Hello Arnaud - here are my thoughts:

1) The current requirement in the IPC-JSTD-001/610 specification set was
developed using transmission X-ray techniques as that was the most
prevalent in the industry so the task team focused on that equipment set.
The AXI criteria was not addressed and left as an "AABUS" item.

2) The IPC-7095 standard is a guideline document that contain information
on voids and AXI such at an individual company can develop void criteria if
they desire.

3) I believe most AXI systems have the ability to capture a transmission
X-ray image in a very reasonable amount of time which could be one solution
for your question.

4) The voiding data gathered by the IPC-JSTD-001/610 task team revealed
that the generation of voids is not an easy thing to do. Voiding was a
primary result of the type of solder paste being used and the reflow
profile for the printed circuit assembly.

The overall result of the  IPC-JSTD-001/610 task team effort was that if
the level of voiding gets anywhere near the specification maximum limit,
there is truly something wrong with the soldering process or the solder
paste which would need to be address immediately. Automated solder reflow
processes that use good materials and have adequate process controls have
voiding levels that are low, usually in the 1%-10% range.

There is an IPC-JSTD-001/610 task group working on the QFN void question
and there should be some information regarding that topic at the Fall
meetings.

Dave Hillman
Rockwell Collins
[log in to unmask]


On Mon, Aug 27, 2018 at 5:21 AM, GRIVON Arnaud <
[log in to unmask]> wrote:

> Hello TechNet,
>
> We all know the IPC-A-610 requirements on BGA voids, namely "30% or less
> voiding of any ball in the X-Ray image area" as acceptable condition.
> The % value indicated pertains to a cumulated value from a 2D topside view
> of an X-ray image like on all MXI (Manual X-ray Inspection) systems.
>
> How to manage such requirement for 3D AXI (Automated X-ray Inspection)
> systems that work on a completely different "slice" approach?
> For a given BGA, 3D AXI systems will rely on few slices (e.g. 3) of the
> BGA joint with a variable section where the voiding criterion will not be
> applicable as is.
>
> Also, how to ensure a 3D AXI process is well under control?
> For manual X-ray inspection, everything relies on the operator and it is
> easy to judge the adequate knowledge and experience of the operator.
> For 3D AXI, there is no operator, everything relies on the programing and
> the program set-up and tuning is rather tricky, typically requiring the
> support of a MXI.
>
> The recently released IPC-7095D provides excellent guidance on both MXI
> and AXI (see an extract below), but does not address the implementation of
> actual voiding requirements in AXI systems.
>
> Extract from IPC-7095D §7.3.3 :
> What generally distinguishes AXI systems from MXI systems is AXI systems
> are in-line capable and do not require an operator to make pass/fail
> decisions. MXI equipment is almost exclusively a transmission X-ray
> technology, whereas AXI
> equipment can be transmission, cross-section or a combination. MXI systems
> deliver much better resolution and much higher quality images, and as a
> result are much more accurate in manual mode. Some MXI systems provide
> automatic routines, including step and repeat, auto BGA, QFN, area void and
> solder pad analysis, but these are secondary to their main function being
> an X-ray microscope.
> AXI systems are also faster. In many cases an MXI system is used to verify
> the AXI results and help fine tune the AXI auto inspection parameters.
>
> Thanks in advance for your feedbacks.
>
> Best regards,
>
> AG
>