Predictive health management for power equipment

Abstract

Embodiments of the present disclosure relate to predictive health management for power equipment. In some examples, an apparatus includes a power structure and a sensing structure electrically isolated from the power structure. The apparatus also includes a processing circuitry configured to determine whether the sensing structure includes a predictive health indicator, where the predictive health indicator is indicative of a health condition of the power structure.

Description

Technical Field

[0001] This invention relates to circuit arrangements for semiconductor devices. Background Technology

[0002] Power devices and semiconductor devices can be exposed to severe stress conditions. This stress can lead to gradual aging and degradation, ultimately causing device failure. The magnitude of stress applied to switches in the field varies depending on the circumstances. Therefore, the actual service life of power devices can be unpredictable.

[0003] Power devices can be designed to withstand the worst-case task distribution throughout the lifespan of the system (e.g., a vehicle) in which they will be installed. Only a small fraction of the device will be exposed to the worst-case task distribution. Therefore, for almost all applications, power devices are over-engineered relative to their actual task distribution. Summary of the Invention

[0004] This disclosure describes techniques for an apparatus including a power structure and a sensing structure electrically isolated from the power structure. The apparatus also includes processing circuitry configured to determine whether the sensing structure includes a predictive health indicator, wherein the predictive health indicator indicates the health of the power structure. Prior to evidence of any health problem in the power structure, the processing circuitry is able to predict a health problem in the power structure based on the detection of the predictive health indicator.

[0005] In some examples, one method includes measuring current in a sensing structure electrically isolated from a power structure. The method also includes determining that the current in the sensing structure is greater than a threshold level. Furthermore, the method includes setting a bit indicating a predicted health problem in response to determining that the current in the sensing structure is greater than the threshold level.

[0006] In some examples, a system includes a power substrate comprising a power structure and a sensing structure, wherein the sensing structure is electrically isolated from the power structure. The system also includes processing circuitry configured to control a driver to generate a driver signal controlling the power structure. The processing circuitry is further configured to measure a current in the sensing structure, which is electrically isolated from the power structure. The processing circuitry is also configured to determine whether the current in the sensing structure is greater than the threshold level. The processing circuitry is further configured to set a bit indicating a predicted health problem in response to determining that the current in the sensing structure is greater than the threshold level.

[0007] Details of one or more examples are set forth in the accompanying drawings and the description below. Other features, objectives, and advantages will be apparent from the specification, the drawings, and the claims. Attached Figure Description

[0008] Figure 1This is a conceptual block diagram of a device including a power structure and a sensing structure according to the present disclosure.

[0009] Figure 2 This is a conceptual block diagram of a sensing structure including one or more antifuse wires and a current sensor according to the present disclosure.

[0010] Figure 3 This is a conceptual block diagram of a sensing structure including one or more antifuse wires and a voltage comparator according to the present disclosure.

[0011] Figure 4 This is a conceptual block diagram of a sensing structure and a power structure including a power switch, based on the technology disclosed herein.

[0012] Figure 5 This is a conceptual block diagram of a system including logic devices and power devices according to the technology disclosed herein.

[0013] Figure 6 This is a flowchart illustrating an example technique for detecting a predicted health indicator in a power structure according to the present disclosure.

[0014] Figure 7 This is a flowchart illustrating an example technique for designing semiconductor devices according to the present disclosure. Detailed Implementation

[0015] This disclosure describes apparatus, methods, and techniques for predicting the health of a power structure using a sensing structure. Instead of directly sensing the health or state of the power structure, the sensing structure may include a predictive health indicator that provides an indication of the health of the power structure. By implementing the sensing structure into a chip that also includes the power structure, or into a separate chip close to the power structure, the apparatus can provide health monitoring capabilities for predictive health management.

[0016] For example, a sensing structure that directly senses the health of a power structure might not sense degradation until it becomes detectable (e.g., apparent), which could be too late to alert the user, mitigate degradation, and / or prevent power structure failure. Instead, the sensing structure of this disclosure may include an aging mechanism that can provide a predictive health indication of the power structure by becoming apparent before degradation becomes noticeable. In this way, processing circuitry can be configured to predict power structure degradation based on the electrical parameters of the sensing structure before degradation becomes apparent.

[0017] Fatigue affects materials in power structures and can ultimately lead to aging and failure. Fatigue can also cause sensing structures to exhibit an electrical signature that indicates a dangerous aging state of the power structure. Sensing structures can be configured to provide an electrical signature when the device deteriorates to a state that could lead to health problems in the power structure. The electrical signature exhibited by the sensing structure can include high or low current, high or low voltage, high or low impedance, resistance changes, impedance changes, capacitance changes, evolving short circuits, and / or evolving open circuits in a dedicated sensing structure. One example signature is breakdown in the oxide of an antifuse, which ultimately leads to a short circuit. Another example signature is the breakdown of a current path in a fusible structure, which ultimately leads to an open circuit. The devices disclosed herein may include processing circuitry capable of providing a warning signal in response to detecting an electrical signature before it becomes apparent in the power structure due to overstress or wear.

[0018] The sensing structure may include dedicated, non-critical sensors that are subject to thermomechanical fatigue. However, the sensing structure does not necessarily have the same failure mechanism as the power structure. Fatigue experienced by the sensing circuitry can generate a characteristic electronic signature on the dedicated circuitry, which can be read by sensors within that dedicated circuitry. In some instances, the time between the characteristic electronic signature and the occurrence of a health problem can be tunable relative to the aging kinetics of the power structure (e.g., through calibration).

