Previous Page  8 / 58 Next Page
Information
Show Menu
Previous Page 8 / 58 Next Page
Page Background

edfas.org

ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 19 NO. 1

8

The difference from the BGA in Fig. 8 is that the molten

solder ball did not wet the surface of the PCB land. The

faulty solder print could also be identified next to the

defective BGA on a bare printed pattern without an

assembled BGA (Fig. 11). Once again, the root cause of

the failure was a dried-out, clogged stencil while printing

the solder paste.

SUMMARY

Based on a number of case studies, it was shown that

failuremay originate at a number of steps in themanufac-

turing process, for example, fromthe solder paste printing

process or from the soldering process itself. The stencil

printing of the solder paste is often a critical process, but

more often, the root cause is faulty PCBs.

Cross sections of solderedBGAs are common for failure

analysis. One newapproach is presented as an additional

method in the course of finding the root cause of failures.

Grinding away the corpus of the BGA until the soldered

balls are still left on the PCB can provide insight into the

failure mechanism.

In the case of perfect solder connections, grinding the

PCB from the backside is another approach, whichwill be

discussed in Part II of this article in the May issue of

EDFA

.

ABOUT THE AUTHOR

Gert Vogel

studied physics in

Stuttgart. He has been with Siemens

for more than 30 years. Dr. Vogel

was a semiconductor technologist

in Siemens’ DRAM production in

Munich and Regensburg for seven

years. He then moved to Siemens

Amberg where, among other topics, he is a specialist

in failure analysis of electronic components on printed

circuit board assemblies. He led a tutorial, “Avoiding

Flex Cracks in Ceramic Capacitors,” at ESREF 2015.

This was followed by a tutorial, “Creeping Corrosion

of Copper on Printed Circuit Board Assemblies,” at

ESREF 2016.

FAILURE ANALYSIS ON SOLDERED BALL GRID ARRAYS: PART I

(continued from page 6)

Fig. 9

After grindingdown theBGAandbeforeblowingaway

the residue of the solder resist fromthe bottomof the

BGA, it was discovered that one ball had dropped off.

Fig. 10

After blowing away the residue of the solder resist,

two missing balls can be seen.

Fig. 11

The faulty solder print can also be seen on a printed

patternwith the same geometry next to the defective

BGA.