TEST METHOD AND TEST SYSTEM FOR TESTING EFUSES

Description

[0001] The invention relates to a method for testing switchable electronic fuses according to the preamble of claim 1. Furthermore, the invention relates to a test system for testing switchable electronic fuses according to the preamble of claim 8.

[0002] Electronic fuses, hereinafter referred to as efuses, are semiconductor circuits used to limit or interrupt the current drawn by an electrical load. Similar to conventional fuses, efuses are destroyed when, in an overcurrent event (for example, a short circuit in the electrical load), a current exceeding a permissible rated current flows through them for a predetermined time. An efuse that has been correctly destroyed in an overcurrent event is typically permanently (irreversibly) non-conductive; however, resettable efuses are also known that can be returned to their initial state after an overcurrent event.

[0003] Switchable efuses are electronically switchable between a conducting and a non-conducting state until an overcurrent occurs. For example, a switchable efuse can be designed as a field-effect transistor, such as an n-channel metal-oxide-semiconductor field-effect transistor, which can be switched between a conducting and a non-conducting state by means of a voltage applied to the gate electrode.

[0004] Manufacturing or operational defects, or age-related degradation of efuses, can lead to a situation where, in the event of an overcurrent, the electrical load being protected is not reliably disconnected from the supply voltage. This compromises the operational safety of the electrical load. For example, if such an electrical load is a safety-critical electrical subsystem of a vehicle, the vehicle's operational safety may be affected. Therefore, it is essential to regularly check the working condition of such efuses.

[0005] For this purpose, the Efuse can be tested in an operating mode with a disconnected electrical load, for example, in a vehicle in a parked state with the safety-critical electrical subsystem switched off. However, this method is not applicable for checking the operating status of an Efuse during operation and is therefore not suitable for continuous monitoring.

[0006] Furthermore, a redundant second efuse can be arranged in parallel to a first efuse between a power supply and an electrical load to be protected. Both efuses can then be alternately disconnected from the electrical load and tested in isolation without affecting the power supply to the electrical load. However, this redundancy leads to increased manufacturing and operating costs, for example, due to increased space requirements on printed circuit boards or chips and / or increased power consumption because additional efuses must be provided.

[0007] Document US 2001 / 0034070 A1 describes a device and method for testing fuses in a storage device. A voltage is applied across a control resistor, which decays over time. The decaying voltage across the control resistor is compared to a reference voltage value using a comparator. The time after which the comparator changes its state is determined. Fuses are tested by comparing this determined time with the fuse under test by applying a decaying voltage across the fuse and determining the time until the comparator switches. If the fuse has a specified resistance value, the comparator switches its state within the time determined from the control resistor.By adjusting the comparator's reference voltage, the time it takes for the comparator to switch for a specified fuse resistance value can be reduced, thus shortening the test time. However, this method is limited to testing non-switchable electrical connections and, in particular, to determining a state of sufficient conductivity (i.e., sufficiently low resistance). Furthermore, document US 2009 / 079439 A1 describes an Efuse system and a method for testing the Efuse system, in which the resistance of the electronic fuse is measured. This resistance depends on whether the electronic fuse has blown or not. Therefore, only complete destruction of the electronic fuse is detected, and not a gradual change in the fuse's performance.

[0008] Therefore, there is a need for an improved test procedure to determine the operating state of a switchable fuse, in particular to differentiate between a safe operating state, in which the electrical load is reliably disconnected from the supply voltage in the event of an overcurrent, and an unsafe operating state, in which the electrical load is not reliably disconnected from the supply voltage in the event of an overcurrent. Furthermore, it may also be necessary to detect operating states in which the electrical load is disconnected in the event of an overcurrent, but certain parameters (for example, the disconnection time or the remaining residual conductivity along the fuse) do not correspond to the specified target values.

[0009] The invention is based, in a first aspect, on the objective of providing an improved method for testing a switchable fuse. This objective is achieved according to the invention by a method with the features of claim 1.

[0010] The invention further aims to provide an improved test system for testing a switchable fuse. This objective is achieved according to the invention by a test system with the features of claim 8.

[0011] Advantageous embodiments of the invention are the subject of the dependent claims.

[0012] The first aspect of the invention relates to a method for testing a switchable efuse.

[0013] In a switchable efuse that can be switched between at least one conductive and one non-conductive state and through which at least one electrical load is electrically connected to a supply voltage source, at least one first, functional operating state and a second, non-functional operating state can be distinguished.

[0014] In the first operating state, at least one electrical load is reliably disconnected from the supply voltage source if the current drawn by the electrical load via the fuse exceeds a predetermined rated current for a predetermined period. In the second operating state, disconnection of the electrical load under these conditions is not guaranteed or not reliably guaranteed.

[0015] The at least one electrical load is designed to tolerate a power supply interruption (i.e., a disconnection from the supply voltage source) for a predetermined maximum interruption duration.

[0016] The maximum interruption duration is the maximum time period during which the operation of the electrical load can be maintained without power from the supply voltage source, taking all tolerances into account, for example by means of internal buffer storage in the electrical load.

