Method, apparatus and non-transitory computer medium for detecting defects of device under test using time domain reflectometry

Abstract

The present invention relates to a method, apparatus and / or system for soft defect modeling of measurements using time domain reflectometry. The soft defect in the detected chip can be rapidly detected based on an electro-optical sampling time domain reflectometry (EOS-TDR). For example, the EOS-TDR may detect soft defects in each pin from a trace-structure point at a relatively high resolution. To interpret the results in a time-sensitive manner, a reference model of the chip can be established from chips known to have met the expected quality criteria. Through the automatic analysis of the characteristics of the waveform of the detected device, the soft defect of the chip, which cannot be detected under time constraint, temperature change, applied current change, applied voltage change, vibration change, moisture change or any other possible change, can be detected.

Description

Background Art

[0001] In certain applications such as chip manufacturing, defects may not actually be detected during various stages of the manufacturing process. Detection of soft defects that affect the long-term sustainability of a chip is particularly important for chips implemented in certain industries such as the automotive industry, where long-term reliability of the chip is necessary for the entire expected lifetime of the end product. Summary of the invention

[0002] Embodiments are directed to a method of determining whether a device under test meets a quality standard. The method includes measuring the device under test using at least one time domain reflectometry measurement of the device under test to generate a device under test waveform. The method includes comparing the device under test waveform to at least one reference waveform. The at least one reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet the quality standard. The method includes determining whether the device under test meets the quality standard based on a result of the comparison.

[0003] Embodiments relate to an apparatus for determining whether a device under test meets a quality standard. The apparatus comprises a measuring unit configured to measure the device under test using at least one time domain reflectometry measurement of the device under test to generate a device under test waveform. The apparatus comprises a comparing unit configured to compare the device under test waveform with at least one reference waveform. The at least one reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet the quality standard. The apparatus comprises a determining unit configured to determine whether the device under test meets the quality standard based on a result of the comparison. BRIEF DESCRIPTION OF THE DRAWINGS

[0004] Example Figure 1 A system for determining whether a device under test (DUT) meets quality standards according to an embodiment is shown.

[0005] Example Figure 2 A method implemented by a system for determining whether a device under test (DUT) meets quality standards according to an embodiment is shown.

[0006] Example Figure 3 A device under test (DUT) is shown connected to a system for determining whether the DUT meets quality standards according to an embodiment.

[0007] Example Figure 4 A circuit representation showing connections between a device under test and a system for determining whether the device under test meets quality criteria according to an embodiment.

[0008] Example Figure 5 is an example image of a computing chip as an example of a device under test (DUT) according to an embodiment.

[0009] Example Figure 6 A measurement unit of a system for determining whether a device under test (DUT) meets quality standards according to an embodiment is shown.

[0010] Example Figure 7 and Figure 8 Each shows an example measurement unit of a system for determining whether a device under test (DUT) meets quality standards using electro-optics sampling based time-domain reflectometry (EOS-TDR) according to an embodiment.

[0011] Example Fig. 9 Example types of soft defects including resistive defects and capacitive defects in a computing chip that may be a device under test (DUT) according to an embodiment are shown.

[0012] Example Fig. 10A and Fig. 10B The resistive defects ( Fig. 10A ) and capacitive defects ( Fig. 10B ) is based on an electro-optical sampling time domain reflectometry (EOS-TDR) response.

[0013] Example Fig.11 An example cross-correlation based detection method using Electro-Optical Sampling-based Time Domain Reflectometry (EOS-TDR) according to an embodiment is shown.

[0014] Example Fig. 12A , Fig. 12B , Fig. 12C and Fig.12D Example device under test waveforms are shown according to an embodiment.

[0015] Example Fig.13 A representation of a device under test waveform is shown along with a reference waveform in accordance with an embodiment.

[0016] Example Fig.14 A visual representation of a Hoteling's T2-test statistical analysis of a device under test waveform of a device under test failing to meet a quality criterion together with a reference waveform according to an embodiment is shown.

[0017] Example Fig.15 A computing device according to an embodiment is shown.

[0018] Example Fig.16 A cloud computing environment according to an embodiment is shown.

[0019] Example Fig.17 Shown are abstract model layers according to an embodiment. DETAILED DESCRIPTION

[0020] Embodiments relate to package-level soft defect detection methods, modeling soft defects in chip packages, and / or detecting soft defects by processing measured electro-optical sampling-based time domain reflectometry (EOS-TDR) waveforms through outlier detection schemes. Embodiments focus on the efficiency and versatility of detecting different soft defect types and locations of soft defects as well as hard defects. In embodiments, EOS-TDR enables rapid early detection of soft defects that cannot be detected using functional testing. Embodiments are particularly important in the automotive industry, but may also be applicable in a wide variety of other industries.

[0021] Example Figure 1 A system for determining whether a device under test (DUT) 4 meets a quality standard according to an embodiment is shown. The system may include an analysis unit 8 and a measurement unit 2.