[0019] Figure 1 This is a conceptual block diagram of a device 100 including a power structure 150 and a sensing structure 110, based on the technology disclosed herein. Figure 1 In some examples, device 100 also includes processing circuitry 160 and an optional diagnostic interface 170, although in some examples, processing circuitry 160 may be located partially or entirely outside of device 100.

[0020] Power structure 150 may include a power switch for regulating the delivery of electrical power to or from a load. Power structure 150 may include a load path for supplying or absorbing electrical power from a load. Power structure 150 may also include an electrical conductor electrically connected to the load path to allow power structure 150 to absorb or supply electrical power. Power structure 150 may be part of power electronics and / or power conversion circuitry.

[0021] Power structure 150 may include, but is not limited to, any type of field-effect transistor (FET) (such as diffused metal-oxide-semiconductor FET (MOSFET)), bipolar junction transistor (BJT), insulated-gate bipolar transistor (IGBT), high electron mobility transistor (HEMT), and / or another element that uses voltage for its control. Power structure 150 may include n-type transistors and / or p-type transistors. Power structure 150 may include semiconductor materials such as silicon, silicon carbide, and / or gallium nitride.

[0022] Processing circuitry 160 is communicatively coupled to sensing structure 110. In some examples, processing circuitry 160 is also communicatively coupled to external device 180 via diagnostic interface 170. Processing circuitry 160 may include any suitable arrangement of hardware, software, firmware, or any combination thereof to perform the techniques accorded to processing circuitry 160 herein. Examples of processing circuitry 160 include one or more of a microprocessor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or any other equivalent integrated or discrete logic circuitry device, and any combination of these components. When processing circuitry 160 includes software or firmware, it also includes any hardware for storing and executing the software or firmware, such as one or more processors or processing units.

[0023] Generally, a processing unit may include one or more microprocessors, DSPs, ASICs, FPGAs, or any other equivalent integrated or discrete logic circuit devices, and any combination of such components. Although not explicitly stated... Figure 1 As shown, however, processing circuitry 160 may include or be coupled to a memory configured to store data. The memory may include any volatile or non-volatile medium, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), electrically erasable programmable ROM (EEPROM), flash memory, etc. In some examples, the memory may be external to processing circuitry 160 (e.g., outside the package housing processing circuitry 160).

[0024] Diagnostic interface 170 is an optional component of device 100 and is configured to communicatively couple to external device 180. External device 180 may be configured to communicate with processing circuitry 160 via diagnostic interface 170. For example, processing circuitry 160 may be configured to send data associated with sensing structure 110 to external device 180. In some examples, processing circuitry 160 may receive command signals (e.g., one or more bits) from external device 180 and may be configured to measure sensing structure 110 in response to receiving a command signal. Additionally or alternatively, processing circuitry 160 may be configured to perform measurements at predetermined time intervals during manufacturing and / or during operation of device 100.

[0025] In some examples, device 100 may include onboard memory, such as one-time programmable (OTP) memory, for storing thresholds used during testing of sensing structure 110. Processing circuitry 160 may be configured to store one or more thresholds into the onboard memory based on communications received from external device 180 via diagnostic interface 170. Processing circuitry 160 may be configured to perform tests on sensing structure 110 and send the test results to external device 180. For example, the tests may occur before device 100 is shipped to a customer. External device 180 may then analyze the results, determine new thresholds, and transmit the new threshold level to processing circuitry 160 for storage in onboard memory. Processing circuitry 160 may then be configured to perform another test using the new threshold level.

[0026] In some examples, processing circuitry 160 and / or external device 180 may be configured to generate an alarm in response to determining that sensing structure 110 includes a predictive health indicator, to notify the user of future health problems of power structure 150. Additionally or alternatively, processing circuitry 160 and / or external device 180 may be configured to shut down and / or operate power structure 150 at lower power levels in response to determining that sensing structure 110 includes a predictive health indicator.

[0027] Power structure 150 can be configured to operate for several years or decades. During the operating life of power structure 150, it can undergo thousands or millions of power cycles. For example, during each power cycle, the power switch of power structure 150 can be turned on and off, causing temperature changes and inrush currents in power structure 150. Power structure 150 will deteriorate over time due to fatigue caused by a large number of power cycles.

[0028] According to the technology disclosed herein, the sensing structure 110 is configured to include a predictive health indicator indicating the health status of the power structure 150. To measure the health status of the power structure 150, the processing circuitry device 160 may be configured to determine whether the sensing structure 110 includes a predictive health indicator. In some examples, the predictive health indicator is a voltage or current exceeding a threshold level in the sensing structure 110. Predictive health indicators can occur due to aging mechanisms of the sensing structure 110. An example of an aging mechanism is the decomposition of oxides in the sensing structure 110.

[0029] Sensing structure 110 may have an aging mechanism different from that of power structure 150. Although the aging mechanisms may differ, they can be designed to be time-dependent, such that the aging mechanism of sensing structure 110 (e.g., a predicted health indicator) becomes apparent before the aging mechanism of power structure 150 becomes apparent. Therefore, by detecting that sensing structure 110 includes a predicted health indicator, processing circuitry 160 can predict health problems in power structure 150 before these problems affect its performance. Sensing structure 110 may include a predicted health indicator when the aging mechanism has developed to the point where the electrical parameters of sensing structure 110 intersect with a threshold level. Sensing structure 110 can be designed such that its aging mechanism becomes apparent before the aging mechanism of power structure 150 becomes detectable.

