Method for determining the state of an inverter

Abstract

The invention relates to a method (74) for determining a state of an inverter (2) on the basis of a plurality of sensor values (94), each of which corresponds to a measured value (50) created by means of an analogue sensor (38) of a sensor unit (18). A clock signal (76), which has a plurality of identical clock pulses (80) having two different clock states (78), is sent to each sensor unit (18). During one of the clock states (78) of each clock pulse (80), a predefined reference value (84) for each sensor unit (18) is changed by the current measured value (50) to an intermediate value (56) by means of a predefined operation. During the other clock state (78) of each identical clock pulse (80), the intermediate value (86) for each sensor unit (18) is changed by an auxiliary value (90) up to the reference value (84) by means of an opposite operation. During each remaining time span of said identical clock pulse (80), a first state (96) and otherwise a second state (98) is used as the relevant sensor value (94), and the sensor values (94) are supplied by means of a supply circuit (102) into a common output line (36) that is directed towards an output point (34). The invention also relates to an inverter (2).

Description

[0001] Description

[0002] Method for determining a state of an inverter

[0003] The invention relates to a method for determining the state of a converter and to a converter. The converter has several sensor units, each with an analog sensor.

[0004] Inverters typically contain several electrical and / or electronic components that heat up during operation due to electrical losses.

[0005] If a critical temperature is exceeded, the individual components will be damaged and must be replaced. To prevent this, multiple temperature sensors are typically installed, and when the temperature is reached, reduced by a safety threshold, the converter shuts down.

[0006] Such converters are increasingly comprising individual modules, each containing one or more power semiconductor switches, by means of which a comparatively high electrical voltage, namely the intermediate circuit voltage, is switched. These modules facilitate the assembly of the converter. The control and / or regulation of the converter, which also operates the power semiconductor switches, is typically carried out by a control unit. The control unit of the converter is usually galvanically isolated from the voltages switched by the power semiconductor switches, thus enabling safe operation for people. This also reduces the requirements for the components used to implement the control unit.

[0007] The modules usually contain a temperature sensor or at least a temperature-dependent resistor, which can be used to determine the temperature of the module and thus also of the power semiconductor switch. The resistor is usually not reliably galvanically isolated from the electrical voltages carried by the respective module. For the control unit to further process the electrical resistance provided in this way, or an electrical voltage resulting from the resistance, galvanic isolation is therefore necessary.

[0008] Since the resulting electrical voltage or resistance is an analog signal, analog galvanic isolation is required for each module, which is comparatively cost-intensive. An alternative is to digitize the respective value using an isolating delta-sigma modulator assigned to each module, which feeds the signal created in this way into an output line. This requires a comparatively cost-intensive delta-sigma modulator for each module. The signals from the different modules are not synchronized with one another, which is why at least four wires must be assigned to the output line for each module. This increases cabling and therefore manufacturing costs.

[0009] In an alternative embodiment, the resulting electrical voltages or the respective resistance are transmitted via a respective assigned line to a common digitization circuit, which performs digitization and, if necessary, minor evaluation. The signals generated by the digitization circuit are transmitted to the controller via a galvanic isolator, such as an optocoupler. This reduces the number of components required for digitization and galvanic isolation. However, a comparatively large number of, and possibly long, cables are required between the modules and the digitization circuit. These cables are not galvanically isolated from the comparatively high intermediate circuit voltage, which therefore places special demands on touch protection. This increases complexity and manufacturing costs.The invention is based on the object of specifying a particularly suitable method for determining a state of a converter and a particularly suitable converter, wherein safety is advantageously increased and manufacturing costs are reduced, and wherein further processing is expediently facilitated.

[0010] With regard to the method, this object is achieved according to the invention by the features of claim 1 and with regard to the converter by the features of claim 8. Advantageous further developments and refinements are the subject of the respective subclaims.

[0011] The method is used to determine the state of a converter. The converter is, for example, an inverter, a rectifier or a DC / DC converter. The electrical power provided by the converter during operation is preferably between 200 W and 500 kW. The converter is, for example, a component of an industrial plant and is used, for example, to operate an actuator by means of which a workpiece is created and / or processed. Alternatively, the converter is used to supply current to an electric motor by means of which a vehicle is driven, such as a motor vehicle, for example a passenger car (car) or a commercial vehicle, such as a lorry (truck) or bus. Alternatively, the vehicle is a snow groomer, a construction machine or an agricultural implement.