[0022] The measurement unit 2 may be configured to measure the device under test 4 using at least one time-domain reflectometry measurement of the device under test 4 to generate a device under test waveform 37. The measurement unit 2 may contact the device under test 4 using a probe 6.

[0023] The analysis unit 8 may include a waveform comparison unit 10 configured to compare the device under test waveform 37 with at least one reference waveform. The at least one reference waveform may be stored in a reference waveform database 14. The at least one reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet the quality standard. The analysis unit 8 may include a quality determination unit 12 configured to determine whether the device under test 4 meets the quality standard based on the result of the comparison. The analysis unit 8 may include a waveform processing unit 16 configured to process the waveform measured by the measurement unit 2 or stored in the reference waveform database 14.

[0024] Example Figure 2 A method implemented by a system for determining whether a device under test (DUT) 4 meets a quality standard is shown according to an embodiment. In the method, step 3 measures the device under test 4 using at least one time domain reflectometry measurement of the device under test 4. Step 5 generates a device under test waveform. Step 9 compares the device under test waveform with at least one reference waveform. The reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet the quality standard. Step 11 determines whether the device under test 4 meets the quality standard based on the result of comparing step 9.

[0025] In an embodiment, each of the correlated reference waveforms is time frame windowed to a relevant time range before generating an average reference waveform from the plurality of correlated reference waveforms. Depending on the embodiment, the time frame windowing of the correlated reference waveforms may be performed before or after the reference waveforms are stored in the reference waveform database 14.

[0026] In an embodiment, in step 7, the DUT waveform is windowed to a relevant time range by time frame before comparing the DUT waveform 4 to the relevant reference waveform (eg, step 9).

[0027] In an embodiment, the relevant time range of the DUT waveform and / or the relevant reference waveform is selected to include at least one peak feature of at least one of the relevant reference waveforms.

[0028] Example Figure 3 A device under test (DUT) 4 according to an embodiment is shown, and the DUT 4 is connected to a system for determining whether the DUT 4 meets the quality standard. In an embodiment, the device under test 4 is as shown in the example Figure 3 In an embodiment, the quality criteria relate to the quality of the wiring 32 in the substrate 22 between the probe 6 of the measurement system and the die 20 of the computing chip 4. In an embodiment, determining whether the device under test 4 meets the quality criteria (e.g., step 11) determines whether there are soft defects in the computing chip 4 that cannot be detected by functional testing.

[0029] In an embodiment, the detector probe 6 can be connected to the solder ball 24 using the first lead 28, and the solder ball is electrically connected to the die 20 through the wiring 32 and the contact 21. In an embodiment, at least one of the solder ball 24, the wiring 32, or the contact 21 is being illustrated. Figure 3 In some embodiments, the detector probe 6 may include a second lead 30 connected to a solder ball 26, which is grounded to the computing chip 4 without being connected to any wiring.

[0030] Example Figure 4 A circuit representation showing connections between a device under test 4 and a system for determining whether the device under test 4 meets a quality criterion according to an embodiment.

[0031] Example Figure 4The circuit diagram shows selected signal pins in a flip-chip ball grid array (BGA) package chip sample under electro-optical sampling-based time domain reflectometry (EOS-TDR) measurement. Pulse signals are injected into the BGA solder balls through a radio frequency (RF) probe, and then they propagate along the traces in the substrate and reach the die. If the impedance of the die-substrate interface is not well matched, the pulse signal may be partially reflected. For example, some die pads incorporate electrostatic discharge (ESD) protection with typical decoupling capacitors in the range of 0.2 to 2 pF, and the negative pulse will be reflected and detected by the EOS-TDR system. In an embodiment, soft open defects in the package can occur anywhere along the trace, such as at solder balls, vias, bumps, or other components. For illustrative purposes, the example Figure 4 Two positions P1 and P2 of a selected signal pin are shown, representing the solder ball-substrate interface and the substrate-die interface, respectively.

[0032] In an embodiment, if the repeatability of the measurement system is well controlled, the soft defect detection sensitivity is determined by the sample-to-sample variation. The repeatability of a semi-automatic probe station can be ±3μm. In an embodiment, an EOS-TDR system can have a dynamic range higher than 90dB and a distance resolution less than 5μm. In an embodiment, the sample-to-sample variation is widely affected by the process and materials in the chip packaging. Table 1

[0033] Table 1 shows the Figure 4 Example parameters in terms of nominal value and tolerance of each circuit element in FIG. The BGA solder ball segment can be modeled by: Series inductance (L B ), the quality factor of the series inductor (Q B,ind ), parallel capacitance between solder balls (C B ) and the quality factor of the shunt capacitor (Q B,cap ). Assume that L B Proportional to the height of the solder ball, according to an embodiment, for a typical BGA solder ball with a 400-μm height and a ±50-μm tolerance, the potential inductance tolerance may be ±12.5%. The trace segment may be modeled by a lossy transmission line parameterized by: trace length (L T )、Characteristic impedance (Z T ), equivalent relative permittivity (ε r ) and quality factor (Q T ). In an embodiment, L T and Z T The tolerance of both is affected by the etching accuracy of the trace layout, and ZT The tolerance may also be affected by ε r According to an embodiment, the device section inside the die can be connected by an equivalent capacitor in capacitance (C L ) and quality factor (Q L,cap ) modeling. In embodiments, the quality factor of components inside the die may be lower than the quality factor of components outside the die. For example, in some embodiments, Q L,cap Comparable Q T 10 times smaller. According to an embodiment, a Monte-Carlo method may be used to compare defective cells with qualified cells.