[0030] Unlike other sensors electrically connected to the power structure, sensing structure 110 is electrically isolated from power structure 150. In other words, each of structures 110 and 150 can receive power from a separate power supply, wherein the power supplies for structures 110 and 150 are electrically isolated. The power supply for each of structures 110 and 150 may include a current source, a voltage source, a power rail, a battery, and / or a connection to mains power. The power supply for sensing structure 110 may be a dedicated power supply used only by sensing structure 110, or sensing structure 110 may be configured to share a power supply with another component of device 100. Due to the electrical isolation between structures 110 and 150, the effects of thermomechanical fatigue on sensing structure 110 may not affect the operation of power structure 150.

[0031] However, in some examples, structures 110 and 150 may share a common power supply. It is conceivable that device 100 may include a power supply for both sensing structure 110 and power structure 150, such that the power supply for sensing structure 110 is not separate from that for power structure 150. Even if structures 110 and 150 share a power supply, they can still have different aging mechanisms. For example, regardless of whether they have a common power supply shared by structures 110 and 150, processing circuitry 160 can predict future health problems of power structure 150 by measuring the progression or state of the aging mechanism of sensing structure 110.

[0032] Processing circuitry 160 may be configured to determine whether sensing structure 110 includes a predicted health indicator by measuring the current in sensing structure 110. Additionally or alternatively, processing circuitry 160 may be configured to determine whether sensing structure 110 includes a predicted health indicator by measuring voltage, resistance, capacitance, and / or impedance in sensing structure 110. Processing circuitry 160 may be configured to determine whether sensing structure 110 includes a predicted health indicator in response to determining that current, voltage, resistance, capacitance, and / or impedance are greater than or less than a threshold level.

[0033] The predictive health indicator can predict health problems of the power structure 150. Health problems can include decomposition events of the power structure 150, such as thermomechanical degradation of the power structure 150. The degradation process of the power structure 150 can be caused by thermomechanical fatigue. Over time, the power structure 150 may degrade due to stress caused by start-up and stop currents, rapid changes in current, rapid turn-on and turn-off, and power dissipation caused by on-resistance. Degradation of the power structure 150 may result in short circuits or open circuits between load terminals, or may result in short circuits between control terminals and one or two load terminals.

[0034] To predict health problems in power structure 150 before they occur, sensing structure 110 can provide predictive health indications. Sensing structure 110 can be designed such that the predictive health indicators become apparent before a health problem occurs in power structure 150. For example, a first, very early indicator may become apparent in sensing structure 110 before a health problem occurs in power structure 150. Other predictive health indicators (e.g., associated with higher current levels) may become apparent in sensing structure 110 after the first indicator but before a health problem occurs in power structure 150. Therefore, sensing structure 110 can provide an early indication of future health problems in power structure 150.

[0035] After determining that the sensing structure 110 includes a predicted health indicator, the processing circuitry 160 can be configured to output a signal to an external device 180 via a diagnostic interface 170, wherein the signal includes data about the predicted health indicator. For example, the signal may include data about the type or severity of the predicted health indicator, such as a very early indicator, an early indicator, or a late indicator. The signal may also include data about the time when the processing circuitry 160 detected the predicted health indicator and / or the number of times each predicted health indicator occurred.

[0036] Figure 2 This is a conceptual block diagram of a sensing structure 210 according to the present disclosure, including one or more antifuse 220s and a current sensor 240. The sensing structure 210 also includes an electrical conductor 212 connecting a voltage source 230 to the antifuse 220. Although in Figure 2 and Figure 3 Antifuses 220 and 320 are shown as examples of sensing structures 210 and 310, but other components can be used to provide predictive health indicators. For example, a capacitor can provide a predictive health indicator based on the degradation of its dielectric barrier layer. As the dielectric barrier layer deteriorates, the capacitance and leakage current of the capacitor can change, providing an indication of future health problems in the power structure.

[0037] Antifuse 220 can be used to monitor the health of a power structure. Compared to a fuse, each antifuse 228A-228N can include an electrical device with reverse electrical characteristics. For example, a fuse starts with low resistance and can be designed to permanently disconnect the path when the current through the conductive path exceeds a specified limit. Therefore, the fuse degrades from low resistance to high resistance. Conversely, antifuse 220 can be configured to start with relatively high resistance (e.g., on the order of gigahertz or megaohms) during manufacturing and degrade to lower resistance.

[0038] Programming the antifuse 220 can convert its high resistance into a permanent conductive path with lower resistance (on the order of tens, hundreds, or thousands of ohms). Programming the antifuse 220 can include storing one or more threshold levels in onboard memory to determine how quickly the antifuse 220 crosses those threshold levels. Programming can include setting thresholds for predicting health indicators at the beginning of chip lifespan (such as during automated test equipment phases). The thresholds can be stored in OTP memory, flash memory, and / or any other type of memory. Although... Figure 2 Multiple antifuse wires 220 are shown, but the sensing structure 210 can also be implemented with a single antifuse wire. The sensing structure 210 can have any number of antifuse wires 220, including one, two, three, or more. The antifuse wires 220 can be connected in parallel and / or in series.

