High speed bondtesting correlates with board level drop testing.

Bondtesting and drop testing

Traditional bondtesting carried out at low speeds (less than a 1mm/sec for shear and 5mm/sec for pull) relies largely on the highly accurate measurement of the forces needed to induce solder failures. Indeed it could be argued that this type of testing is not examining the ball to bond interface at all, but is really acting as a materials tester for the solder.

In order to subject the bonds themselves to a robust examination, device manufacturers have had to resort to board level drop testing. In this method, a number of packages under test are assembled on a circuit board, connected by an electrical daisy chain, and dropped under controlled conditions until failure occurs. This technique is expensive and time consuming in terms of set up, execution and analysis. The essence of drop testing is that the solder joints are subjected to similar loadings that they might experience during manufacture or use in a portable device. The principal failure mode is brittle fracture failure, especially in Pb-free devices.

The key concept of high-speed bondtesting is that it does exactly the same by using high strain rates during the test, comparable to those of drop testing. Just as in drop testing, high speed bondtesting produces brittle fracture failures in Pb-free materials. But now the testing is at the component rather than board level. Tens of bonds can be assayed in an hour with detailed information on fracture force, energy and failure mode. High speed bondtesting can quickly provide data on the differences between solder alloys, surface finishes and the quality of bumping processes.

High speed Bondtesting versus Drop - An independent study

Over the past six months Dage has commissioned an independent study involving a leading academic institute and a top-rank electronics manufacturing consultant. The prime focus of this study has been a direct comparison of high speed shear and pull testing using the Dage 4000HS with board level drop testing. Seven different packages were used in the study with a variety of solder alloys and surface finishes. All were subjected to thermal ageing for different periods.

The results of this work have been outstanding, showing a strong relationship between high speed bondtesting and board level drop testing. This work will be reported in detail at ECTC2007 on May 29th to June 1st in Reno, Nevada. Following this publication, further details will be appearing on this web site.

© 2008 Dage Holdings Limited