[0034] Example Figure 5 is an example image of a computing chip as an example of a device under test (DUT) according to an embodiment. Figure 4 A cross-sectional view of the chip under test is shown. Each pad is directly connected to the die through a copper pillar. Both the upper and lower surfaces of the copper pillars may be susceptible to soft open defects.

[0035] Example Figure 6 A measurement unit 2 of a system for determining whether a device under test (DUT) 4 meets quality standards according to an embodiment is shown. According to an embodiment, the measurement unit 2 includes a signal generator 36 that can generate a pulse waveform 35. The pulse waveform 35 can be electrically transmitted to the detector probe 6. When the detector probe 6 is connected to the device under test 4, the pulse waveform 35 can be internally transmitted through the device under test 4 and reflected in the form of a reflection waveform 37. According to an embodiment, the characteristics and properties of the reflection waveform 37 will depend on the impedance characteristics of the device under test 4, which can indicate whether the device under test 4 meets the quality standards. The signal detector 38 can detect the reflection waveform 37 for transmission of the characteristics, properties or representation of the reflection waveform 37 to be sent to the analysis unit 8.

[0036] Example Figure 7 An example measurement unit of a system for determining whether a device under test (DUT) 4 meets quality standards using electro-optical sampling based time domain reflectometry (EOS-TDR) according to an embodiment is shown. In the EOS-TDR, according to an embodiment, an optical reference signal source 40 may generate an optical reference signal with a relatively high resolution. In an embodiment, an optical detector and an electrical signal generator 42 may receive the optical reference signal and generate a relatively high resolution electrical signal 35 to be transmitted to a detector probe 6. The electrical signal 35 may be internally reflected at the device under test 4, thereby generating a reflection signal 37. The reflection waveform 37 may be detected by an electrical signal detector 44.

[0037] In an embodiment, the at least one time domain reflectometry measurement 37 of the device under test 4 is at least one electro-optical sampling based time domain reflectometry measurement 37 of the device under test 4. In an embodiment, the at least one time domain reflectometry measurement 37 of at least one device known to meet the quality standard is at least one electro-optical sampling based time domain reflectometry measurement 37 of at least one device known to meet the quality standard.

[0038] Example Figure 8 An example measurement unit of a system for determining whether a device under test (DUT) 4 meets quality standards using electro-optical sampling based time domain reflectometry (EOS-TDR) according to an embodiment is shown.

[0039] Electro-optical sampling-based time domain reflectometry (EOS-TDR) can quickly detect soft defects in each pin from a trace-structure perspective, regardless of the operation of the transistors inside the die. Figure 8 As shown in , similar to an oscilloscope-based time domain reflectometry system, an electro-optical sampling-based time domain reflectometry (EOS-TDR) system injects an electrical pulse waveform into a device under test and monitors the reflection. In embodiments, the EOS-TDR system can provide 20-30 times better distance resolution than other oscilloscope-based time domain reflectometry systems, which can be important when detecting small waveform deviations caused by soft defects along a trace. In embodiments, the narrow pulse waves generated by the EOS-TDR system can better discern defect locations than the step waves from an oscilloscope-based time domain reflectometry system.

[0040] For example, using an oscilloscope-based time domain reflectometry system, soft defects can only be detected by functional testing under extreme conditions, while using an EOS-TDR system, soft defects can be detected under normal conditions (e.g., ambient temperature, no excessive stress on the pad, power failure, etc.). For example, if a functional test is used, a local open circuit can only be detected at 70°C, but if measured by EOS-TDR, a waveform deviation compared to a device known to meet quality standards can be detected at room temperature. Similarly, a through-hole crack occurring at 135°C can be detected by a waveform deviation detected at room temperature before rupture. In an embodiment, the waveform deviation is caused by sample-to-sample variations between defects in a defective device and / or a good unit (GoodUnit) (e.g., a device known to meet quality standards). In an embodiment, in order to detect soft defects among many devices by EOS-TDR waveforms, an understanding of waveform deviations from different soft defect situations may be necessary. In an embodiment, statistical analysis may be required to identify outliers.

[0041] Embodiments implement soft defect detection through analysis of EOS-TDR waveforms. In embodiments, a circuit model of typical soft defects may be generated after analysis of the resulting EOS-TDR response. In embodiments, a cross-correlation based outlier waveform detection method may be implemented to detect general soft defects.

