Advanced Detection Techniques for Counterfeit Electronic Components

By Dr. Nicholas Williams, Technical Director SMT Corp.
September 16, 2026

Counterfeit electronic components detection is not always straightforward.

Traditional visual inspection remains an important part of counterfeit mitigation and can reveal signs of remarking, resurfacing, refurbishment, or other forms of alteration. However, some counterfeits are much more difficult to detect. Cloned devices, in particular, can closely resemble authentic components and may not exhibit an obvious defect during initial testing.

In my presentation at the 2026 Symposium on Counterfeit Parts and Materials, I reviewed several advanced counterfeit component detection techniques included within SAE AS6171, with particular focus on burn-in and 3D radiography. The objective was to examine what these methods can tell us when evaluating suspect components and where they are most useful.

Before going deeper into the individual techniques, it is important to understand how testing levels are structured within AS6171 and where these advanced methods fit within the overall risk-based approach.

What Makes a Test Method “Advanced”?

AS6171 includes a range of counterfeit test methods beyond the standard moderate risk testing methods.

In AS6171, Moderate Risk Model 2 represents the minimum default testing. For the purposes of my presentation, I used the term “advanced” to describe the additional methods that may be included in high-risk and critical-risk test plans.

These methods include:

• Bond Pull and Die Attach — AS6171/4
• 3D Radiography — AS6171/5
• Acoustic Microscopy — AS6171/6
• Burn-In, Temperature Cycling, Fine and Gross Leak — AS6171/7
• RAMAN Spectroscopy — AS6171/8
• FTIR Spectroscopy — AS6171/9
• Thermogravimetric Analysis (TGA) — AS6171/10
• Design Recovery — AS6171/11

These methods provide different types of information.

For example, RAMAN and Fourier Transformer Infrared (FTIR) spectroscopy can be used to examine material characteristics by producing a spectrum that acts as a material fingerprint. These techniques utilize the absorption and emission of light to provide qualitative and quantitative information about the chemical composition of materials.

For FTIR spectroscopy, an infrared light source transmits through the ATR crystal to the sample which reflects a portion of the infrared light back through the ATR crystal to a detector. In RAMAN spectroscopy, the material excitation from the laser results in scattering of phonons at different wavelengths than the incident laser.

RAMAN and FTIR spectroscopy are particular useful for detecting remarked counterfeits. For example, FTIR can detect the difference in material composition due to the remarking layer on the top surface of a component when compared against an established plastic encapsulation model. Whenever there is a layer of remarking material, the correlation between the remarked component and the plastic encapsulation model is poor.

Burn-in, on the other hand, allows evaluation of electrical behavior following extended exposure to controlled temperature and electrical conditions.

3D radiography provides a more detailed and in-depth perspective by allowing examination of the entire internal construction of a component.

With that framework in mind, we can look more closely at the individual methods and what they can tell us when evaluating counterfeit components.

Burn-In Testing for Counterfeit Mitigation

The traditional application of burn-in testing is to identify early-life failures associated with manufacturing defects, damage, or design flaws. The basic concept is to operate a component under specified conditions for an extended period and then verify its performance post-burn-in.

For counterfeit mitigation, burn-in can serve another purpose.

For traditional counterfeit components, such as remarked, resurfaced, or refurbished devices, burn-in may help identify components where previous use has already consumed a significant portion of the device’s useful life or where electrical damage has been introduced during refurbishment.

For cloned components, the question is somewhat different.

A cloned semiconductor may have been produced using manufacturing processes that do not meet the same standards as the authentic device. Burn-in provides an opportunity to place those devices under controlled stress and observe whether the counterfeit manufacturer properly screened and the qualified components to industry standard methods.

For this evaluation, I focused on burn-in for microcircuits according to MIL-STD-883 Method 1015, as referenced by AS6171/7. AS6171 refers to various military standards for burn-in methods which depend upon the component type being tested.