[0039] For example, antifuse 228A may include oxide 224 located between metal lines 222 and 226, wherein metal line 222 may be arranged parallel to metal line 226. Oxide 224 may be configured to degrade and leak due to thermomechanical stress experienced by sensing structure 210, which may be related to thermomechanical stress experienced by power structure. Therefore, the leakage current through antifuse 228A may be configured to increase as oxide 224 of antifuse 228A degrades due to thermomechanical stress. Furthermore, due to the degradation caused by thermomechanical stress, the equivalent impedance of antifuse 220 may decrease, and the voltage across antifuse 220 may be reduced.

[0040] The current sensor 240 can be configured to sense leakage current through the antifuse 220. The leakage current is related to the voltage across the antifuse 220 and the impedance of the antifuse 220 by Ohm's law. When the impedance of the antifuse 220 decreases due to thermomechanical stress, the leakage current through the antifuse 220 may increase if the voltage output from the voltage source 230 remains stable. Since the antifuses 228A-228N are connected in series, the equivalent impedance of the antifuses 228A-228N can be equal to the sum of the impedances of each antifuse 228A-228N. Figure 2 As shown in the voltage source example, antifuse 228A-228N can be connected in series so that all antifuse 228A-228N conduct the same current.

[0041] In some examples, voltage source 230 is decoupled from and electrically isolated from the power supply of the power structure. In other examples, voltage source 230 may be configured to supply power to the power structure, or voltage source 230 may receive power from the power supply of the power structure. Conductor 212 connects voltage source 230 and antifuse 220, and conductor 212 may also be electrically isolated from the power structure. In some examples, voltage source 230 may be configured to output a relatively stable voltage of one volt, two volts, three volts, or five volts.

[0042] The processing circuitry can be configured to receive a signal from the current sensor 240, wherein the signal indicates leakage current through the antifuse 220. The processing circuitry can be configured to determine whether the sensing structure 210 includes a predictive health indicator by determining whether the leakage current exceeds a threshold level. In response to determining that the leakage current exceeds the threshold level, the processing circuitry can be configured to set a warning bit. In some examples, the processing circuitry can be configured to apply multiple threshold levels to the leakage current, where each threshold level represents a different warning type (e.g., very early, early, etc.). Each threshold level can be based on the antifuse specifications and arrangement. An early detection threshold can correspond to a resistance greater than 1%, 2%, 5%, or 10% of the original resistance, while a late detection threshold can correspond to a resistance less than 1% of the original resistance.

[0043] The threshold level and other parameters of the sensing structure 210 (e.g., the arrangement of the antifuse 220) can be adjusted through a calibration process. For example, the processing circuitry can be configured to determine the threshold level based on commands from an external device. The calibration process can be used to correlate the sensing structure 210 with a major failure mechanism of the power structure (e.g., an aging mechanism), such as thermomechanical degradation. The threshold level is adjustable to allow for good detection before any health problems occur in the power structure, or the threshold level can be adjusted to be triggered closer to the time when a health problem occurs in the power structure. Therefore, the time between when the aging mechanism of the sensing structure 210 becomes apparent and when the aging mechanism of the power structure becomes apparent is adjustable.

[0044] Figure 3 This is a conceptual block diagram of a sensing structure 310 according to the present disclosure, including one or more antifuse 320s and a voltage comparator 340. The sensing structure 310 also includes an electrical conductor 312 that connects a current source 330 to the antifuse 320 and the voltage comparator 340.

[0045] Current source 330 is configured to drive current through antifuse 320 connected in parallel. Processing circuitry can be configured to determine the voltage drop across antifuse 320 based on the output of voltage comparator 340. In some examples, voltage comparator 340 may include one or more comparators configured to apply one or more threshold levels. In examples where antifuse 320 has relatively high impedance (e.g., during manufacturing), the voltage drop across antifuse 320 may be approximately the same as the supply voltage generated by current source 330. In examples where the impedance of antifuse 320 is reduced, the voltage drop may be closer to zero volts.

[0046] The leakage current through the antifuse 320 can be configured to increase as the oxides of the antifuse 320 deteriorate due to thermomechanical stress. Furthermore, due to the deterioration caused by thermomechanical stress, the equivalent impedance of the antifuse 320 can be reduced, and the voltage across the antifuse 320 can be decreased.

[0047] The processing circuitry can be configured to determine the sensing structure, including a predicted health indicator, by determining that the voltage across the antifuse is less than a voltage threshold level. The processing circuitry can be configured to apply more than one voltage threshold level, where the highest voltage threshold level corresponds to very early detection, and lower voltage threshold levels correspond to earlier detection.

[0048] Figure 4 This is a conceptual block diagram of a sensing structure 410 and a power structure 450 including a power switch 456, according to the present disclosure. Figure 4In one example, device 400 includes a sensing structure 410, a power structure 450, and a power supply 454. However, in other examples, the power supply 454 may be external to device 400. For example, the power supply 454 may be a stand-alone device and / or an external device.

[0049] In some examples, device 400 may include a single semiconductor substrate comprising sensing structure 410 and power structure 450. The single semiconductor substrate may be a single semiconductor material, a single block of semiconductor material, and / or a single semiconductor material chip. In other examples, sensing structure 410 may be off-chip, such that structures 410 and 450 are disposed on a separate semiconductor substrate.