[0042] Example Fig. 9 Example types of soft defects including resistive defects and capacitive defects in a computing chip that may be a device under test (DUT) according to an embodiment are shown. Fig. 9 Examples of typical soft defects in chip packages including resistive defects and capacitive defects are shown. Both resistive defects and capacitive defects can increase or decrease load impedance. If there is a resistive open circuit or a capacitive open circuit defect, the load impedance may be high. If there is a resistive short circuit or a capacitive short circuit defect, the load impedance may be low.

[0043] Example Fig. 10A and Fig. 10B The resistive defects ( Fig. 10A ) and capacitive defects ( Fig. 10B ) is based on an electro-optical sampling time domain reflectometry (EOS-TDR) response.

[0044] Example Fig. 10A The simulated EOS-TDR response to a resistive defect is shown. The reflected peak amplitude increases with the equivalent defect resistance (R Open or R Short ) increases. In an embodiment, the reflection from a resistive open circuit is positive, while the reflection from a resistive short circuit is negative. Because the peak reflection time of a resistive defect does not depend on R Open and R Short , so detection can be achieved based on the amplitude change of the TDR waveform.

[0045] Example Fig. 10B 1 shows the simulated EOS-TDR response to a capacitive defect. Similar to a resistive defect, the reflection from a capacitive open circuit is positive, while the reflection from a capacitive short circuit is negative. In an embodiment, the reflection peak amplitude varies with the equivalent defect capacitance (C Open or C Short ) is reduced. In an embodiment, due to the energy storage characteristics of the capacitor, the time delay of the reflection peak increases with C Open and C Short In an embodiment, both amplitude and time changes may be considered to detect capacitive defects.

[0046] Example Fig.11 An example cross-correlation based detection method using Electro-Optical Sampling-based Time Domain Reflectometry (EOS-TDR) according to an embodiment is shown.

[0047] In an embodiment, there are multiple devices that are known to meet the quality criteria. Each of the multiple devices that are known to meet the quality criteria has an associated reference waveform. In an embodiment, a method generates an average reference waveform by averaging all associated reference waveforms of all multiple devices that are known to meet the quality criteria.

[0048] In an embodiment, a cross correlation of the averaged reference waveform with each of the correlated reference waveforms is performed to generate at least one cross correlation of the correlated reference waveforms.In an embodiment, a cross correlation of the averaged reference waveform with the device under test waveform is performed to generate a cross correlation of the device under test waveform.

[0049] In an embodiment, extracting at least one peak feature from at least one cross-correlation of a related reference waveform is performed. In an embodiment, extracting at least one peak feature from a cross-correlation of a device under test waveform is performed.

[0050] In an embodiment, at least one peak feature from at least one cross-correlation of the correlated reference waveform comprises at least one of an amplitude peak feature of the correlated reference waveform or at least one hysteresis peak feature of the correlated reference waveform. In an embodiment, at least one peak feature from the cross-correlation of the device under test waveform comprises at least one of an amplitude peak feature of the device under test waveform or at least one hysteresis peak feature of the device under test waveform.

[0051] In an embodiment, a method includes comparing at least one peak feature extracted from a cross-correlation of a device under test waveform with at least one peak feature extracted from a cross-correlation of a correlated reference waveform. In an embodiment, a significant deviation of a delayed peak feature of the device under test waveform from a delayed peak feature of the correlated reference waveform indicates a capacitive defect in the device under test.

[0052] In an embodiment, a method includes comparing at least one peak feature extracted from a cross-correlation of a device under test waveform with at least one peak feature extracted from a cross-correlation of a related reference waveform. In an embodiment, a significant deviation of the amplitude peak feature of the device under test waveform from the amplitude peak feature of the related reference waveform indicates a resistive defect in the device under test.

[0053] In an embodiment, a method includes comparing at least one peak feature extracted from a cross-correlation of a device under test waveform with at least one peak feature extracted from a cross-correlation of a related reference waveform.

[0054] In an embodiment, a method includes determining that a device under test does not meet a quality criterion if a significant deviation of at least one peak feature extracted from a cross-correlation of a device under test waveform from at least one peak feature extracted from a cross-correlation of a related reference waveform is determined to exist.

[0055] In an embodiment, the presence of a significant deviation may be determined based on a statistical analysis of at least one peak feature extracted from a cross-correlation of the device under test waveform and at least one peak feature extracted from a cross-correlation of a related reference waveform. In an embodiment, the statistical analysis comprises a Hotelling t-squared statistical analysis. In an embodiment, the statistical analysis is performed by at least one of artificial intelligence, machine learning, or a statistical algorithm.