[0017] In a test procedure to verify the operating state of a switchable fuse, the switchable fuse is switched from a conductive state to a non-conductive state for a test period, whereby the test period is at most equal to the shortest maximum interruption duration of all electrical loads supplied by the switchable fuse.

[0018] During the test period, the residual current flowing through the switchable fuse in the non-conducting state is recorded as the actual residual current value and compared with at least one target residual current value. The switchable fuse is assigned an initial (functional) operating state if the recorded actual residual current value is below at least one target residual current value. If at least one target residual current value is less than or equal to the recorded actual residual current value, then the switchable fuse is assigned a further operating state.

[0019] This method makes it possible to test the functional reliability of a switchable fuse without affecting the ongoing operation of the electrical load it supplies and without using a redundant structure of switchable fuses. This enables a particularly cost-effective, energy-efficient, and frequent functional testing of switchable fuses.

[0020] In one embodiment of the method, the actual residual current value is determined by measuring the discharge time of a capacitor connected in parallel to the electrical load from the supply voltage source via a resistor. The measured discharge time is then compared with at least one target value for the discharge time, which is determined for a switchable fuse in each operating state.

[0021] For example, a first target value for the discharge time in a fully functional operating state of a switchable fuse and a second target value for the discharge time in a still functional but already degraded operating state can be determined. The actual discharge time of a capacitor connected to the supply voltage source via a switchable fuse under test is then measured and compared with the first and second target values. If the measured discharge time is below the first target value, the switchable fuse under test is assigned the operating state "functional". If the measured discharge time is below the second target value but not below the first target value, the switchable fuse under test is assigned the operating state "partially functional".If the actual value of the recorded discharge time is not below the second target value, the switchable efuse being tested is assigned the operating state "not functional".

[0022] In a further development of this embodiment of the method, a measurement voltage across the capacitor is detected and compared with a reference voltage, for example by means of a comparator. The reference voltage can be provided, for example, by an adjustable reference voltage source.

[0023] A clock signal with a predetermined clock period can be counted during the time the measurement voltage is above the reference voltage and the fuse has switched to the non-conducting state. The actual discharge time can be determined from the product of the counter value and the clock signal period. For example, if the reference voltage is chosen to be sufficiently low (close to the voltage value of the measurement voltage for a fully discharged capacitor), the product of the counter value and the period can be calculated directly. To shorten the measurement time, however, it is also possible to choose a higher reference voltage and extrapolate the resulting (correspondingly lower) counter value, which would be obtained for a reference voltage close to the measurement voltage above the nearly fully discharged capacitor.

[0024] The counter required for time measurement according to this further development of the method, as well as a precise clock generator, is very easy and inexpensive to implement or may already be available. Furthermore, the counting process can be easily limited to the time interval in which the capacitor is discharged through the resistor. Counter circuits or counter modules typically have a "capture" input, which must be supplied with a positive (logically true) level and / or a rising (or falling) edge in order to trigger the counting process or capture the counter state. For example, this level and / or edge for such a "capture" input can be derived from the output voltage of a comparator, which compares the measured voltage with the reference voltage.A logical connection (for example with a signal voltage that triggers the switching of the switchable efuse from the conducting to the non-conducting state) is also very easy to implement, for example via transistor-transistor logic (TTL).

[0025] Thus, the discharge time provides a particularly simple way to determine whether the fuse is disconnecting the electrical load as intended when switching from the conductive to the non-conductive state. An actual discharge time value that exceeds a predetermined range around a corresponding target value (corresponding to an insufficiently fast capacitor discharge) indicates an excessive leakage current in the non-conductive state or a short circuit across the fuse. In this case, the operational safety of the electrical load is compromised because it is not disconnected from the supply voltage in the event of a fault (exceeding the rated current).

[0026] If the actual value of the discharge time is below this predetermined corridor, insufficient charging of the capacitor can be detected, which may be caused, for example, by the fact that the switchable efuse does not fully switch on when it is activated in the conducting state (i.e., the resistance along the efuse is too high).

[0027] In a further development of the procedure, a current temperature value and / or a current value of the supply voltage and / or a deviation of the capacitance value of the capacitor and / or a deviation of the resistance value of the resistor from the respective nominal value is determined and used to correct the actual value of the discharge time duration.

[0028] For example, the actual capacitance value of the capacitor may be lower than its nominal capacitance value, resulting in a shorter actual discharge time. To prevent a switchable fuse from being falsely classified as faulty, a corrected (corrected) target value for the discharge time is determined based on the actual (corrected) capacitance value of the capacitor. Such a correction can be obtained particularly easily and with sufficient approximation by scaling the target value for the discharge time (according to the ratio of the target value to the actual capacitance value). Of course, the actual value of the discharge time can also be scaled inversely (with equivalent computational results).

[0029] Similarly, a correction of the actual resistance value is possible if there is a deviation from the nominal resistance value.