Under Method 1015, burn-in includes operation at a specified temperature for a defined duration, pre- and post-electrical testing, and either a static or dynamic stimulus depending on the test condition. One standard condition for military-grade devices is +125°C for 160 hours.

To better understand how burn-in could be applied as a counterfeit mitigation technique, I evaluated several clone devices alongside known authentic components.

Burn-In Case Study: Authentic vs. Clone Components

Four clone counterfeit clones were evaluated:

• Clone Example A — FPGA configuration memory
• Clone Example B — FPGA configuration memory
• Clone Example C— logic device
• Clone Example D— linear regulator

The samples were serialized and subjected to pre- and post-electrical testing, with delta measurements used to evaluate changes in electrical parameters. Known-good devices were included for comparison. The burn-in temperature was +125°C for a duration of 160 hours. Dynamic stimulus was used for the Clone A, Clone B, and Clone C, while static stimulus was used for the Clone D.

Component	Part Number	Stimulus	Post-Burn-In Findings
FPGA Configuration Memory	Clone A	Dynamic	No gross failures. Clone devices showed an average ~150% change in leakage current, compared with ~55% for authentic devices.
FPGA Configuration Memory	Clone B	Dynamic	No gross failures. Clone devices showed an average ~135% change in leakage current, compared with ~84% for authentic devices.
Linear Regulator	Clone C	Static	No gross failures. Clone devices exhibited greater changes in output voltage at higher output currents, including >50 to 300 mV changes at 0.5 to 1.5 A.
Logic Device	Clone D	Dynamic	No gross failures. No significant differences in delta measurements were observed between clone and authentic devices.

The results are worth looking at more closely, particularly because the devices did not all respond in the same way.

For the Clone A and Clone B FPGA configuration memory devices, both authentic and clone devices retained their memory contents after burn-in, and no gross failures were observed. The difference appeared in the electrical measurements: the clone devices showed approximately two to three times greater percentage changes in leakage current than the authentic devices.

The Clone C Linear Regulator showed a different type of behavior. Again, there were no gross failures, but the clone devices exhibited greater changes in output voltage at higher output currents. Changes of greater than 50 mV and up to approximately 300 mV were observed at output currents between 0.5 A and 1.5 A.

The Clone D logic device provided an important contrast. No significant differences in the delta measurements were observed between the clone and authentic devices.

This also highlights an important limitation of burn-in as a counterfeit detection technique: a meaningful distinction will not necessarily appear in every device or every parameter.

Why Establishing a Baseline Matters

The pre-burn-in electrical measurements establish a baseline for each component. After burn-in, the measurements can be repeated and compared to determine whether the device has experienced an acceptable amount of parameter drift.

In a counterfeit evaluation, that comparison becomes particularly important when there is no gross failure.

If a device simply stops functioning, the result is relatively straightforward. The more difficult situation is when the component continues to operate but its electrical characteristics have changed.

In those cases, having a known authentic device for comparison provides valuable context.

The results from this evaluation demonstrated that point clearly. All of the devices remained functional after burn-in, but several of the clone devices showed greater parameter drift than their authentic counterparts. With burn-in, we are looking for any deviations to our baseline measurements.

That distinction is important when determining how and when to use burn-in as part of a counterfeit component evaluation.

Burn-in should therefore be selectively used for high- and critical-risk cases, particularly when the additional information gained justifies the time and resources required.

Looking Beyond Electrical Behavior

Burn-in gives us valuable information about how a component behaves electrically under stress but by itself does not provide the full picture of component authenticity.

We may also need to dig deeper into the physical construction of the device being evaluated. This brings us to another advanced method included in AS6171: 3D radiography.

3D Radiography for Counterfeit Component Detection

Conventional 2D radiography is an established method for examining the internal construction of electronic components. It can provide useful information about features that cannot be inspected externally.

The limitation is that a 2D image is still a two-dimensional representation.

Internal features located at different depths can overlap in the resulting image, which can make certain structures difficult to interpret.

AS6171/5 describes 3D radiography as a method intended to enhance the inspection and imaging performed using conventional 2D radiography.