[0056] In an embodiment, the EOS-TDR waveform deviation may be attributed to impedance-dependent and / or position-dependent soft defects. For example, if the pulse amplitude of the pulse reflection is impedance-dependent, the pulse timing may be position-dependent. Example Fig.11 An embodiment of cross-correlation based soft defect detection is shown, which includes time window identification, feature extraction and / or Hotelling T 2 According to an embodiment, the features extracted by cross-correlation are relatively sensitive to amplitude changes and temporal changes.

[0057] In an embodiment, the EOS-TDR waveform may include a defect-dependent reflection (DDR) and / or a defect-independent reflection (DIR). In an embodiment, in order to detect defects with known locations, a time window of interest may be identified with the aid of the DDR, which may be obtained by open-short normalization and / or comparison of the waveform of a defective cell with the waveform of a non-defective cell (e.g., a good cell). However, in an embodiment, if the defect location is unknown, the time window may need to cover the reflection time of the farthest device. In an embodiment, the same time window will be applied to the EOS-TDR waveforms of both the non-defective cell and the defective cell for further calculations.

[0058] In an embodiment, after windowing, the average non-defective cell EOS-TDR waveform is and DUT waveform (d i ) becomes a vector of equal length, where i∈[w1,w2], and w1 and w2 are the left and right boundaries of the time window. The waveform deviation caused by the defect can be measured by cross-correlation analysis and d i The similarity of is detected, thereby generating two characteristic correlation factors (X) and hysteresis (L). In an embodiment, X and L may correspond to the impedance change and time shift change caused by the defect, respectively.

[0059] In an embodiment, [X D ,L D ]and and respectively represent the normalized feature vectors extracted from the device under test and the qualified unit (e.g., a device known to meet the quality standards or a defect-free unit, etc.). Hotelling T can be performed by the following operation 2 test: T 2 =[X D ,L D ] T S -1 [X D ,L D ] Where S is The covariance matrix of .

[0060] In an embodiment, when Following the chi-square distribution, the soft defect detection area can be defined as: T 2 >X 2 (α,k) where (1-α) is the confidence interval and k is the degrees of freedom (eg, k is equal to 2 here).

[0061] Example Fig. 12A , Fig. 12B , Fig. 12C and Fig.12D Example device under test waveforms are shown according to an embodiment.

[0062] Fig. 9 The example resistive short defect shown in Figure 4 At the P1 position, with a 1kΩ R Short The TDR waveform of a failed cell can be simulated by generating another 100 samples (for example) with the remaining parameters shown in Table 1 within the tolerance. Fig. 12A For example purposes, simulated TDR waveforms of 100 good cells and 100 bad cells are shown. Fig. 12A In the example in FIG. 1 , the difference between the waveform of the qualified cell and the waveform of the unqualified cell can be relatively clearly observed. In order to solve such problems in the detection of marginal unqualified cells, the embodiment is most applicable.

[0063] In an embodiment, a time window may be identified for the correlation time window. According to an embodiment, because the defect location may be unknown, the correlation time window should cover the complete reflection from the detection location to the device point. As an example, based on the parameters shown in Table 1, the actual device point may be approximately 50 ps. However, in some examples, the reflection from the device point may be attributed to C L As an example, the time window may be selected to be 0 to 130 ps, ​​within which most of the significant reflections are contained.

[0064] In an embodiment, the TDR waveform features may be extracted by a cross-correlation method. Fig. 12A The extracted features of all waveforms are shown. The failed cells may be distinguishable from the qualified cells. In an embodiment, the Hotelling T 2 The test is applied to the extracted features of the qualified cells to obtain confidence interval (CI) circles. The CI circles serve as a good measure to detect soft defects because the extracted features of the failed cells are likely to appear anywhere outside the CI circles, experiencing the TDR waveform of the qualified cells and a type of soft defect.

[0065] As an example, in an embodiment, a resistive open defect may be placed in an example Figure 4 At the P2 position, with a 1Ω R Open The TDR waveform of the failed unit can be generated by generating another 100 samples (for example) within the tolerance with the remaining parameters shown in Table 1. Example Fig. 12B The generated TDR waveform of a failed cell is shown along with the waveform of a passed cell. As shown, the waveform of the failed cell almost overlaps with the waveform of the passed cell. As an example, a time window of 0 to 130 ps can be applied for outlier detection. Example Fig. 12B Extracted features of the waveform of a failed cell compared to the waveform of a qualified cell and CI circles are shown according to an embodiment. In an embodiment, a 1-Ω resistive open may be detected even if the 1-Ω resistive open is located close to the device in the die. In an embodiment, it may be challenging to detect marginal resistive opens in a chip package simply by functional testing because the current allowed to flow through the device is extremely small.