[0030] The influence of the temperature of the switchable fuse and / or other components, as well as fluctuations in the supply voltage, can also be taken into account by computationally correcting the target value of the discharge time (or, equivalently, the actual value of the discharge time). It is possible that such fluctuations or deviations may be incorporated into the computational correction non-linearly. In this case, it will prove advantageous to include correction tables, for example, a correction table to account for the operating temperature of the fuse, in which a correction term is defined for each of a series of temperature values. This term is to be applied additively, multiplicatively, or according to another functional relationship to the recorded actual value of the discharge time once the respective temperature has been recorded. A correction table can also be created analogously for different values ​​of the supply voltage.

[0031] Such correction tables can be stored in non-volatile memory and retrieved by a test evaluation unit after the actual value of the discharge time has been recorded as described. In particular, correction values ​​to account for deviations in the actual resistance value and / or the actual capacitance value can also be stored in such non-volatile memory; these values ​​are recorded once (for each switchable fuse) in a calibration step.

[0032] This further development of the embodiment allows the use of less precise components (resistors, capacitors) and enables the method to be carried out in very different situations (with varying, highly fluctuating supply voltages and temperatures). It thus improves the robustness of the method and allows for simpler and more cost-effective implementation.

[0033] In one exemplary embodiment of the method, the test time (shutdown time), during which the switchable efuse is placed in the non-conducting state, is limited to 100 microseconds. Since electrical loads, especially those used in automobiles, have buffer memories (e.g., internal capacitances) that are sufficiently large to bridge such a short interruption without any limitations, this embodiment of the method allows efuses to be tested in a wide variety of operating situations (with a diverse range of electrical loads).

[0034] In one embodiment of the method, recorded actual residual current values ​​(for example, actual residual current values ​​determined from actual discharge duration values) are continuously stored. From this, a dynamically adjusted target residual current value is continuously determined such that the switchable fuse is assigned a further operating state (i.e., an operating state different from the "functional" state) if a currently recorded actual residual current value deviates statistically significantly from previously recorded actual residual current values. In this way, gradual processes (for example, age-related drift), which do not impair the functionality of the switchable fuse, can be distinguished from potentially faulty sudden changes. This enables the detection of faulty switchable fuses with particularly high specificity and sensitivity.

[0035] In one embodiment of the method, the switchable fuse is arranged to supply an electrical load in a vehicle. Such electrical loads (for example, control units, sensors for monitoring the vehicle's environment, or components of a powertrain) are particularly relevant to safety and reliability. Their shutdown in the event of a fault (e.g., when a rated current is exceeded) is therefore especially important.

[0036] In one embodiment of the method, a first operating state, "functional," is assigned to the switchable fuse when the actual residual current value is below a first target residual current value. A second operating state, "partially functional," is assigned when the actual residual current value lies between the first and a second target residual current value that is higher than the first target residual current value. A third operating state, "non-functional," is assigned when the actual residual current value is above the second target residual current value.

[0037] Similarly, if the actual residual current value is determined based on the actual value of the discharge time duration, a first (comparatively smaller) target value and a second (comparatively larger) target value of the discharge time duration can be used as the basis for assigning the first to third operating states.

[0038] This embodiment has the advantage that already degraded, but still functional, switchable efuses (for example, with an increased leakage current in the blocked, non-conductive state that is still within the intended tolerance range) are detected in time so that they can be replaced during scheduled maintenance work.

[0039] According to a further aspect, the invention relates to a test system configured to carry out the method described above. According to the invention, the test system comprises a test circuit, a test control unit, and a test evaluation unit.

[0040] The test circuit is set up to detect the actual residual current value flowing through the switchable efuse when it is switched to the non-conducting state.

[0041] The test control unit is designed to switch the switchable efuse from the conductive state to the non-conductive state for a predetermined test time.

[0042] The test evaluation unit is designed to compare the actual residual current value detected by the test circuit with at least one target residual current value.

[0043] The advantages of the test system correspond to those of the method that can be carried out with it according to the first aspect of the invention.

[0044] In one embodiment, the test circuit comprises an RC circuit with a resistor and a capacitor connected in parallel, as well as a comparator and a timer.

[0045] The RC circuit can be connected to a power supply via the switchable fuse. The comparator is designed to compare a measurement voltage applied across the RC circuit with a reference voltage, which is provided, for example, by a preferably adjustable reference voltage source.

[0046] The timer is connected to the test control unit and the comparator and is designed to record the time elapsed from the time the switchable efuse is switched from the conducting to the non-conducting state by the test control unit until the time the comparator determines that the measured voltage across the RC circuit reaches or falls below the reference voltage.

[0047] With such a test circuit, the actual discharge time of the capacitor across the resistor can be recorded, from which the actual residual current through the switched-off (non-conducting) fuse can be determined. Alternatively, instead of comparing the actual residual current with a target residual current, the recorded actual discharge time can also be compared with a target discharge time corresponding to the target residual current.

[0048] This embodiment thus enables a particularly simple and accurate test of the actual residual current value and therefore a test of the functionality of the switchable efuse.

[0049] In one embodiment, the test evaluation unit comprises an analog-to-digital converter (ADC) configured to detect the supply voltage and / or, in conjunction with a temperature sensor, to detect the temperature of the switchable fuse. Alternatively or additionally, according to this embodiment, the test evaluation unit comprises non-volatile memory for storing at least one calibration value and / or at least one previously recorded actual residual current value.