The difference becomes more apparent when we look at how the 3D imaging process works.

How 3D X-Ray CT Works

For the evaluation presented, a Nordson Dage Quadra 5 X-ray system with micro-computed tomography (µCT) was used.

Rather than capturing a single image from one direction, the component is rotated through 360 degrees while multiple X-ray images are collected at each degree angle. The resulting data can then be reconstructed into a three-dimensional representation of the component’s internal structure.

This provides the ability to examine internal features from different perspectives.

Instead of viewing only the topside angle of a component, for example, we can examine the reconstructed structure from different angles, including from the top, bottom, and angled side views.

That can be particularly useful when internal features overlap in a conventional 2D radiograph.

Note that a practical consideration when exploring this test method is that high-resolution 3D imaging can require substantially more time than conventional radiography. In the evaluation presented, imaging times of approximately 45 minutes to two hours were noted. The additional imaging time also results in higher radiation exposure to the component, which can limit applicability to destructive samples or failure-analysis applications.

The presentation included examples involving components such as the:

Molex 749603018:

Broadcom AMMP-6120-TR1:

Semiconductor NLAST4599DTT1G:

The value of the additional information becomes even clearer when we consider a specific case.

3D Radiography Case Study

One of the components evaluated was an Agilent HSMP-3822-TR1:

During 2D radiographic examination, a bond wire appeared to be broken. However, using 3D x-ray inspection, it was observed that the bond wire was bonded to part of the lead frame. Electrical testing later verified that connection when forward conduction was measured for both diodes.

This is a good example of why different inspection methods should not necessarily be viewed in isolation.

The 2D radiographic image raised a question about the physical condition of the device. The electrical results provided another piece of information about how the device was functioning.

Neither observation should be treated as the complete answer. Instead, the findings need to be considered together and, where appropriate, investigated using additional methods.

Using Multiple Test Methods to Build Evidence

When evaluating a suspect component, the important question is not simply whether a particular test produces a pass or fail.

The more useful question is:

What does this test tell us, and what additional information do we need?

Burn-in can tell us about changes in electrical behavior following exposure to controlled stress.

2D radiography can provide information about internal construction.

3D radiography can provide a more detailed view of internal structures when conventional radiography does not provide sufficient information.

Typical AS6171 methods such as External Visual Inspection, XRF, SEM and decapsulation can provide important visual and materials information about the device.

Other advanced AS6171 methods, including RAMAN, FTIR, acoustic microscopy, bond pull, die attach evaluation, and other analytical techniques, provide additional types of evidence.

The appropriate combination depends on the component, the risk level, and the questions that need to be answered.

Looking at these methods together brings us back to the broader purpose of counterfeit component testing: building enough evidence to determine whether a component is consistent with what it is represented to be.

Conclusion: What Advanced Testing Can Tell Us

The evaluations presented at CALCE demonstrated both the value and the limitations of advanced counterfeit detection techniques.

During burn-in, all of the devices remained functional, but several of the clone devices demonstrated greater parameter drift than their authentic counterparts. Leakage current changes were particularly notable in Clone A and Clone B, while output voltage changes were observed in Clone C. Clone D did not demonstrate a significant difference.

3D radiography provided another form of evidence by allowing internal construction to be examined in greater detail than conventional 2D radiography. However, the additional imaging and processing time limits its throughput and can affect how the technique is applied.

The main takeaway is that no single test method provides complete counterfeit detection coverage.

Advanced testing is most useful when we understand the question we are trying to answer and select the appropriate method to provide that information.

A component does not necessarily have to fail catastrophically for testing to reveal something important. Likewise, passing one test does not necessarily resolve every question about its authenticity.

Counterfeit detection is ultimately about building evidence.

The more sophisticated the suspected counterfeit, the more important it becomes to use complementary methods, understand their limitations, and evaluate the results together.

^{1  R. Shaw, “GaAs MMIC Reliability Assurance Guideline for Space Applications”, December 15, 1996}