[0050] The size 458 of the power structure 450 can be larger than the size 418 of the sensing structure 410. For example, the size 458 can be at least ten, one hundred, two hundred, one thousand, or two thousand times larger than the size 418. The sizes 418 and 458 of structures 410 and 450 can refer to the chip area covered by structures 410 and 450 or the volume occupied by structures 410 and 450. The size 458 may include the size of the power switch 456 and some or all of the electrical conductors 452. In some examples, the size 458 may include multiple power switches of the power structure 450. Therefore, the sensing structure 410 can provide a predictive health indicator for the power structure 450, even if the sensing structure 410 may be much smaller than the power structure 450. By designing the sensing structure 410 to be smaller than the power structure 450, the sensing structure 410 can increase the size of the device 400 without substantially increasing the size of the device.

[0051] Furthermore, the cross-sectional area of ​​conductor 452 can be larger than the cross-sectional area of ​​the conductor of sensing structure 410. For example, the cross-sectional area of ​​conductor 452 can be at least ten, twenty, one hundred, or one thousand times larger than the cross-sectional area of ​​the conductor of sensing structure 410. Conductor 452 can be connected to the load path of power switch 456 so that current can flow through conductor 452 when power switch 456 is activated.

[0052] The distance 460 between structures 410 and 450 can be small enough that structures 410 and 450 experience associated thermomechanical stress over a period of time. Structures 410 and 450 can experience thermomechanical stress caused by the power cycling of power switch 456. Distance 460 can be less than one hundred, fifty, twenty, ten, five, or two power unit spacings of power switch 456, wherein power switch 456 comprises multiple power units and adjacent power units are separated by a center-to-center power unit spacing (e.g., for trench devices). Distance 460 can be less than five hundred, one hundred, fifty, twenty, ten, five, or two micrometers. Distance 460 can be measured as the distance between the nearest edges of structures 410 and 450. For example, distance 460 can be measured from the nearest edge of sensing structure 410 to the nearest edge of electrical conductor 452 or power switch 456.

[0053] Figure 5 This is a conceptual block diagram of a system 590 including a logic device 560 and a power device 500, based on the technology disclosed herein. Figure 5 In the example shown, processing circuitry 562 is part of logic device 560, which also includes a driver 580 for driving power structure 550. In some examples, processing circuitry 562 may control driver 580, while another processing element may be configured to determine whether sensing structure 510 includes a predictive health indicator. In some examples, logic device 560 may include digital circuitry, while in other examples, logic device 560 may consist partly or entirely of analog circuitry, as a replacement or complement to digital circuitry.

[0054] exist Figure 5 In the example shown, the power supply 554 for the power structure 550 is outside the power device 500. In other examples, the power supply 554 may be integrated into the power device 500, or the power supply 554 may be wholly or partially external to the system 590. The sensing structure 510 may include a power supply that is separate from and electrically isolated from the power supply 554, or the power supply for the sensing structure 510 may be part of or connected to the power supply 554. The power supply for the sensing structure 510 may be integrated on the power device 500 (e.g., inside the sensing structure 510), or the power supply for the sensing structure 510 may be wholly or partially external to the power device 500.

[0055] exist Figure 5 In the example shown, driver 580 is part of logic device 560. In other examples, driver 580 may be separate from logic device 560. For example, driver 580 may be a standalone device, and / or driver 580 may be partially or fully integrated into device 560.

[0056] Figure 6 This is a flowchart illustrating an example technique for detecting a predictive health indicator in a power structure according to the present disclosure. (Reference) Figure 1 The circuit shown is used to describe Figure 6 The technology is similar to that used in other components.

[0057] exist Figure 6 In the example, processing circuitry 160 measures electrical parameters (600) in a sensing structure 110 that is electrically isolated from power structure 150. Electrical parameters may include voltage, current, impedance (e.g., resistance), derivatives of the electrical parameters, and / or any combination of electrical parameters. Processing circuitry 160 may be configured to receive signals output by sensors within sensing structure 110. Processing circuitry 160 may be configured to determine the magnitude or value of the electrical parameters based on the signals received from sensing structure 110.

[0058] exist Figure 6 In one example, processing circuitry 160 determines whether an electrical parameter in sensing structure 110 is greater than a threshold level (602). In some examples, processing circuitry 160 may be configured to compare the electrical parameter with multiple threshold levels. For example, processing circuitry 160 may compare the electrical parameter with a first threshold level representing very early detection of a health problem in power structure 150. Very early detection can occur long before an actual performance problem in power structure 150 occurs. Processing circuitry 160 may compare the electrical parameter with a second threshold level representing early detection of a health problem in power structure 150. Processing circuitry 160 may compare the electrical parameter with a third threshold level representing late detection of a health problem in power structure 150. Late detection can occur before an actual performance problem in power structure 150, but processing circuitry 160 may be configured to output an alarm in response to determining that the third threshold level has been exceeded. By using multiple threshold levels, processing circuitry 160 may be configured to accurately pinpoint the expected lifetime of power structure 150.

[0059] exist Figure 6 In the example, processing circuitry 160 sets a bit (604) indicating a predicted health problem of power structure 150. This bit can be stored in the memory of device 100, wherein the memory is coupled to processing circuitry 160. Processing circuitry 160 can be configured to set the first bit in response to determining that an electrical parameter is greater than a first threshold level, set the second bit in response to determining that an electrical parameter is greater than a second threshold level, and set the third bit in response to determining that an electrical parameter is greater than a third threshold level.