[0050] This embodiment has the advantage that fluctuations in the operating temperature of the fuse or the supply voltage, as well as manufacturing tolerances of the electronic components, for example, the resistor and the capacitor in the RC circuit, can be taken into account during the testing of the fuse. This enables a particularly specific and sensitive test of the fuse while simultaneously allowing for a particularly simple and cost-effective design of the test system. Further advantages correspond to those of the corresponding embodiment of the method described above according to the first aspect of the invention.

[0051] Exemplary embodiments of the invention are explained in more detail below with reference to drawings.

[0052] This shows: Fig. 1 schematically shows a test system for a switchable efuse, Fig. 2 schematically shows the time course of a test voltage during a test phase, and Fig. 3 schematically shows a flow chart for the test evaluation of a measurement on a switchable efuse.

[0053] Corresponding parts are marked with the same reference symbols in all figures.

[0054] Figure 1 Figure 100 shows, purely as an example and schematically, a test system 100 for checking the operating state of a switchable fuse 1, which is arranged between a supply voltage source 2 and an electrical load 3.

[0055] In the present embodiment, the switchable fuse 1 is configured as an n-channel metal-oxide-semiconductor field-effect transistor (NMOS FET), but other semiconductor switching element configurations are equally possible. Here, the source electrode 1.S of the NMOS FET is connected to the supply voltage source 2. The drain electrode 1.D is connected to the electrical load 3 via a diode 4. The forward current of the diode 4 points from the switchable fuse 1 to the electrical load 3.

[0056] The supply voltage source 2 can, for example, be designed as a vehicle battery of a vehicle not shown in detail, which provides a supply voltage UVThe supply voltage source 2 typically outputs between 8 and 17 volts. Alternatively or additionally, it can be part of the vehicle's electrical system. The electrical load 3 can be, for example, a control unit, active sensor, actuator, or other electrically operated subsystem of the vehicle.

[0057] The electrical load 3 must be continuously connected to the supply voltage. UV However, interruptions in the supply voltage can occur (for example, via internal capacities or other buffer storage). UV to bridge gaps for a certain period of time. Typically, supply interruptions of up to a maximum duration of approximately 100 microseconds can be bridged.

[0058] The test system 100 comprises a microcontroller 110 with external circuitry, which (typically together with the switchable fuse 1) is arranged on a printed circuit board (not shown). A timer 111, a non-volatile memory 112, and a multi-channel analog-to-digital converter (ADC) 113 are integrated on the microcontroller 110.

[0059] Furthermore, the microcontroller 110 has an execution unit for executing program instructions, which is located in Figure 1 not shown in detail. The executable program code of this execution unit comprises Efuse test software 114 and application software 115, which may be stored in a common or separate program memory of the microcontroller 110. The program memory(s) of the microcontroller 110 is / are in Figure 1not shown in detail. For example, a non-volatile memory 112 can be designed as flash memory and configured for storing both program instructions and non-volatile data.

[0060] The multi-channel ADC 113 samples the value of the supply voltage on a first channel. UV and on a second channel the value of a peripheral temperature sensor 101. The temperature sensor 101 detects the temperature of the circuit board on which the Efuse 1 and the external circuitry of the microcontroller 110 are arranged.

[0061] The timer 111 can, for example, be configured as a resettable counter that, upon a counter start, counts an internal clock signal of the microcontroller 110.

[0062] Furthermore, the microcontroller 110, controlled by the program instructions of the Efuse test software 114, provides a signal voltage USready, which is connected to the gate electrode 1.G of the switchable fuse 1. With the signal voltage US The switchable Efuse 1 can switch between a conductive state (for example, when US ≈ 5 volts) and a non-conductive state (for example, when US ≈ 0 volts) can be switched.

[0063] Between the drain electrode 1.D of the switchable fuse 1 and ground, an RC circuit 102, 103 is arranged, which is formed by a resistor 102 and a capacitor 103 connected in parallel to it and which has a time constant τ exhibits, which results from the product of the resistance value R and the capacitance of the capacitor C ( τ = R · C ) .

[0064] The measuring voltage U RC The voltage above the RC element 102, 103 is detected by a comparator 104 and compared with a reference voltage. URcompared to the voltage provided by a reference voltage source 105. The comparator output voltage UK The output of comparator 104 assumes a high level (which is a positive value, for example) UK (approximately 3 volts) when the measuring voltage U RC greater than or equal to the reference voltage UR is. The comparator output voltage UK The output of comparator 104 assumes a low level (corresponding to a comparatively smaller value, for example). UK ≈ 0 volts), if the measuring voltage U RC smaller than the reference voltage UR is.

[0065] In the embodiment according to Figure 1 The test system 100 comprises a test circuit 120, a test control unit 121 and a test evaluation unit 122.

[0066] The test circuit 120 is formed in this case by the timer 111, the RC element 102, 103 with the resistor 102 and the capacitor 103, the comparator 104 and the reference voltage source 105.

