AGING DETECTOR FOR AN ELECTRICAL CIRCUIT COMPONENT, METHOD FOR MONITORING THE AGING OF A CIRCUIT COMPONENT, COMPONENT AND CONTROL UNIT

Description

[0001] The present invention relates to an aging detector for an electrical circuit component, a method for monitoring the aging of a circuit component, a component and a control unit. State of the art

[0002] Integrated circuits today contain a multitude of active and passive components, also referred to synonymously as components, such as transistors, resistors, and capacitors. These components are subject to aging mechanisms in both active and passive operation, for example, temperature instability caused by negative bias (NBTI). negative bias temperature instability ), HCl and / or electromigration, which can lead to a degradation of circuit component parameters, for example changes in threshold voltage, saturation current, internal resistance and / or breakdown voltage, and thus to a malfunction or failure of the integrated circuit.

[0003] Information can be programmed into the components in a way that allows for non-destructive reading, from which a serial number and / or a production lot, a wafer number, an installation position of the component and / or a packaging lot can be determined.

[0004] In safety-critical applications, such as in automotive applications, malfunctions or failures must be avoided at all costs. Therefore, it is desirable to gather information about aging processes.

[0005] Using characterization data on the type and extent of parameter degradation obtained from test structures and example components using standardized measurement methods, such as JEDEC JP001.01 and AEC-Q100, the influence of different aging mechanisms on an integrated circuit can be compensated for, at least for one or more typical applications, through appropriate design measures. However, in real-world field applications, very different aging degradation patterns arise due to varying circuit types, operating conditions, and environmental parameters, such as temperature, as well as manufacturing-related variations in the semiconductor material's properties.

[0006] DE000010161998A1 discloses a method for monitoring the operation of a system with multiple components, wherein representative parameters for the aging of the components are recorded, wherein the degree of aging of the individual components is determined by evaluating the parameter values ​​for the individual components and the exceeding of a threshold value of an aging factor is reported, wherein the aging factor is determined by cumulating the degrees of aging.

[0007] Further state of the art is described in EP 2 884 663 A1, US 2011 / 101990 A1, US 2009 / 113358 A1, US20120119825 A1, US20090113358 A1, US20100214007 A1, by Keane et al, "On-Chip Silicon Odometers for Circuit Aging Characterization", in T. Grasser (ed.), "Biss Temperature Instability for Devices and Circuits", Springer Science+Business Media New York 2014, pp. 679-717, and by Kumar et al, "On-Chip Aging Compensation for Output Driver", IEEE Reliability Physics Symposium, 2014, pages CA.3.1 - CA.3.5. Disclosure of the invention

[0008] According to the invention, an aging detector according to claim 1, a method according to claim 6, a component according to claim 9 and a control unit for a vehicle according to claim 10 are provided.

[0009] The aging detector is designed for an electrical circuit component to monitor the aging of the circuit component.

[0010] The aging detector according to the invention comprises at least one input for detecting an aging-specific parameter of the circuit component of a device and is configured to use at least the detected parameter to determine an associated reaction threshold and / or a reaction, or to adjust the reaction threshold and / or the reaction. The aging detector is further configured to trigger the reaction, at least in response to the detected parameter exceeding the determined reaction threshold. The aging detector is characterized in that it is further configured to use the detected parameter to adjust the reaction threshold, wherein the adjustment of the reaction threshold is achieved using information obtained by correlating readout field data of the detected parameter with manufacturing data and application conditions to identify a failure behavior.

[0011] The ability to determine or adjust the reaction threshold and / or reaction allows data to be used for determining the reaction threshold and / or reaction that was not available at the time the component was manufactured.

[0012] According to the invention, the measured parameter is the frequency and / or duration of a component-critical application of the circuit component. This allows the influence of the component-critical application on aging to be determined.

[0013] The aging detector may include at least one output and be equipped to apply a voltage to it, and the aging detector may further be equipped to detect an electrical resistance or a threshold voltage drift in response to the application of a voltage.