[0066] As an example, in an embodiment, a capacitive short circuit defect may be placed in an example Figure 4 At the P1 position, there is a 60fF C Short According to an embodiment, the TDR waveform of the failed cell can be generated by generating another 100 samples (for example) within the tolerance with the remaining parameters shown in Table 1. Example Fig. 12C , according to an embodiment. For example, the 60-fF short-circuit capacitance in the failed cell can be reduced (Example Figure 4 The impedance at P1 of the cell is changed to cause a reflection amplitude of about 0ps, which is lower than the reflection amplitude of the qualified cell. For example, such waveform differences can continue to be carried in the presence of multiple reflections in the signal trace, resulting in higher reflection amplitudes close to 25ps and 100ps. As an example, these time slots exposed to waveform changes are covered by a selected time window of 0-130ps. Example Fig. 12C The extracted features of the waveform of a failed cell compared to the waveform of a passed cell and the CI circles are shown, wherein all failed cells are distinguished, according to an embodiment.

[0067] As an example, in an embodiment, a resistive open defect may be placed in (Example Figure 4 At the P2 position, there is a 1pF C Open In this example, if another capacitor is stacked on top of C L Above, C L It is still possible to charge above the threshold in this way, which may cause the functional test to ignore this failure. The TDR waveform of the failed cell can be generated by generating 100 samples (for example) within the tolerance with the remaining parameters shown in Table 1. Example Fig.12D 1 shows the resulting waveform according to an embodiment. According to an embodiment, within 40 to 90 ps, ​​as a result of the reduction in total load capacitance, the waveform of the failed cell may be higher than the waveform of the qualified cell (C Open and C L As an example, a time window of 0 to 130 ps can be applied to outlier detection. Fig.12D Extracted features of a waveform of a failed cell compared to a waveform of a passed cell are shown along with CI circles, where the failed cells are distinguished, according to an embodiment.

[0068] Example Fig.13 A representation of a device under test waveform along with a reference waveform is shown according to an embodiment. Fig.13 Measured EOS-TDR waveforms of good and bad cells (fs) according to an embodiment are shown. In this example, there are 22 good cells and one bad cell.

[0069] Example Fig.14 A visual representation of a Hotelling's T2 test statistical analysis showing a device under test waveform for a device under test that failed to meet quality criteria along with a reference waveform in accordance with an embodiment. Fig.13 The Hotelling T for the measured EOS-TDR waveform according to an embodiment is shown. 2 Testing. In this example, according to an embodiment, a non-conforming unit can be detected with a confidence interval of 99.9995%.

[0070] Example Fig.15 A computing device 1200 according to an embodiment is shown. The computing device 1200 is only one example of a computing device and is not intended to imply any limitation on the scope of use or functionality of the embodiments described herein. In any case, the computing device 1200 is capable of implementing and / or performing any of the functionalities stated above. The term processing node is a logical concept. Any number of central processing units with any number of cores or machines may be in one computing device or distributed among multiple computing devices.

[0071] In computing device 1200, there is a computer system / server 1202, which can operate in numerous other general or special purpose computing system environments or configurations. Examples of well-known computing systems, environments, and / or configurations that may be suitable for use with computer system / server 1202 include, but are not limited to, personal computer systems, server computer systems, thin clients, fat clients, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputer systems, mainframe computer systems, and distributed cloud computing environments including any of the above systems or devices, etc.

[0072] Computer system / server 1202 may be described in the general context of computer system executable instructions, such as program modules, executed by a computer system. Generally, program modules may include routines, programs, objects, components, logic, data structures, etc. that perform specific tasks or implement specific abstract data types. Computer system / server 1202 may be practiced in a distributed cloud computing environment where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computer system storage media, including memory storage devices.

[0073] Embodiments are directed to a non-transitory storage medium having stored thereon a program for causing at least one processor to perform a method. In embodiments, the method includes measuring a device under test using at least one time domain reflectometry measurement of the device under test to generate a device under test waveform. In embodiments, the method includes comparing the device under test waveform to at least one reference waveform, wherein the at least one reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet a quality standard. In embodiments, the method includes determining whether the device under test meets the quality standard based on a result of the comparison.

[0074] like Fig.15 As shown, a computer system / server 1202 in computing device 1200 is shown in the form of a general-purpose computing device. Components of computer system / server 1202 may include, but are not limited to, one or more processors or processing units 1204, a system memory 1206, and a bus 1208 that couples various system components including system memory 1206 to processor 1204.

[0075] Bus 1208 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example and not limitation, such architectures include an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, an Enhanced ISA (EISA) bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0076] Computer system / server 1202 typically includes a variety of computer system readable media. Such media can be any available media that can be accessed by computer system / server 1002, and it includes both volatile and nonvolatile media and both removable and non-removable media.

[0077] The system memory 1206 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 1210 and / or cache memory 1212. The computer system / server 1202 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, a storage system 1214 may be provided for reading and writing to a non-removable non-volatile magnetic medium (not shown and generally referred to as a "hard disk"). Although not shown, a magnetic disk drive for reading and writing to a removable non-volatile magnetic disk (such as a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk such as a CD-ROM, DVD-ROM or other optical media may be provided. In such a case, each may be connected to the bus 1208 via one or more data media interfaces. As will be further depicted and described below, the memory 1206 may include at least one program product having a set (e.g., at least one) of program modules configured to implement the functions of various embodiments of the present invention.

