Abstract

The use of Dynamic Infrared (IR) Imaging is presented as a novel, valuable and non-destructive approach for the analysis and isolation of failures at a system/component level. Over the years, the Failure Analysis (FA) community has strived to develop low cost, efficient and time saving methods for the non-destructive isolation of failures. In the PC industry, more complicated designs and smaller, highly populated systems are an ever-increasing challenge for timely and effective FA. Due to these constant technological advancements and changes in the incoming System Under Test (SUT), many tools have become obsolete and are no longer effective for FA. Utilizing a test setup integrating an IR camera and image subtraction software, static IR system level fault isolation is a proven FA approach [1]. This approach is implemented by placing a known good SUT and then a suspect failing SUT under the IR camera. Both are electrically exercised into a known time static mode, and then a snapshot thermal image of each SUT is acquired. These images can be subtracted providing a method for isolating differences in the thermal signatures between known good and suspect systems. This technique provides a non-destructive but limited approach in isolating a failing component on the SUT. Limitations of the static approach include the inability to detect “timedependent” or “test specific” failures (that are not easily detectable in static environments), the lack of real-time comparisons, and the time consumption required for the SUT to stabilize in a known static state in order to capture a relevant image. A more effective approach has been developed through the use of Dynamic IR Imaging for system level fault isolation. Similar to the static IR approach, a good known SUT and then a suspect failing SUT are placed under the IR camera. In contrast to the static approach, the failure analyst does not have to wait for the SUT to stabilize in a known static state before acquiring the thermal image. In addition, timedependent and test specific failures can be identified. In any exercised state, a real time thermal image can be captured immediately and saved in an Audio/Video Interleaved (AVI) file format. This saves time and allows the investigation of dynamic, time-based failure conditions through the captured thermal images. While no commercially available software has been identified to subtract two AVI files, Dell has developed a method to separate two AVI files into frames, subtract the individual images frame by frame and reconstruct the result into a third AVI file, thus allowing a dynamic image subtraction comparison for the isolation of time-based system level failures that may be specific to a particular dynamic condition. This paper will briefly review the theory of IR thermography including the environmental (i.e. room temperature vs. SUT temperature) and material (i.e. differences in emissivity) constraints that have to be taken into account. It will describe in detail how this Dynamic IR Imaging approach is useful in isolating time-based or test specific failures. It will also describe the software and the programming methods used to achieve Dynamic image subtraction. It will discuss the advantages and disadvantages of conducting Dynamic IR Imaging analysis using the AVI file format. It will conclude by summarizing the results of our experiments and stating the advantages as well as disadvantages of using this Dynamic IR Imaging system level failure isolation approach.

This content is only available as a PDF.
You do not currently have access to this content.