[0014] This is advantageous because the electrical resistance can be used to monitor electromigration and / or the threshold voltage drift to monitor a temperature instability effect (NBTI effect) of a MOS transistor caused by negative bias.

[0015] The method presented according to the invention serves to monitor the aging of an electrical circuit component of a device. It comprises the following steps: Determining at least one aging-specific parameter, determining an associated parameter-specific reaction threshold, triggering a reaction at least in response to the reaction threshold being exceeded by the determined parameter, and adjusting the reaction threshold as well as using the determined parameter to adjust the reaction threshold, wherein the adjustment of the reaction threshold is carried out by means of information obtained by correlating read field data of the recorded parameter with manufacturing data and application conditions to identify a failure behavior.

[0016] The ability to determine or adjust the reaction threshold and / or reaction allows data to be used for determining the reaction threshold and / or reaction that was not available at the time the component was manufactured.

[0017] According to the invention, the specified parameter is a frequency and / or a duration of a component-critical application of the circuit component.

[0018] The procedure may further include: capturing manufacturing data during the manufacture of the circuit component and using the specified manufacturing data to adjust the reaction threshold and / or the reaction.

[0019] This allows fluctuations in the aging resistance of different production batches to be compensated for even after the fact by adjusting the reaction threshold and / or the reaction itself.

[0020] The component presented according to the invention comprises the aging detector according to the invention. The control unit presented according to the invention is suitable for a vehicle and comprises the component according to the invention.

[0021] In a preferred embodiment, the control unit is prepared to carry out the method according to the invention.

[0022] Advantageous embodiments of the invention are specified in the dependent claims and described in the description. Drawings

[0023] Exemplary embodiments of the invention are explained in more detail with reference to the drawings and the following description. The drawings show: Figure 1 a block diagram of a component monitored according to an embodiment of the invention, Figure 2 a block diagram of a control unit monitoring components according to an embodiment of the invention, Figure 3 an adjustment of the reaction threshold, and Figure 4 An example of the shift ΔP of an aging parameter P.

[0024] If faults occur in safety-critical electronic systems under special stress conditions, these systems must be replaced – among other reasons for product liability – to ensure that the component-critical combination of high stress conditions and lack of resilience does not occur.

[0025] The load-bearing capacity is an individual circuit component property that is subject to fluctuations in the actual manufacturing process.

[0026] In turn, a fault can depend on a specific stress that has occurred previously, for example on the delivery route, at the installation site or during the installation of the component, by the place of use, for example its climate, and by user behavior.

[0027] However, if it is not known that a fault only occurs as a result of a specific stress, this can lead to entire production batches having to be recalled, even though only the circuits subjected to the specific stress would need to be recalled.

[0028] In the embodiments of the invention described below, a method is presented which includes monitoring one or more degradation mechanisms of a component and uses a reaction threshold. The reaction threshold is specific to the component and its resilience (for example, to temperature or humidity stresses). The reaction threshold is adjusted using the monitored degradation mechanisms.

[0029] If degrading operating conditions or combinations thereof (e.g., supply voltage, humidity, high temperature) persist over extended periods, appropriate corrective measures can be initiated when predefined thresholds are reached. These measures can then prevent or delay a failure. Depending on the type of fault, activating an alternative operating mode (e.g., slowing down a processor clock speed) or implementing redundancy may be used as a remedy. In extreme cases and with safety-critical components, the need for preventive component replacement can be signaled.

[0030] An example of a monitored degradation mechanism is the frequency and / or duration of device-critical operating conditions. Examples of such conditions include overvoltages, reverse polarity, and overtemperature. While the components are designed to withstand these conditions without damage for specified frequencies and / or durations, if these conditions occur more frequently and / or for longer periods than originally specified, this can lead to sustained component degradation and ultimately component failure.

[0031] Exemplary training courses also adapt the type of reaction to reaching or exceeding a reaction threshold using the monitored degradation mechanisms.

[0032] The possibility of subsequent adjustment means that data which were not available at the time of the component's manufacture can also be used to determine the reaction threshold and / or the reaction itself.