[0067] The test control unit 121 is formed here by the microcontroller 110 with the Efuse test software 114. The test evaluation unit 122 is formed here by the microcontroller 110 with the ADC 113 and the non-volatile memory 112 as well as the temperature sensor 101. However, it is also possible that the test control unit 121 and the test evaluation unit 122 are implemented independently of each other.

[0068] The functionality of test system 100 is explained below using the following examples: Figure 2 further explained, in which the schematically shows the course of the signal voltage US , the measuring voltage U RC as well as the comparator output voltage UK are shown.

[0069] At a first point in time tThe Efuse test software 114 lowers the signal voltage to 0. US such a thing (for example, from US Approximately 5 volts US (approximately 0 volts), that the switchable fuse 1 is switched from the conducting to the non-conducting state by the NMOS FET being switched off. The diode 4 prevents internal capacitances or buffer memories of the electrical load 3 from further supplying the capacitor 103 of the RC network 102, 103.

[0070] Synchronous with the reduction of the signal voltage US The Efuse test software 114 starts the timer 111.

[0071] The capacitor 103 then discharges through the resistor 102, so that the measuring voltage U RC from the value of the supply voltage UV starting in the manner of a decaying exponential function, it decreases U RC t = U V ⋅ e t 0 − t R ⋅ C , t ≥ t 0 bis sie zu a second point in time t 1 the value of the reference voltage UR falls below a certain threshold.

[0072] When the reference voltage falls below UR The comparator output voltage changes. UK from the high level (corresponding to the logical value one or "true") to the low level (corresponding to the logical value zero or "false"). This results in the following at the second time point: t 1 of the timer 111 stopped (for example, by reducing the comparator output voltage) UK (is placed on a capture input of a counter of timer 111).

[0073] The value of timer 111 (counter reading at the second time) t 1) thus represents the discharge time T, during which the capacitor 103 is discharged via the resistor 102 to the value of the reference voltage. UR was unloaded.

[0074] At a third point in time t S The Efuse test software 114 raises the signal voltage US again in this way (for example from US , ≈ 0 volts on US(approximately 5 volts), that fuse 1 switches from the non-conducting to the conducting state by turning the NMOS FET on. Thus, from the third point in time onwards, t S The electrical load 3 is supplied again by the supply voltage source 2.

[0075] The interval between the first point in time t 0 and the third point in time t S is chosen so that the shutdown duration TS , during which the electrical load 3 is not supplied, the voltage is sufficiently large for the capacitor 103 to discharge to the reference voltage. UR of comparator 104, but considerably shorter than the maximum interruption duration during which the electrical load 3 can remain without power without malfunctioning. This prevents any impairment of the operation of the electrical load 3.

[0076] Synchronous with increasing the signal voltage US at the third point in time t SThe Efuse test software 114 resets the timer 111 back to zero.

[0077] Following on from the second point in time t 1. The Efuse test software 114 evaluates the discharge time duration value T supplied by the timer 111. For example, the discharge time duration measured in this way can be T with a target value T soll a comparison can be made, which results from the above-mentioned relationship for the unloading process: U RC T soll = U R = U V ⋅ e − T Soll R ⋅ C

[0078] In various operating environments, especially in a vehicle, the supply voltage can UV The values ​​may vary considerably. Furthermore, due to manufacturing tolerances or aging processes, the resistance value R and / or the capacitance value C may deviate from their respective nominal values. Additionally, the switching behavior of a switchable fuse 1 and / or the discharge behavior of an RC circuit 102, 103 may also depend on the temperature.

[0079] It can therefore prove advantageous, when evaluating the switchable Efuse 1, to consider not only the measured discharge time T, but also at least correction values ​​for the precise determination of the resistance value R and / or the capacitance value C, and in particular the current (at the first time point) t 0 recorded) value of the supply voltage UV to be taken into account.

[0080] This can lead to changed values ​​based on the aforementioned connection. U V ′ , R', C' the supply voltage UV , of the resistance 102 and / or the capacitance of the capacitor 103 a corrected setpoint T' for comparison with the discharge time T to be determined: U RC T soll ′ = U R = U ′ V ⋅ e T soll ′ R ′ ⋅ C ′

[0081] Similarly, temperature-related variations can also be taken into account, for example by recording characteristic curves for the switching behavior of the switchable Efuse 1.

[0082] Correction values, for example a correction value Δ R = R' - R for the resistance 102 and / or a correction value Δ C = C - C' The capacitance of capacitor 103 can advantageously be recorded in a calibration step and stored on the non-volatile memory 112.

[0083] In one embodiment, the recorded values ​​of the discharge time T can also be continuously stored on the non-volatile memory 112, at certain time intervals, or for a certain past period. By comparing a currently recorded value with past values, a gradual degradation of the switchable Efuse 1 can be detected particularly early using the Efuse test software 114, especially even when its functionality is still provisionally guaranteed.

[0084] For example, the operating state of the switchable Efuse 1 can be determined as a value from the list {"functional", "functional but degraded", "not functional"} and passed from the Efuse test software 114 to the application software 115.

[0085] Based on this value, the application software 115 can trigger an error response, for example, deactivating certain functions of a vehicle controlled by the affected electrical load 3 or issuing a warning message.