[0060] In some examples, the processing circuitry 160 may be configured to set a bit only after several measurements have exceeded a threshold level to reduce the influence of noise on the determination. For example, in response to determining that an electrical parameter is greater than a threshold level, the processing circuitry 160 may be configured to repeat the process by taking at least one additional measurement to confirm the determination. The processing circuitry 160 may be configured to set a bit in response to determining that N out of M measurements are greater than a threshold level, where N is an integer and M is an integer greater than or equal to N. For example, the processing circuitry 160 may set a bit in response to determining that three out of ten measurements are greater than a threshold level.

[0061] Additionally or alternatively, the processing circuitry 160 may be configured to set a bit in response to an electrical parameter exceeding a threshold level for a predetermined duration. For example, the processing circuitry 160 may determine that the electrical parameter exceeds a threshold level, set a timer, and then check whether the electrical parameter still exceeds the threshold level after the timer expires.

[0062] exist Figure 6 In the example, the processing circuitry 160 outputs an indication of a predicted health problem to an external device 180 (606) via a diagnostic interface 170. The indication may include data about the predicted health problem, such as the time of occurrence, the number of occurrences, and the severity or level of each occurrence.

[0063] Figure 7 This is a flowchart illustrating an example technique for designing semiconductor devices according to the present disclosure. Figure 7 In the example, the designer determines data (700) related to the behavior of the sensing structure through loops. The data may relate to the decomposition of oxides (e.g., impedance) in the antifuse across multiple power cycles. Additionally or alternatively, the data may relate to the current or voltage in the antifuse across multiple power cycles.

[0064] exist Figure 7 In the example, the designer establishes an initial understanding of the fatigue of the selected technology (702). Fatigue data may include the degradation of the power structure over multiple power cycles. The designer then designs a first product chip and / or calibrates the sensing structure (704). As an example, the designer may select the oxide thickness based on the robustness of the power structure. Calibration of the sensing structure may include trimming the antifuse and / or adjusting the threshold level applied by the processing circuitry.

[0065] exist Figure 7In the example, the designer performs tests on the device (706). The tests may include determining when the sensing structure develops a predictive health indicator and whether this development occurs before a health problem in the power structure becomes apparent. The designer can then determine if the sensing structure is effective (708). In an example where the designer determines the sensing structure is ineffective, the designer coordinates the characteristics of the sensing structure and / or modifies the readout circuitry based on the test results (710). The designer may coordinate threshold levels and / or adjust the arrangement and connections of antifuse wires. Figure 7 In the example, the designer verifies the final implementation (712). Once the design is verified, it can be used to produce a batch of devices for distribution to customers.

[0066] This disclosure assigns functionality to processing circuitry arrangement 160. Processing circuitry arrangement 160 may include one or more processors. Processing circuitry arrangement 160 may include any combination of integrated circuit devices, discrete logic circuit devices, and analog circuit devices, such as one or more microprocessors, digital signal processors (DSPs), application-specific integrated circuit devices (ASICs), and / or field-programmable gate arrays (FPGAs). In some examples, processing circuitry arrangement 160 may include multiple components, such as any combination of one or more microprocessors, one or more DSPs, one or more ASICs, or one or more FPGAs, as well as other discrete or integrated logic circuit devices and / or analog circuit devices.

[0067] The techniques described in this disclosure can also be embodied or encoded in products including non-transitory computer-readable storage media, such as memory or storage cells associated with processing circuitry device 160. In some examples, the memory may be local and electrically integrated with the processing circuitry device, or in other examples, the memory may be external and electrically connected to the processing circuitry device 160, such as via a data bus or direct connection. Exemplary non-transitory computer-readable storage media may include RAM, ROM, programmable ROM (PROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, magnetic media, optical media, or any other computer-readable storage device or tangible computer-readable medium. The term "non-transitory" may indicate that the storage medium is not embodied in a carrier wave or propagating signal. In certain instances, non-transitory storage media are capable of storing data that may change over time (e.g., in RAM or cache).

[0068] The following numbered examples illustrate one or more aspects of this disclosure.

[0069] Example 1. An apparatus including a power structure and a sensing structure electrically isolated from the power structure. The apparatus further includes processing circuitry configured to determine whether the sensing structure includes a predictive health indicator, wherein the predictive health indicator indicates the health of the power structure.

[0070] Example 2. The device according to Example 1, wherein the power structure and the sensing structure are formed on a single semiconductor substrate.

[0071] Example 3. A device according to Example 1 or 2, wherein the first aging mechanism of the power structure is different from the second aging mechanism of the sensing structure.

[0072] Example 4. An apparatus according to Examples 1-3 or any combination thereof, wherein the processing circuitry is configured to determine whether the sensing structure includes a predictive health indicator by measuring a second aging mechanism.

[0073] Example 5. A device according to Examples 1-4 or any combination thereof, wherein the first aging mechanism includes thermomechanical degradation of the power structure.

[0074] Example 6. A device according to Examples 1-5 or any combination thereof, wherein the second aging mechanism includes the decomposition of oxides in the sensing structure.

[0075] Example 7. A device according to Examples 1-6 or any combination thereof, wherein the sensing structure and the power structure experience associated thermomechanical stress over a period of time.