[0033] Essentially, it's about the possibility of feeding back information about manufacturing problems or reliability weaknesses into ASICs already installed in the field, so that critical degradations can be detected and rectified before a malfunction occurs.

[0034] For this purpose, the status data of detectors already implemented in the field are read out (field observation) and correlated with manufacturing data and application conditions to identify potential reliability problems. Based on the information obtained, the detector thresholds in the ASIC are adjusted only for affected ASICs (for example, limited to a production batch, a specific production period, or a specific application) so that the ASIC can warn in time before reaching a critical state.

[0035] For example, the evaluation of field data for a specific ASIC type reveals that in regions with high humidity, a limited number of vehicle types (specific installation situation) experience a cluster of failures. Correlation with manufacturing data narrows down the affected ASICs to a narrowly defined production period. Furthermore, the evaluation of detector data shows that in all affected ASICs, a specific detector (e.g., humidity) exhibited elevated values, but had not yet reached its originally set reaction threshold. The invention presented here allows the reaction threshold for the humidity detector to be lowered during inspection of ASICs from the corresponding production period that are installed in the affected vehicles, so that the detector reaches its threshold before the ASIC fails.By limiting the recall to the truly affected ASICs, both the costs of recall campaigns and the number of field failures can be significantly reduced.

[0036] Further exemplary embodiments of the invention relate to aging detectors for direct or indirect monitoring of aging mechanisms or component-critical application situations.

[0037] One embodiment of an aging detector is configured to apply a voltage via an output to a conductor arrangement not directly used in the application circuit, to detect its electrical resistance and to use it for monitoring electromigration.

[0038] Another embodiment of an aging detector is configured to apply a voltage via an output to one or simultaneously or successively to several MOS transistors, to detect a threshold voltage drift and to use it for monitoring the NBTI effect.

[0039] Another embodiment of an aging detector is configured to detect one or more overvoltages at an input, which can only occur in a component-critical application situation, for example, charge dump when disconnecting a vehicle battery, and to use a counter to count the number of individual events that have occurred.

[0040] Yet another embodiment of an aging detector is configured to use a diode as a temperature sensor. When a specific temperature is exceeded, the duration of the overtemperature can be recorded and summed up in combination with a timer signal based on an oscillator.

[0041] The method can also be used with other aging detectors, the monitoring of which serves to determine and store one or more or a combination of several aging parameters.

[0042] The evaluation of the key parameters can be carried out, for example, during every start-up procedure of the circuit in the field.

[0043] For safety-critical applications, an evaluation can also be carried out in-situ permanently or at regular time intervals in the field.

[0044] The timing and type of evaluation can be controlled either internally in the integrated circuit or externally by the control unit.

[0045] This could be the control unit in which the component or integrated circuit is installed. In applications with multiple control units, for example in the automotive sector, another control unit can also take over this control function.

[0046] In some embodiments, it is additionally or alternatively possible to control the system externally, for example during an inspection or maintenance, for example via a serial peripheral interface (SPI). serial peripheral interface) or other communication interfaces, such as mobile communications, mobile internet or other radio technologies in conjunction with internal vehicle communication, to trigger an evaluation.

[0047] During the evaluation, these parameters are compared with reaction thresholds. The specific reaction thresholds can be defined or adjusted at different times.

[0048] Example time points include a time of wafer testing or final testing, a time point after installation in a control unit through communication with the component, a time point after installation in a vehicle or other application object, maintenance or inspection times, and operating times during normal operation or in parked mode.

[0049] After installation in the control unit, the evaluation is carried out, for example, via SPI, mobile communications or other radio technologies, for example in conjunction with the internal vehicle communication.

[0050] The determination of the reaction thresholds is initially based on simulation or qualification data or the aging-specific component parameters.

[0051] For continuously changing aging parameters, such as parameter drift, the monitored parameters are evaluated after completion of component testing (e.g., using AEC-Q100), and a tested parameter range is documented based on their maximum value. For example, during testing, the drift (and / or displacement) ΔP = PE - P0 of the aging parameter P is characterized. In series production, the reaction threshold PRS is then determined from the initial value Pinitial measured during component testing and the drift characterized during testing: PRS = Pinitial + 0.9 * ΔP. The scaling factor (0.9 in this example) is set such that a warning can be issued during operation before the tested parameter range is exceeded.