[0086] Based on the in Figure 2 The schematic voltage-time curves shown illustrate the functionality of the [device] in Figure 1 The schematic circuit shown is explained in more detail. The upper part shows... Figure 2 the time course of the signal voltage US , in the middle part the time course of the measured voltage U RC as well as, in the lower part, the time course of the comparator output voltage UK along the timeline t .

[0087] At the time t = 0 (in which the supply voltage UV (is switched on or applied) the signal voltage US The voltage is switched to a high value (typically between 10 and 15 volts), which is sufficient to conduct the switchable fuse 1, designed as an NMOS FET. The measuring voltage applied across the RC network 102, 103 is adjusted accordingly. U RC approximately the value of the supply voltage UV assume (reduced only by the comparatively minor voltage drop along the source-drain path, which is typically 100 millivolts to 1 volt).

[0088] The comparator 104 compares the measured voltage U RC with the reference voltage UR The reference voltage UR is set to a value slightly above zero. The reference voltage URFor example, the permissible leakage current, which is just considered acceptable for a blocked switchable fuse 1, can be determined by calculating the resulting voltage drop across the RC network 102, 103. If this voltage drop is not exceeded, the switchable fuse 1 can be considered functional. Typically, the reference voltage is used for this purpose. UR approximately 10 to 20 percent of the supply voltage value UV be.

[0089] With the switchable fuse 1 functioning correctly (in a conducting state), the measuring voltage is therefore U RC significantly larger than the reference voltage UR . Therefore, a high positive voltage value of the comparator output voltage is present at the output of comparator 104. UK (depending on the circuit technology and the power supply of the comparator 104, typically 2 volts to 5 volts).

[0090] At a first point in time t At point 0, the Efuse test software 114, running on the microcontroller 110, triggers a brief disabling of the switchable Efuse 1 to test its functionality. For this purpose, up to a third time point... t S (i.e., for a shutdown period) TS = t S - t 0) the signal voltage US set to a low level, corresponding to a value of 0 volts or close to 0 volts.

[0091] With a functioning, switchable fuse 1, the source-drain current is completely or almost completely interrupted. Capacitor 103 then discharges through resistor 102, with the voltage across capacitor 103 (which is equal to the measured voltage) U RC is) according to the decaying exponential function already described at the beginning, through the time constant τ = R · Cof the RC element 102, 103. After a period of time designated as the discharge time T, a second time is determined. t 1 = t 0 + T the value of the reference voltage UR The threshold is undershot. Synchronously or with negligible delay, the output of comparator 104 switches from the high to the low level of the comparator output voltage. UK um.

[0092] The shutdown duration TS is sufficiently large to allow (with a functioning, switchable fuse 1) the decay of the measuring voltage U RC to be able to observe safely. It will prove advantageous to have a shutdown duration that is more than [a certain duration]. TS to provide a multiple of the discharge time T, within which, with a functioning (in particular, sufficiently good blocking) switchable fuse 1, the measuring voltage U RC from the supply voltage UV to the reference voltage UR falls off.

[0093] At the same time, the shutdown duration TS dimensioned so small that the electrical load 3 causes a failure of the supply voltage UV tolerated without impairment during this period. Such a period, during which the electrical load 3 can be safely tolerated without impairment (that is: taking into account possible tolerances of both the electrical load 3 and the supply voltage). UV as well as further tolerances) even in the event of a loss of supply voltage UV The maximum duration of interruption that can be operated is referred to here and in the following as the maximum interruption duration.

[0094] For example, starting from a known maximum interruption duration, the shutdown duration can be calculated. TS as a fraction (for example, half or a third) of the maximum interruption duration. Then the RC circuit 102, 103 and the reference voltage are URdimensioned in such a way that, with a maximum permissible leakage current through a blocked switchable fuse 1 and with a maximum permissible supply voltage UV The capacitor 103 is discharged via the resistor 102 to the value of the reference voltage during a discharge period T. UR is discharged, which is a fraction (for example, one fifth or one tenth) of the shutdown duration TS amounts.

[0095] To determine the discharge time T, for example by means of another, in Figure 1 A capture signal is generated by a comparator (not shown in detail) or by conversion to transistor-transistor logic (TTL), which assumes a high level and / or a rising edge when the signal voltage US assumes a low level and simultaneously the comparator output voltage UKThe signal remains at a high level, while otherwise it assumes a low level and / or a falling edge. Using such a capture signal, a counting process of a known clock signal can be triggered at the timer 111, namely at a high level and / or a rising edge, or stopped, namely at a low level and / or a falling edge. In this case, the counter value of the timer 111, multiplied by the period of the known clock signal, corresponds exactly to the discharge time T. Naturally, the counting process can also be stopped if the counter value of the timer 111 exceeds a maximum value. Furthermore, the application software 115 can send a start and / or stop signal to the timer 111 to trigger and / or terminate the counting process.

