> I would like to have some feedback on how a "not approved" BGA solder joint look like" (except
> for an open or bridge). I am referring to eutectic collapsable solder joints.
>      
> Can you have "too fat" or "too thin" joints?
>      
> Hans Bogren

>      To add to Hans' question what is an allowable amount of voids in a BGA 
>      joint.  Do IPC standards exist on this?
>      
>      Thad_McMillan
The obvious rejectable defects would be opens, shorts,  non-reflowed 
solder, and solderballs.  The subjective characteristics would be voids, 
insufficient, and excess.  

With respect to voids, responses will vary between the old mil-spec type of 
"voids are bad" answers to the "it depends on your environment" 
answer.  One school of thought is that the voids provide a stress 
relief if a crack were to develop, while others believe that the void 
will decrease the  distance through which the crack needs to 
propagate before there is a functional failure.  I'm sure there is a 
practical limit,  for lack of better numbers I would be concerned if 
you are seeing greater than 25% voiding.  The best approach to take 
is that of a  process indicator.  If your process did not produce 
solder voids in  BGA's and suddenly everything has voids in it, you 
should be  concerned about your process. This could be indications of 
an incorrect reflow profile.

As far as the "too fat" or "too thin" question, the real item of 
interest is why is there so much variability in the solder volume. 
Usually this is an indication of a stencil printing problem however 
I have seen a case through our helpline where there were very thin connections at the 
center of the BGA and severely excess connections as you moved away from the 
center.  In this case, the "thin" connections had slightly smaller 
radius and less thickness  than expected.  We believe this was a 
result of popcorning the device during reflow.


There is a gentleman in England with the European version of IPC that 
has taken a stab at a spec for X-Ray inspection of solder 
connections, unfortunately his answer may be a little more vague 
than you would like.

***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************