[0052] For parameters in the form of counters, the reaction threshold is determined based on occurrence frequencies specified in the requirements document, which have been verified in testing.

[0053] For warnings regarding state variables (e.g., temperature), the threshold is determined based on maximum values ​​specified in the requirements document. For this purpose, the detector signal (e.g., current or voltage of the temperature diode) must be calibrated accordingly during component testing. This can be achieved by measuring the detector signal at various temperatures (typically room temperature 25°C and high temperature 140-150°C) and storing the temperature characteristics as derived temperature coefficients in the component memory. Alternatively, the reaction threshold can be derived directly from the high-temperature signal (e.g., diode current at 150°C). To do this, it is first characterized on test components how much the diode current I180 at the desired temperature threshold (e.g., 180°C) exceeds the diode current I150 at the measurement temperature used in the component test (e.g., 150°C).Using this correction factor KF = I180 / I150 and the initial measured value Pinitial during the component test at 150°C, the reaction threshold can be derived in series production: PRS = Pinitial*KF.

[0054] In addition, application-specific information such as vehicle, control unit, installation location, maintenance intervals, and known environmental parameters (e.g., expected temperature) can be used to determine the reaction threshold. The goal in defining the reaction threshold is to trigger an alarm before a component-critical fault occurs.

[0055] Furthermore, the reaction thresholds can initially be undefined, i.e., without a value. In this case, the aging parameter is initially determined continuously without being compared to a reaction threshold, and a reaction threshold is only determined at a later time.

[0056] Once installed, the aging parameters can be read out regularly. Example times for reading the parameters include maintenance and inspection intervals, as well as operating times during normal operation or in parked mode. Reading the parameters can be performed via SPI, cellular networks, or other wireless technologies, for example, in conjunction with the vehicle's internal communication system.

[0057] The aging data collected in this way can be evaluated in a component- and application-specific manner and used to further adjust the reaction thresholds.

[0058] In addition, information about incidents during wafer processing or packaging or the transport of individual production batches, wafers or components can be used subsequently to adjust the reaction thresholds.

[0059] Furthermore, any other information that indicates a possible failure of the component can also be used to adjust the reaction thresholds.

[0060] The type of reaction to reaching or exceeding a reaction threshold can be defined differently.

[0061] Example responses include: no response, deactivation of the component or individual functional groups, warning during inspection or maintenance, warning to a user during operation by means of a warning signal device, for example a yellow or red warning light, and warning to the manufacturer of the component, for example via SPI, mobile communications or other radio technologies, for example in connection with internal vehicle communication.

[0062] The type of response can be defined or adjusted at various adjustment points. Examples of such points include the time of wafer testing or final testing, the time after installation in the control unit through communication with the component, the time after installation in the vehicle or other application, maintenance or inspection times, and operating times during normal operation or in parked mode.

[0063] In the Figure 1Figure 1 shows a block diagram of a component 100 monitored according to an embodiment of the invention. The component 100 comprises an evaluation unit 10, detectors 20-1, ..., 20-x for aging, and detectors 30-1, ..., 30-y for component-critical applications. The evaluation unit 10 communicates, preferably, but not necessarily, bidirectionally with the aging detectors 20-1, ..., 20-x and with the detectors 30-1, ..., 30-y for component-critical applications. The evaluation unit 10 also communicates, again preferably, but not necessarily, bidirectionally with an associated control unit, with the application in which the component 100 is used, and with devices outside the application.

[0064] In the Figure 2 A block diagram of a control unit 200 monitoring components according to an embodiment of the invention is shown.

[0065] The control unit comprises, for example, two monitored components 100-1, 100-2 and a control unit 40. The control unit 40 communicates, preferably, but not necessarily bidirectionally, with the evaluation units of the monitored components 100-1, 100-2. The control unit 40 also communicates, again preferably, but not necessarily bidirectionally, with the application object in which the component 100 is used and with devices outside the application object.