[0096] Alternatively, the comparator output voltage can be UKcan also be directly connected to a negated capture input of timer 111 in such a way that timer 111 generates a clock signal during the time interval between the second time point t 1 and the third point in time t S counts, while the comparator output voltage UK at the low level. In this embodiment, the discharge time T results from the difference between the precisely known shutdown time (triggered by the Efuse test software 114). TS and the measured time period T = T S − t S − t 1 .

[0097] Figure 3 This schematically and purely exemplarily shows a flowchart to explain the evaluation performed by the Efuse test software 114. The process starts at point A and ends at point E.

[0098] Starting from point A, in a first step S1 the current value of the supply voltage is determined. UVThe current temperature value, which is detected by the temperature sensor 101 (typically located near the switchable fuse 1), is also read in. Both values ​​are sampled and discretized using the ADC 113.

[0099] In a second step, S2 is at the second time point t 1 the signal voltage US The signal is lowered to a low level, thereby disabling (setting to the non-conducting state) switchable fuse 1. Simultaneously, in the second step S2, timer 111 is started.

[0100] In a subsequent first decision step E1, it is checked whether the time period recorded by the timer 111, during which the switchable fuse 1 is blocked, is the predetermined shutdown period. TS has already been reached or exceeded. In other words, it is checked whether the current time has already been reached or exceeded. t before the third point in time t S lies t < t 0 + TS = t S .

[0101] If the shutdown duration TS If the target value has not yet been reached, the subsequent third step S3 pauses the process for a predetermined waiting period, and then the first decision step E1 is executed again. The predetermined waiting period is sufficiently short to ensure that reaching the predetermined shutdown time is not a critical factor. TS can be detected with high accuracy. Preferably, the predetermined waiting time is one clock period of a clock signal supplied to and counted by the timer 111.

[0102] If, in the first decision step E1, it is determined that the predetermined shutdown duration TS Once this has been achieved, in a subsequent fourth step S4 the switchable fuse 1 is switched back to conduction (into a conducting state).

[0103] In a fifth step, S5 then reads the discharge time T, for example as the counter reading of a counter multiplied by the clock period, which in the period between the first time t 0 and the second point in time t 1 , in which the comparator output voltage UK at a high level, the signal voltage US but is at a low level, counting the clock signal.

[0104] In a subsequent sixth step S6, the read discharge time T is adjusted by using calibration data regarding the resistance value R and the capacitance value C of the RC circuit 102, 103 and / or the current (i.e.: recorded in the first step S1) value of the supply voltage. UV and / or the temperature detected by temperature sensor 101 is taken into account.

[0105] For example, the deviation ΔR from a nominal resistance value R and the deviation ΔC from a nominal capacitance value C during a (in Figure 3 The calibration step (not shown in detail) is recorded and stored in non-volatile memory 112. Additionally, characteristic curves depicting the time-switching behavior of the NMOS FET forming fuse 1 and / or characteristic curves depicting the source-drain leakage current of the switchable fuse 1 as a function of temperature may be stored in non-volatile memory 112. This data is read out in the sixth step S6 and used to adjust the read-out discharge time. T used in such a way that a corrected value for the discharge time duration T The behavior of the switchable Efuse 1 is determined independently of manufacturing tolerances, temperature, and supply voltage. UV describes.

[0106] In a subsequent second decision step E2, the discharge time corrected in this way is T compared to an initial threshold value. Is the discharge time... T Below this first threshold, the switchable fuse 1 is considered to be fully functional. In a first output step A1, an initial output code is then displayed as "functional" (which can, of course, also be encoded as a bit pattern or literal).

[0107] Is the unloading time T If the discharge time duration is not below the first threshold, the second decision step E2 is followed by a third decision step E3 to check whether the discharge time duration is within the limit. T below a second threshold that is chosen to be larger than the first threshold.

[0108] Is the unloading time TBelow this second threshold, the switchable fuse 1 is considered to have limited functionality. This could mean, for example, that the switchable fuse 1 is likely to only function for a certain remaining lifetime and should be replaced before this lifetime expires. In a second output step A2, a second output code, different from the first, indicating "limited functionality" is then output (which can, of course, also be encoded as a bit pattern or literal).

[0109] Is the unloading time T If the discharge time T is not below the first threshold, then in the second decision step E2, a third decision step E3 checks whether the discharge time T is below a second threshold that is chosen to be greater than the first threshold.

[0110] If the discharge time T is not below this second threshold, the switchable fuse 1 is considered non-functional. In a third output step A3, a third output code, different from the first and second output codes, is then output indicating "non-functional" (this code can, of course, also be encoded as a bit pattern or literal).

[0111] The threshold values ​​used in the second and third decision steps E2 and E3 can be determined, in particular, through statistical investigations. For example, values ​​for the discharge time T can be continuously recorded for a sufficiently large sample of switchable efuses 1. A threshold value for the discharge time T can then be retrospectively determined for a subset of failed switchable efuses 1 from this sample. This threshold value would have allowed a sufficiently large proportion of these failures to be detected in time without provoking an unacceptably high number of false positive detections (i.e., switchable efuses 1 marked as imminent failure based on this threshold value, but which are actually still functional).

