with the solder ball
connections on the periphery, etc.).
Oftentimes, however, there IS a problem with the solder joints, and it
is difficult to determine whether the solder joints are bad or some
other issue such as timing issues is causing the circuit to malfuntion.
Here are the steps to verify the problem is the BGA or the circuit
design:
1. When building a new design that uses BGAs, build one "golden unit"
but leave the BGA(s) off and solder an emulation socket(s) in its place.
The socket should be of the crowned-pin style (not spring contact
style).
2. You can now use the golden board to test BGAs removed from the
failing assemblies. After removal, simply run a solder pencil with
plenty of flux applied over the bottom of the BGA to remove the excess
solder and leave a low-profile mound of solder on the BGA pads. Then
wash this suspect BGA and put it into the emulation socket on the golden
unit and power up and test. If the BGA works on the golden unit, you
know it was either a performance issue or the solder joints, but
probably not the BGA itself.
3. Next, screen several new BGAs using the golden unit. If some fail, it
then becomes apparent there are timing issues with some assemblies or
some other performance issue associated with the BGA. But take one of
the BGAs that passes test using the golden unit and rework it onto the
failed unit. If it works, the performance issue is in the BGA, if it
does not work then the problem is in the assembly.
Using crowned pins on the emulation socket provides a means of testing
either new BGAs with solder balls attached, or removed BGAs with a
coating of solder on the pads.
With this method, you can use the golden unit as a test system to screen
good BGAs from the lot, to confirm that removed BGAs are in fact
defective, or that there are performance issues between some lots of
BGAs or some lots of assemblies. It is a valuable tool to have in the
test department.
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Tempea, Ioan
Sent: Friday, September 23, 2005 2:40 PM
To: [log in to unmask]
Subject: [TN] Soldering or design problem?
Hi Technos,
we are faced here with a very delicate problem: the customer suspects a
workmanship issue on our side, we are suspecting a design or component
problem.
Here it goes: 50% of the cards do not function due to the same problem,
something wrong with the same PBGA. We tweaked the thermal recipe, we
are at about 220C peak; no improvement. We played with lower
temperatures for fear not to thermally stress the part; no improvement.
Now I get into the weird zone:
* if the BGA is heated, it instantly starts working, but also
starts working if the BGA is frozen
* it also works if we freeze the opposite side of the PCB
* We use the flying probe to test for open joints, everything
seems to be OK, but the card does not work
* we did all kinds of re-reflow of the BGA on the bad boards, in
the oven, but also on the rework station, even at extreme temperatures
that send the top of the part at 280C. Same pattern, heat/freeze, it
works, ambient temperature - fail. So we didn't even manage to kill the
part for good.
* we change the component, bingo, it works in 90% of the cases
I am sure the soldering is OK and the joints are properly formed, but
don't have a solid proof, for some people BGA soldering is still the
cause of all problems.
My question is: if it's not the soldering, what else can it be? Can an
electronic design be borderline so that it does not work with the first
time reflowed BGA, but works right after replacing the part with parts
from the same tray? Can something be in the design of the PCB, in the
manufacturing of the PCB?
And a last one, what is the test that can confirm the absence of open
joints under a BGA? Can the balls be cross-sectioned parallell with the
PCB, like half-ball, so an eventual open be redily spotted?
Thanks,
Ioan
---------------------------------------------------
Technet Mail List provided as a 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/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 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/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------
|