[0066] Figure 3This shows the adjustment of a reaction threshold. An ASIC, which monitors various aging mechanisms (measurement curves 310-1, 310-2, ..., 310-n recorded by a respective sensor A) and / or critical situations (measurement curves 320-1, 320-2, ..., 320-n recorded by a respective sensor B), is used in various vehicles 10-1, 10-2, ..., 10-n in the field 300. During regular servicing and / or in the event of faults, these sensors are read out together with the ASIC ID at a service center 400. From the correlation 600 of this field data with production data from the manufacturing plant 500, conspicuous production batches are identified. For example, that all ASICs from a specific production period may fail unexpectedly upon reaching a reduced threshold of sensor B.Based on this evaluation, the threshold for sensor B can be lowered during the next service or as part of a recall, so that the system can warn the driver in good time before reaching the newly defined failure range.

[0067] Figure 4 This shows an example of the shift ΔP of the aging parameter P during component testing from the initial value P0 to the final value PE. As an example, the corresponding reaction threshold PRS for a component delivered to the field can be derived from its initial measured value Pinitial and the characterized shift ΔP as follows: PRS = Pinitial + 0.9 * ΔP.

Claims

1. Ageing detector (30) for an electrical circuit component (100, 100-1, 100-2) of a component for monitoring ageing of the circuit component (100, 100-1, 100-2), wherein the ageing detector (30) comprises at least one input for detecting an ageing-specific parameter of the circuit component (100, 100-1, 100-2) and is configured for adapting a response threshold, wherein the ageing detector (30) is also configured to trigger a response at least in response to transgression of the response threshold by the detected parameter, characterized in that the ageing detector (30) is configured to use the detected parameter for adapting the response threshold, wherein the response threshold is adapted through information obtained by correlating field data relating to the detected parameter that are read out by the ageing detector (30) with production data and application conditions in order to identify a failure behaviour, wherein the detected parameter is a frequency and / or a duration of a component-critical application of the circuit component (100, 100-1, 100-2).

2. Ageing detector (30) according to Claim 1, wherein the ageing detector (30) is also configured to use the detected parameter for adapting the response.

3. Ageing detector according to either one of Claims 1 and 2, wherein the ageing detector (30) comprises at least one output and is configured to apply to it a voltage, and the ageing detector (30) is also configured to detect electrical resistance or a threshold voltage drift in response to application of a voltage.

4. Ageing detector according to Claim 3, wherein the electrical resistance is used for monitoring an electromigration and / or the threshold voltage drift is used for monitoring a negative-bias temperature instability effect (NBTI effect) of a MOS transistor.

5. Ageing detector according to Claim 4, wherein the electrical resistance is used for monitoring an electromigration and / or the threshold voltage drift is used for monitoring a negative-bias temperature instability effect (NBTI effect) of a metal-oxide semiconductor transistor (MOS transistor).

6. Method for monitoring ageing of an electrical circuit component of a component, with the steps: determining at least one ageing-specific parameter, determining an associated parameter-specific response threshold, and triggering a response at least in response to transgression of the response threshold by the determined parameter and adapting the response threshold, characterized in that the method also comprises using the determined parameter for adapting the response threshold, wherein the response threshold is adapted through information obtained by correlating field data relating to the determined parameter that are read out with production data and application conditions in order to identify a failure behaviour, wherein the determined parameter is a frequency and / or a duration of a component-critical application of the circuit component (100, 100-1, 100-2).

7. Method according to Claim 6, also comprising using the determined parameter for adapting the response.

8. Method according to one of Claims 6 to 7, furthermore comprising: detecting production data during the production of the circuit component and using the detected production data for adapting the response threshold and / or the response.

9. Component (100, 100-1, 100-2) comprising an ageing detector (30) according to one of Claims 1 to 5.

10. Control device (200) for a vehicle, wherein the control device (200) comprises a component (100, 100-1, 100-2) according to Claim 9.

11. Control device according to Claim 10, configured for carrying out a method according to Claims 6 to 8.