[0076] Example 8. A device based on Examples 1-7 or any combination thereof, wherein the predicted health indicator becomes apparent before the thermomechanical degradation of the power structure becomes apparent.

[0077] Example 9. A device according to Examples 1-8 or any combination thereof, wherein the power structure is powered by a first power source and the sensing structure is powered by a second power source electrically isolated from the first power source.

[0078] Example 10. A device according to Examples 1-9 or any combination thereof, wherein the sensing structure includes an antifuse.

[0079] Example 11. A device according to Examples 1-10 or any combination thereof, wherein the sensing structure includes a current sensor configured to sense the current flowing through the antifuse.

[0080] Example 12. An apparatus according to Examples 1-11 or any combination thereof, wherein the processing circuitry is configured to determine that the sensing structure includes a predictive health indicator by determining that the current flowing through the antifuse is greater than a threshold level.

[0081] Example 13. A device according to Examples 1-12 or any combination thereof, wherein the antifuse includes an initial resistance of more than one hundred megohms during manufacturing.

[0082] Example 14. A device based on Examples 1-13 or any combination thereof, wherein the threshold level corresponds to a resistance of less than one megaohm.

[0083] Example 15. The device according to Examples 1-14 or any combination thereof also includes a programmable memory configured to store a threshold level.

[0084] Example 16. The device according to Examples 1-15 or any combination thereof also includes a diagnostic interface configured to be communicatively coupled to an external device.

[0085] Example 17. A device according to Examples 1-16 or any combination thereof, wherein the processing circuitry is configured to store new values ​​for setting threshold levels in a programmable memory based on communications received from an external device via a diagnostic interface.

[0086] Example 18. A device according to Examples 1-17 or any combination thereof, wherein the antifuse is a first antifuse, and the sensing structure further includes a second antifuse connected in series with the first antifuse.

[0087] Example 19. A device according to Examples 1-18 or any combination thereof, wherein a current sensor is configured to sense current through a first antifuse and through a second antifuse.

[0088] Example 20. A device according to Examples 1-19 or any combination thereof, wherein the sensing structure includes a current source configured to drive current through an antifuse.

[0089] Example 21. A device according to Examples 1-20 or any combination thereof, wherein the sensing structure includes a voltage comparator configured to compare the voltage across the antifuse with a voltage threshold level when a current source drives current through the antifuse.

[0090] Example 22. An apparatus according to Examples 1-21 or any combination thereof, wherein the processing circuitry is configured to determine that the sensing structure includes a predictive health indicator by determining that the voltage across the antifuse is less than a voltage threshold level.

[0091] Example 23. A device according to Examples 1-22 or any combination thereof, wherein the antifuse is a first antifuse, and the sensing structure further includes a second antifuse connected in parallel with the first antifuse.

[0092] Example 24. A device according to Examples 1-23 or any combination thereof, wherein the current source is configured to drive current through a first antifuse and through a second antifuse.

[0093] Example 25. A device according to Examples 1-24 or any combination thereof, wherein the voltage comparator is configured to compare the voltage across the first and second antifuses with a voltage threshold level as the current source drives the current through the first antifuse and through the second antifuse.

[0094] Example 26. A device according to Examples 1-25 or any combination thereof, wherein the size of the power structure is at least one hundred times larger than the size of the sensing structure.

[0095] Example 27. A device according to Examples 1-26 or any combination thereof, wherein the distance from the sensing structure to the power structure is less than ten times the spacing of the power cells in the power structure.

[0096] Example 28. A device according to Examples 1-27 or any combination thereof, wherein the power structure includes a power switch.

[0097] Example 29. A device according to Examples 1-28 or any combination thereof, wherein the power structure includes a first conductor electrically connected to the load path of the power switch.

[0098] Example 30. A device according to Examples 1-29 or any combination thereof, wherein the sensing structure includes an antifuse and a second conductor electrically connected to the antifuse.

[0099] Example 31. An apparatus according to Examples 1-30 or any combination thereof, wherein the cross-sectional area of ​​the first conductor is at least ten times greater than the cross-sectional area of ​​the second conductor.

[0100] Example 32. A system comprising a power substrate, the power substrate comprising the power structure and sensing structure of Examples 1-31 or any combination thereof.

[0101] Example 33. The system according to Example 32, wherein the processing circuitry is configured to control the driver to generate a driver signal for the control power structure.

[0102] Example 34. A system according to Example 32 or 33 or any combination thereof, wherein the processing circuitry is configured to measure current in a sensing structure electrically isolated from the power structure.

[0103] Example 35. A system according to Examples 32-34 or any combination thereof, wherein the processing circuitry is configured to determine whether the current in the sensing structure is greater than a threshold level, and in response to determining that the current in the sensing structure is greater than the threshold level, set a bit indicating a predicted health problem.

[0104] Example 36. A system according to Examples 32-35 or any combination thereof, wherein the processing circuitry is configured to measure the current in the sensing structure, including measuring the leakage current through the antifuse; and determining whether the current is greater than a threshold level includes determining whether the leakage current is greater than the threshold level.

[0105] Example 37. A method comprising measuring current in a sensing structure electrically isolated from a power structure. The method further comprises determining that the current in the sensing structure is greater than the threshold level. The method further comprises setting a bit indicating a predicted health problem in response to determining that the current in the sensing structure is greater than the threshold level.

[0106] Example 38. The method according to Example 37 also includes any technique performed by the processing circuitry of Examples 1-31 or any combination thereof.

