Derek Snider at dual beam stage with star wars R2D2 toy

Semiconductor and Electronic Failure Analysis Blog

Welcome to the Semiconductor and Electronics Failure Analysis Blog, and discussion forum for all things related to electrical, integrated circuit (IC) board and electronics failure analysis.  Please subscribed to our feed and feel free to leave a comment or question. Thanks for visiting.

A Study in Printed Circuit Board (PCB) Failure Analysis, Part 2

Continued from A Study in Printed Circuit Board Failure Analysis, Part 1

Printed Circuit Board Failure AnalysisThe next step in the failure analysis process, revealing the defect, would almost certainly involve destruction of the board; as a result, a strong hypothesis was necessary before embarking upon any further analysis. In order to determine the best course of action, our analyst reviewed the facts as they stood before proceeding.

 

  1. The failure can be thermally modulated – as board temperature increases, the failure becomes more pronounced.
  2. In the failing condition, high resistance is measured between two points on the same node. This high resistance results in reduced output current.
  3. No signs of solder quality issues – cracking or non-wetting – were noted in the area of the failure.
  4. X-Ray inspection did not reveal any signs of damage to the copper trace between the two suspect points.

Given this list of facts, our analyst determined that the most likely cause of failure was an intermittent contact between the two points in question that became worse under thermal expansion (as the board materials heated and expanded, less material remained in contact to conduct electricity). The most likely location for this type of failure would be at the connection between the copper trace and the barrel of a via or plated thru-hole; given this hypothesis, the analyst elected to cross-section through the PTH for the suspect pin.

The area of interest was cut away from the bulk of the PCB and encapsulated in epoxy. A cross-section was performed by grinding into the suspect pin with progressively finer grits of polishing abrasive, finishing with a sub-micron polishing compound to bring the sample surface to a finish suitable for high magnification imaging. The PTH was imaged with a high power optical microscope; as hypothesized, an incomplete connection between the copper trace and the PTH barrel was noted (reference figures 5 and 6). The analyst had the proverbial smoking gun; now, the only remaining step was to tie the physical defect to its most likely cause.

Though the physical defect had been revealed, the analyst’s job was not over; the goal of any failure analysis project is to find the root cause of failure and determine the most likely origin of any existing defect. Of the many possible explanations for this type of failure, two were considered as the most likely candidates:

  • Mechanical stresses (vibration, thermal cycling, board flexure) may have broken a trace that was originally well connected
  • Insufficient etchback or smear removal (followup after drilling holes in the board) was performed during the PCB manufacturing process, preventing a good bond between the buried traces and the barrel

If mechanical stresses were the root cause of this failure, an analyst would expect to see much more damage to the PCB’s copper traces (some degree of tearing or other stress-related cracking); other than the separation from the barrel wall, no such damage was noted. Improper cleaning and etchback during manufacturing, on the other hand, could very well result in an incomplete bond between a buried trace and the via barrel. The defect was therefore classified as most likely occurring during manufacturing. Corrective action was implemented by adding additional inspection and destructive physical analysis on incoming PCBs per IPC-A-600 as part of production screening; as a result, other failures similar to this were found before reaching the end user, and the PCB manufacturer was able to identify and correct an inadequacy in their process.

Conclusion

In this case study, we examined how the failure analysis process enables a defective part to produce actionable data that suppliers and manufacturers can use to improve their product. Despite starting with little more than a nebulous problem description – “this doesn’t work” – the analyst was able to methodically work towards a more comprehensive explanation of the failure; in doing so, an expensive chunk of scrap was transformed into a valuable source of knowledge, identifying a process weakness and helping to prevent further defective product from reaching end users. Future columns will continue to provide other approaches to failure analysis of printed circuit boards (PCB), components, and other electronic devices; in the mean time, keep an open mind, and remember that failure is nothing more than an opportunity to improve!

Derek Snider is a failure analysis engineer at Insight Analytical Labs, where he has worked since 2004. He received his Bachelor’s of Science in Electrical Engineering from the University of Colorado at Colorado Springs.

This article originally appeared in the September 2013 issue of SMT Magazine.

Read 7467 times

Search Our Site

Contact Us

Please type your full name.
Please type your full name.
Invalid phone format. Please use xx-xxx-xxxx
Invalid email address.
Invalid Input
Invalid Input

ISO-9001:2007 Certified

Follow Our Feed

Enter your email address:

Delivered by FeedBurner

Login

 

Need to Determine the Root Cause of a Failure in an Electronic Component?  We get back to you with a quote in 24 hours once we have your information.

Request Failure Analysis Quote