DfR Solutions Reliability Designed and Delivered

3 Real Life Examples of “After a Failure” Investigations

Posted by Robert Manzanares on Feb 14, 2018 11:11:00 AM

3 Ways to Achieve Electronic Reliability.jpg

 

“After a failure” investigations are typically performed to identify root cause of failure, calculate risk exposure and develop mitigation and remediation solutions. Just like with “before a failure” investigation, there are two specific test methods that could be applied to either of the two categories – non-destructive physical analysis (NPA) and destructive physical analysis (DPA).

Examples of NPA are electrical characterization, acoustical, optical, electron and x-ray microscopy inspection and exposure to various environmental stresses. 

Examples of DPA are specimen cross sectioning, chemical decapsulation, Digital Image Correlation (DIC), “die and pry” solder joint analysis, sheer and pull testing of solder joints and bond wires, shock and vibration of assemblies.

Typical “after a failure” investigations are performed after an unexpected product failure occurs. Here are 3 examples of how these investigations can provide answers and improve product development.

 

1.  Detecting design errors

A manufacturer of a computer subsystem who was having a problem with memory chips on a printed circuit board assembly (PCBA) failing after final assembly.  The initial optical inspection of the failure found that the solder joints of the memory chips were cracking.  With an additional visual inspection and some dimensional measurements, it was discovered that the cast aluminum housing the PCBA was being attached to had a design error.  Several screws at various points around the perimeter and interior area of the PCBA held it in place.  The post for the screw in the middle of the PCBA, next to the failing memory chip was 1/16 of an inch shorter than all the others. The PCBA would pass all the post manufacturing tests, and then fail the final assembly test because nobody noticed the board was being bent. After the investigation, the mount post issue was resolved and the product was able to pass in final assembly.

 

2.  Determining causes of thermal damage

A company submitted an assembly to be investigated that had suffered extensive thermal damage caused by an unexpectedly high current within the PCBA. After careful examination of the pattern of charred PCBA, and information on how it was oriented during the failure event, a point of origin for the failure was determined.  Examination of the design documents, particularly the Gerber files for the board layout and bill of materials, lead to a tantalum capacitor that was identified as the root cause of failure. 

There is a well-known failure mode of tantalum capacitors, of which this particular one succumbed to, that is referred to as “surge current failure.”  At the time of power-on, or during a rapid change in supply voltage due to a changing load, a large in-rush current flowed in to the capacitor to charge it to the same voltage level of the power supply.  This current flow caused a rise in temperature of the capacitor great enough that it ignited.  When the capacitor ignited, it caused the PCB material underneath it to thermally decompose and short the interior power and ground planes.  The shorted power and ground planes then proceeded to draw many tens of amps of current from the power supply, thermally decomposing an ever-increasing area of the PCBA until the power was shut off.  The report to the customer included the recommendation to change to tantalum capacitors with built in fuses which would prevent future catastrophes of this type.

 

3.  Generating required failure analysis reports

Failures of electronics used in medical devices and nuclear power plants are required to have a failure analysis report generated, even if it’s not the first occurrence of a particular failure. 

A medical device manufacturer had several failures to a line of products that was traced to a batch of poorly fabricated PCBs.  As each failed product was returned to the manufacturer they requested DfR Solutions perform cross sectioning and optical microscopy to confirm it was the known failure.  

Several nuclear power plants have engaged DfR Solutions to perform failure analysis on components that have failed in service. Several times relays used in backup generator systems have failed, and submitted for failure analysis. Relays would be tested and measured in order to attempt to reproduce the original failure.  Optical and X-ray microscopy were utilized at several stages of disassembly to inspect for possible causes of the failure.  A root cause of failure wasn’t always determined, but a report detailing the findings of the investigation was submitted to the customer to be included with their report to the appropriate regulatory agency. 

Even seemingly simple devices have required a failure analysis to be performed because they were used in a nuclear power plant.  On more than one occasion a large thermal fuse was submitted for a failure analysis.  In one particular case, the root cause of failure was old age.  The fuse had been originally placed in service sometime in the early 1980s.  After more than 30 years, and an indeterminate number of times that the power had cycled off and on, the fusing element developed a high resistance which led to a hot spot and eventual failure. 

 

DfR Solutions offers many services: reliability and accelerated life testing, design review and root cause analysis, that can enhance a company’s bottom line and increase customer satisfaction by producing more reliable, long lasting products. Contact us for a quote today.

 

Join DfR Solutions in March for the 2018 Design for Reliability Conference for Real-World Applications of Physics of Failure Solving Today's Technology Challenges.

Topics: Failure Analysis, reliability testing