[0078] A program / utility 1216 having a set (at least one) of program modules 1218 (by way of example and not limitation) and an operating system, one or more application programs, other program modules, and program data may be stored in memory 1206. Each or some combination of the operating system, one or more application programs, other program modules, and program data may include an implementation of a networking environment. Program modules 1218 generally implement the functions and / or methods of various embodiments of the present invention as described herein.

[0079] The computer system / server 1202 may also communicate with one or more external devices 1020, such as keyboards, pointing devices, displays 1222, one or more devices that enable a user to interact with the computer system / server 1202, and / or any device that enables the computer system / server 1202 to communicate with one or more other computing devices (e.g., a network card, a modem, etc.). Such communications may occur via an I / O interface 1224. Furthermore, the computer system / server 1202 may communicate with one or more networks, such as a local area network (LAN), a general wide area network (WAN), and / or a public network (e.g., the Internet) via a network adapter 1226. As depicted, the network adapter 1226 communicates with other components of the computer system / server 1202 via a bus 1208. It should be understood that, although not shown, other hardware and / or software components may be used in conjunction with the computer system / server 1202. Examples include, but are not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, among others.

[0080] Example Fig.16 A cloud computing environment 1302 according to an embodiment is shown. As shown, the cloud computing environment 1302 includes one or more cloud computing nodes 1200, with which local computing devices such as personal digital assistants (PDAs) or cell phones 1304, desktop computers 1306, laptop computers 1308, and / or automotive computer systems 1310 used by cloud consumers can communicate. The nodes 1200 can communicate with each other. They can be physically or virtually grouped in one or more networks such as the private, community, public, or hybrid clouds described above, or combinations thereof (not shown). This allows the cloud computing environment 1302 to provide cloud consumers with infrastructure, platforms, and / or software as a service for which they do not need to maintain resources on local computing devices. It should be appreciated that Fig.16 The types of computing devices 1304, 1306, 1308, 1310 shown in are illustrative only, and computing node 1200 and cloud computing environment 1302 can communicate with any type of computerized device via any type of network and / or network addressable connection (e.g., using a web browser).

[0081] Example Fig.17 The cloud computing environment 1302 ( Fig.16 ) provides a set of functional abstraction layers. It should be understood in advance that Fig.17 The components, layers, and functions shown in are merely illustrative, and the embodiments are not limited thereto. As depicted, the following layers and corresponding functions are provided:

[0082] The hardware and software layer 1402 includes hardware and software components. Examples of hardware components include mainframes; servers based on RISC (Reduced Instruction Set Computer) architecture; storage devices; network and networking components. Examples of software components include network application server software and database software.

[0083] Virtualization layer 1404 provides an abstraction layer from which the following examples of virtual entities can be provided: virtual servers; virtual storage devices, virtual networks, including virtual private networks; virtual applications and operating systems; and virtual clients. In one example, management layer 1406 can provide the functionality of processing unit 68. Workload layer 1408 provides examples of functionality for which a cloud computing environment can be utilized.

[0084] Aspects of the present invention have been discussed above with reference to flowchart illustrations and / or block diagrams of methods, devices (systems) and computer program products according to various embodiments of the present invention. It should be understood that each box in the flowchart illustration and / or block diagram and the combination of boxes in the flowchart illustration and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine, so that instructions executed by the processor of the computer or other programmable data processing device create components for implementing the functions / actions specified in one or more boxes of the flowchart and / or block diagram.

[0085] These computer program instructions may also be stored in a computer-readable medium, which may direct a computer, other programmable data processing equipment, or other devices to operate in a specific manner, such that the instructions stored in the computer-readable medium for measuring the device under test (DUT) and determining whether the DUT meets quality standards include instructions for implementing the functions / actions specified in one or more boxes of the flowchart and / or block diagram.

[0086] The computer program instructions may also be loaded onto a computer, other programmable data processing device or other apparatus to cause a series of operational steps to be performed on the computer, other programmable device or other apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide a process for implementing the functions / actions specified in one or more boxes of the flowchart and / or block diagram.

[0087] It will be obvious and clear to those skilled in the art that various modifications and variations can be made in the disclosed embodiments. Therefore, it is intended that the disclosed embodiments cover obvious and clear modifications and variations as long as they are within the scope of the appended claims and their equivalents.

Claims

1. A method, characterized in that include: measuring the device under test using at least one time domain reflectometry measurement of the device under test to generate a device under test waveform; comparing the device under test waveform to at least one reference waveform, wherein the at least one reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet quality standards; and A determination is made based on a result of the comparison whether the device under test meets a quality standard.

2. The method according to claim 1, characterized in that: The at least one time domain reflectometry measurement of the device under test is at least one electro-optical sampling based time domain reflectometry measurement of the device under test; and The at least one time domain reflectometry measurement of the at least one device known to meet quality criteria is at least one electro-optical sampling-based time domain reflectometry measurement of the at least one device known to meet quality criteria.