[0112] The term "output of an output code" refers specifically to the transmission of such an output code to the application software 115. For example, if the Efuse test software 114 has determined the output code "limited functionality," the application software 115 can issue a warning and / or a notification indicating when the switchable Efuse 1 should be replaced. If the Efuse test software 114 has determined the output code "not functional," the application software 115 can shut down a safety-critical subsystem supplied by the switchable Efuse 1.

[0113] Following the first, second, and third output steps A1, A2, and A3, the value of the discharge time T is stored in non-volatile memory 112 in a seventh step S7. The endpoint E of the process is then reached.

[0114] For the sake of simplicity, it was assumed in the preceding section that steps S5, S6, and S7, decision steps E2 and E3, and output steps A1, A2, and A3 are performed after the fourth step S4, in which the switchable fuse 1 is switched back to conduction (a conducting state). In principle, it is also conceivable that the aforementioned steps could be performed from the earliest point at which the discharge time is available. T, namely from the second point in time t 1. This would have the technical advantage that the reactivation of the fuse would depend on the result of the evaluation of the discharge time duration. T It could become dependent. For example, efuses deemed non-functional might no longer be activated.

Claims

1. Test method for verifying the operating state of an efuse (1) which can be switched between at least one conductive and one non-conductive state and via which an electrical load (3) is electrically connected to a supply voltage source (2), the electrical load (3) tolerating a supply interruption for a predetermined maximum interruption duration, characterized in that - the switchable efuse (1) is switched from the conductive to the non-conductive state for a test time, TS, which is at most equal to the maximum interruption duration of the electrical load (3), - the residual current which flows through the switchable efuse (1) during the test time, TS, is recorded as the actual residual current value by means of a test circuit (120) and - is compared with at least one target residual current value, a first operating state being assigned if the actual residual current value is below the at least one target residual current value, and otherwise a further operating state being assigned.

2. Test method according to claim 1, characterized in that the actual residual current value is recorded across a resistor (102) by way of an actual value of a discharge time period, T, of the discharge of a capacitor (103) supplied by the switchable efuse (1) in parallel with the electrical load (3) and is compared with at least one target value of the discharge time period, T, assigned to a target residual current value determined for the switchable efuse (1) for one operating state in each case.

3. Test method according to claim 2, characterized in that a measuring voltage, URC, is recorded across the capacitor (103) and compared with a reference voltage, UR, and a clock signal is counted down during a time period during which the measuring voltage, URC, lies above the reference voltage, UR, the actual value of the discharge time period, T, being determined from the product of a counter reading and of a period duration of the clock signal.

4. Test method according to either of claims 2 or 3, characterized in that a current temperature value and / or a current value of a supply voltage, UV, and / or a deviation of the capacitance value of the capacitor (103) and / or of the resistance value of the resistor (102) is / are each determined from a nominal value and is / are used to correct the actual value of the discharge time period, T.

5. Test method according to any of the preceding claims, characterized in that the target residual current value is adjusted on the basis of previously recorded actual residual current values in such a way that the switchable efuse (1) is assigned a further operating state if a currently recorded actual residual current value deviates statistically significantly from previously recorded actual residual current values.

6. Test method according to any of the preceding claims, characterized in that the switchable efuse (1) is arranged to supply the electrical load (3) in a vehicle.

7. Test method according to any of the preceding claims, characterized in that a first operating state is assigned as "functional" if the actual residual current value is below a first target residual current value, a second operating state is assigned as "limited functionality" if the actual residual current value is between the first and a higher second target residual current value, and a third operating state is assigned as "non-functional" if the actual residual current value is above the second target residual current value.

8. Test system (100) for carrying out a method according to any of the preceding claims, comprising a test circuit (120), a test control unit (121) and a test evaluation unit (122), characterized in that - the test circuit (120) is configured to record the actual residual current value which flows through the switchable efuse (1) in the non-conductive state, - the test control unit (121) is configured to switch the switchable efuse (1) between the conductive and the non-conductive state for the predetermined test time, TS, and - the test evaluation unit (122) is configured to compare the actual residual current value recorded by the test circuit (120) with at least one target residual current value.

9. Test system (100) according to claim 8, characterized in that the test circuit (120) comprises an RC member (102, 103) comprising a resistor (102) and a capacitor (103) connected in parallel therewith, a comparator (104) and a timer (111), - it being possible for the RC member (102, 103) to be connected to a supply voltage source (2) via the switchable efuse (1), - the comparator (104) being configured to compare the measuring voltage, URC, applied above the RC member (102, 103) with a reference voltage, UR, and - the timer (111) being started by the test control unit (121) with the switching of the switchable efuse (1) from the conductive into the non-conductive state and being stopped by the comparator (104) if the measuring voltage, URC, reaches or falls below the reference voltage, UR.

10. Test system (100) according to claim 7 or claim 8, characterized in that the test evaluation unit (122) comprises - an analog-to-digital converter (113), o which is configured to record an actual value of the supply voltage, UV, and / or o which is configured to record the temperature of the switchable efuse (1) by means of a temperature sensor (101), - and / or a non-volatile memory (112) for storing at least one calibration value and / or at least one actual residual current value.