[0107] Example 39. A method comprising a computer-readable medium having executable instructions stored thereon, the executable instructions being configured to be executed by a processing circuit means to cause the processing circuit means to perform the techniques of Examples 1-31 or any combination thereof.

[0108] Example 40. A system comprising means for performing the methods of Examples 1-31 or any combination thereof.

[0109] Various examples of this disclosure have been described. Any combination of the described systems, operations, or functions is contemplated. These and other examples are within the scope of the appended claims.

Claims

1. An apparatus comprising: Power structure; The sensing structure is electrically isolated from the power structure; as well as A processing circuit device is configured to determine whether the sensing structure includes a predicted health indicator, wherein the predicted health indicator indicates the health of the power structure; The first aging mechanism of the power structure differs from the second aging mechanism of the sensing structure. The processing circuitry is configured to determine whether the sensing structure includes the predicted health indicator by measuring the second aging mechanism.

2. The device according to claim 1, wherein, The power structure and the sensing structure are formed on a single semiconductor substrate.

3. The device according to claim 1, in, The first aging mechanism includes the thermomechanical degradation of the power structure, and The second aging mechanism includes the decomposition of oxides in the sensing structure.

4. The device according to claim 3, in, The sensing structure and the power structure experience relevant thermomechanical stresses over a period of time, and The predicted health indicator becomes apparent before the thermomechanical degradation of the power structure becomes apparent.

5. The device according to claim 1, in, The power structure is powered by a first power source, and The sensing structure is powered by a second power source that is electrically isolated from the first power source.

6. The device according to claim 1, in, The sensing structure also includes: anti-fuse wire; and A current sensor is configured to sense the current flowing through the antifuse, and The processing circuitry is configured to determine that the sensing structure includes the predicted health indicator by determining that the current flowing through the antifuse is greater than a threshold level.

7. The device according to claim 6, in, The antifuse wire is manufactured with an initial resistance exceeding one hundred megohms, and The threshold level corresponds to a resistance of less than one megaohm.

8. The device according to claim 6, further comprising: A programmable memory is configured to store the threshold level; as well as The diagnostic interface is configured to be communicatively coupled to an external device. The processing circuitry is configured to store new values ​​for the threshold level in the programmable memory based on communication received from the external device via the diagnostic interface.

9. The device according to claim 6, in, The antifuse is the first antifuse. The sensing structure further includes a second antifuse wire connected in series with the first antifuse wire, and The current sensor is configured to sense the current passing through the first antifuse and the second antifuse.

10. The device according to claim 1, in, The sensing structure also includes: anti-fuse wire; A current source is configured to drive current through the antifuse; and A voltage comparator is configured to compare the voltage across the antifuse with a voltage threshold level when the current source drives the current through the antifuse. The processing circuitry is configured to determine that the sensing structure includes the predicted health indicator by determining that the voltage across the antifuse is less than the voltage threshold level.

11. The device according to claim 10, in, The antifuse is the first antifuse. The sensing structure further includes a second antifuse wire connected in parallel with the first antifuse wire. The current source is configured to drive current through the first antifuse and through the second antifuse, and The voltage comparator is configured to compare the voltage across the first and second antifuses with a voltage threshold level when the current source drives the current through the first and second antifuses.

12. The device according to claim 1, wherein, The size of the power structure is at least one hundred times larger than the size of the sensing structure.

13. The device according to claim 1, wherein, The distance from the sensing structure to the power structure is less than ten times the spacing between the power units in the power structure.

14. The device according to claim 1, in, The power structure includes: Power switches; and The first conductor is electrically connected to the load path of the power switch. The sensing structure includes: anti-fuse wire; and The second conductor is electrically connected to the antifuse wire, and The cross-sectional area of ​​the first conductor is at least ten times larger than the cross-sectional area of ​​the second conductor.

15. A method comprising: Measure the current in a sensing structure that is electrically isolated from the power structure; Determine that the current in the sensing structure is greater than a threshold level; as well as In response to determining that the current in the sensing structure is greater than the threshold level, a bit indicating a predicted health problem is set; The sensing structure includes an antifuse wire. Measuring the current in the sensing structure includes measuring the leakage current through the antifuse, and Determining whether the current is greater than the threshold level includes determining whether the leakage current is greater than the threshold level.

16. A system comprising: A power substrate includes a power structure and a sensing structure, wherein the sensing structure is electrically isolated from the power structure; The processing circuit device is configured as follows: Control the driver to generate driver signals that control the power structure; Measure the current in the sensing structure that is electrically isolated from the power structure; Determine whether the current in the sensing structure is greater than a threshold level; and In response to determining that the current in the sensing structure is greater than the threshold level, a bit indicating a predicted health problem is set; The sensing structure includes an antifuse wire, and The processing circuit device is configured as follows: Measuring the current in the sensing structure includes measuring the leakage current through the antifuse; and Determining whether the current is greater than the threshold level includes determining whether the leakage current is greater than the threshold level.

17. The system according to claim 16, in, The sensing structure includes oxides. The processing circuitry is configured to determine whether the sensing structure includes a predictive health indicator by measuring the decomposition of the oxide. The sensing structure and the power structure undergo relevant thermomechanical stress over a period of time, and The decomposition of the oxides becomes apparent before the thermomechanical degradation of the power structure becomes significant.

Images