Stephen R. Gregory wrote:

> In a message dated 5/6/99 5:36:37 AM Pacific Daylight Time,
> [log in to unmask] writes:
>
> <<
>  Steve: Thanks for your explanation of IPC's activities on BGA
> voiding. Just
>  two 'small' questions (maybe 'big' answers??):
>
>  i./ Is that 20-30% by cross-sectional area, or volume, or what?
>
>  ii./ Is there any comment also about WHERE the voids are located -
>  solder/UBM or solder/pad?
>
>  Ta,
>
>  Andy>>
>
> Hi Andy!
>
> I'm sure it was cross-section, because you can't tell how big the void
> really
> is using just transmission x-ray...and as to where is it is another
> gray area
> using transmission x-ray, you just don't know. It does make a
> difference
> where it is, I'm sure. But that points out why one should err on the
> side of
> caution considering the size of the void.
>
> I'm sort of guessing now, but I feel that one might assume that a void
> is
> going to be in the solder fillet between the ball and the PCB pad
> unless you
> can somehow determine that the BGA has voids in the balls themselves
> before
> assembly. In which case I would send them back from where they came
> from.
>
> Just another lil' update on the conference I'm attending. I went to
> the first
> couple lectures this morning, and won't be going back until this
> afternoon...all the lectures between now and then are about down at
> the wafer
> level of these lil' puppies.
>
> But what was interesting from the first lecture this morning, which
> was
> titled; Key Factors in CSP BGA Reliability, by a Dr. Reza Ghaffarian
> from
> JPL, was that all the reliability studies and tests that have been
> done, and
> there's been a TON of them! (He had more charts and slides than I
> could keep
> up with!) was that they were all done and performed with BGA's and
> CSP's that
> were soldered with traditional Tin/Lead solders. I asked if there had
> been
> any testing done with any of the alternative solders, and he had said
> no.
>
> "Things that make ya' go hmmmmm...."
>
> Sorry, I just can't drop this lead free thing...as the Georgia sheriff
> says;
> "I think yo' inna heap of trouble boy!"
>
> -Steve Gregory-
>
> ################################################################
> TechNet E-Mail Forum provided as a free service by IPC using LISTSERV
> 1.8c
> ################################################################
> To subscribe/unsubscribe, send a message to [log in to unmask] with
> following text in the body:
> To subscribe:   SUBSCRIBE TechNet <your full name>
> To unsubscribe:   SIGNOFF TechNet
> ################################################################
> Please visit IPC's web site (http://www.ipc.org) "On-Line Services"
> section for additional information.
> For technical support contact Hugo Scaramuzza at [log in to unmask] or
> 847-509-9700 ext.312
> ################################################################

  Dear Technetters!
In Bob Willis' s  web site about transmission X-ray for BGA the 20-30%
voids
refers to void size as seen through the whole ball. And more, the X-ray
machine I use, is able to calculate the void % using an image processing
programme from Nicolet.
What I have learned during my short experience, is that although cross
sectioning is very accurate and the surest method to see the
distribution of the voids and their position, it has many drawbacks if
you don't have an X-ray machine, the biggest being that you never know
which solder ball row to cut in order to see a defect.
In short, if you decide to check your process for voids, you can not use
one without the other.
The interpretation of the 20-30% criterion is still a mystery for me.
The vertical cross section of the solder ball will reveal only voids
found in this specific area in the XZ plane.. The X-ray shows all the
voids in the ball projected on the the projection of the whole ball as
seen in the XY plane.
There are recent statements that voids are even beneficial in avoiding
the propagation of cracks, so where do we stand now??
Gaby

################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe:   SUBSCRIBE TechNet <your full name>
To unsubscribe:   SIGNOFF TechNet 
################################################################
Please visit IPC's web site (http://www.ipc.org) "On-Line Services" section for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################