3. The method according to claim 1, characterized in that: The at least one device known to meet the quality standard is a plurality of devices known to meet the quality standard, each of the plurality of devices known to meet the quality standard having an associated reference waveform of the at least one reference waveform; and The method includes generating an average reference waveform by averaging all of the associated reference waveforms for all of the plurality of devices known to meet a quality criterion.

4. The method according to claim 3, characterized in that The at least one reference waveform and the average reference waveform are stored in a reference waveform database.

5. The method according to claim 3, characterized in that: Prior to generating the average reference waveform, each of the related reference waveforms is time frame windowed to a related time range; and Prior to comparing the DUT waveform to the associated reference waveform, the DUT waveform is windowed by time frame to the associated time range.

6. The method according to claim 5, characterized in that The correlated time range of the device under test waveform and the correlated reference waveform is selected to include at least one peak feature of at least one of the correlated reference waveforms.

7. The method according to any one of claims 3 to 6, characterized in that The comparison includes: cross-correlating the averaged reference waveform with each of the correlated reference waveforms to generate at least one cross-correlation of the correlated reference waveforms; and The averaged reference waveform is cross-correlated with the device under test waveform to generate a cross-correlation of the device under test waveform.

8. The method according to claim 7, characterized in that The comparison includes: extracting at least one peak feature from the at least one cross-correlation of the correlated reference waveform; and At least one peak feature is extracted from the cross-correlation of the device under test waveform.

9. The method according to claim 8, characterized in that: the at least one peak feature from the at least one cross-correlation of the correlated reference waveform comprising at least one of an amplitude peak feature of the correlated reference waveform or at least one lag peak feature of the correlated reference waveform; The at least one peak characteristic from the cross-correlation of the device under test waveform includes at least one of an amplitude peak characteristic of the device under test waveform or a hysteresis peak characteristic of the device under test waveform.

10. The method according to claim 9, characterized in that: The comparing comprises comparing the at least one peak feature extracted from the cross-correlation of the device under test waveform with the at least one peak feature extracted from the cross-correlation of the correlated reference waveform; and A significant deviation of the delayed peak characteristic of the device under test waveform from the delayed peak characteristic of the associated reference waveform indicates a capacitive defect in the device under test.

11. The method according to claim 9, characterized in that: The comparing comprises comparing the at least one peak feature extracted from the cross-correlation of the device under test waveform with the at least one peak feature extracted from the cross-correlation of the correlated reference waveform; and A significant deviation of the amplitude peak characteristic of the device under test waveform from the amplitude peak characteristic of the associated reference waveform indicates a resistive defect in the device under test.

12. The method according to claim 8, characterized in that The comparing includes comparing the at least one peak feature extracted from the cross-correlation of the device under test waveform to the at least one peak feature extracted from the cross-correlation of the correlated reference waveform.

13. The method according to claim 12, characterized in that The determining includes determining that the device under test does not meet a quality criterion if it is determined that there is a significant deviation of the at least one peak feature extracted from the cross-correlation of the device under test waveform from the at least one peak feature extracted from the cross-correlation of the correlated reference waveform.

14. The method according to claim 13, characterized in that The presence of the significant deviation is determined based on a statistical analysis of the at least one peak feature extracted from the cross-correlation of the device under test waveform and the at least one peak feature extracted from the cross-correlation of the correlated reference waveform.

15. The method according to claim 14, characterized in that The statistical analysis included Hotelling's t-squared statistical analysis.

16. The method according to claim 14, characterized in that The statistical analysis is performed by at least one of artificial intelligence, machine learning, or a statistical algorithm.

17. The method according to claim 1, characterized in that: The device under test is a computing chip; The quality criteria relate to the quality of the wiring in a substrate between solder balls and a die of the computing chip; and Determining whether the device under test meets quality standards may determine whether soft defects are present in the computing chip that cannot be detected by functional testing.

18. The method according to claim 1, characterized in that The quality criterion relates to at least one of a soft defect and a hard defect.

19. A device, characterized in that It includes: a measurement unit configured to measure a device under test using at least one time domain reflectometry measurement of the device under test to generate a device under test waveform; a comparison unit configured to compare the device under test waveform to at least one reference waveform, wherein the at least one reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet quality standards; as well as A determination unit is configured to determine whether the device under test meets a quality standard based on a result of the comparison.

20. A non-transitory storage medium, characterized in that: A program is stored thereon for causing at least one processor to perform the following operations: measuring the device under test using at least one time domain reflectometry measurement of the device under test to generate a device under test waveform; comparing the device under test waveform to at least one reference waveform, wherein the at least one reference waveform represents at least one time domain reflectometry measurement of at least one device known to meet quality standards; and A determination is made based on a result of the comparison whether the device under test meets a quality standard.

Images