[0012] In particular, the converter has a plurality of power semiconductor switches, wherein the state of the converter is determined in particular based on the state of the power semiconductor switches or is at least related thereto. The power semiconductor switches are preferably present in modules, wherein, for example, two power semiconductor switches are always combined to form such a module. In particular, a bridge circuit, preferably a B6 circuit, is provided by means of the power semiconductor switches. Suitably, three modules are present, wherein each module provides a bridge branch. The converter has a plurality of sensor units, each comprising an analog sensor. For example, all sensors are identical to one another and thus also the analog sensors. Alternatively, the sensor unit or at least the analog sensors differ.However, it is particularly preferred that the same physical quantity is measured using the analog sensors. Suitably, each module is assigned a corresponding analog sensor, so that a B6 circuit preferably has three sensors and thus three sensor units.

[0013] Preferably, each analog sensor is a temperature sensor, so that by means of it, in particular, a temperature of the respective module or other component of the converter can be measured. Preferably, the state of the converter to be determined corresponds to a temperature of the converter, wherein the state corresponds, for example, to the temperature, or the state has, for example, only two states, such as operation at operating temperature (normal temperature range) and overheating of the converter. Preferably, each analog sensor has a temperature-dependent resistor, which is expediently interconnected by means of further components such that an electrical voltage is provided by means of each analog sensor, which voltage is expediently functionally related to the respective current resistance of the temperature-dependent resistor and thus to the temperature acting on it.For example, temperature and electrical voltage are directly proportional to each other; a linear relationship exists between them. In particular, the electrical voltage provided by the analog sensor is an analog signal that does not have different discrete levels. The analog sensor is advantageously optimized to be relatively / sufficiently accurate within a temperature range between 60°C and 100°C.

[0014] In summary, the analog sensor generates a measured value during operation that is analog and, for example, directly measured. However, preprocessing is preferably performed using the analog sensor, for example, converting a changing resistance due to physical effects to be measured into an associated electrical voltage. Alternatively, or in combination with this, amplification occurs, facilitating further processing. The value generated in this way corresponds to the measured value provided by the respective analog sensor.

[0015] The method provides that a clock signal is first sent to the sensor units. The clock signal is provided in particular by means of a clock generator and has several identical clock pulses, which thus do not differ and which expediently follow one another directly in time. Each clock pulse has two different clock states. For example, each clock pulse has only exactly the two different clock states, so that the clock signal is binary. For example, each clock pulse is evenly distributed between the two clock states. However, it is particularly preferred for at least one of the clock states to be longer in time, preferably the later of the two clock states.

[0016] According to the method, during one of the clock states of each clock, a predefined reference value is changed by the respective current measured value, which is provided by the respective analog sensor, for each sensor unit by means of a predefined operation to a respective intermediate value. In other words, as long as one of the clock states lasts, the predefined operations are carried out one after the other so that the initially predefined reference value is changed to the intermediate value. The reference value is, for example, the same for all sensor units and, for example, predefined centrally, or particularly preferably it is predefined separately for each sensor unit so that the reference values ​​can differ between the individual sensor units, for example due to manufacturing tolerances. This reduces manufacturing costs.In particular, the operation is applied with a specific time constant, which is conveniently the same for all sensor units.

[0017] In summary, the operation is applied multiple times to each sensor unit during one clock state of each clock cycle, so that the respective reference value is successively changed based on the current measured value, which is determined by the respective analog sensor. After the clock state has expired, the intermediate values ​​are available, which differ between the individual sensor units due to the different measured values. However, since the clock signal is predetermined in time for all sensor units, the intermediate values ​​of the different sensor units are always reached at the same time.

[0018] In a subsequent work step, during the other clock state of the same clock cycle, the respective intermediate value for each sensor unit is changed by a respective auxiliary value using an opposite operation until it reaches the reference value. The opposite operation is, for example, the inverse operation to the operation by which the reference value was changed to the intermediate value. Once the reference value has been reached, the opposite operation is no longer performed; instead, the reference value is no longer changed, i.e., the intermediate value is no longer changed. In particular, the auxiliary value is constant. The auxiliary value is, for example, specified externally for all sensor units, or expediently created by each sensor unit itself, whereby these preferably differ only relatively slightly between individual sensor units.At least for each sensor unit, the auxiliary value has a fixed ratio to the reference value, with the components of the sensor unit being designed in particular such that the ratio preferably exhibits only slight fluctuations among the individual sensor units. In particular, the interconnection of the sensor units and / or the components used to specify the auxiliary value has low error tolerances, thus increasing the comparability of the sensor values ​​to be generated between the sensor units.

[0019] Changing the reference value to the intermediate value and the intermediate value to the reference value is done primarily using analog technology. In other words, this takes place in the analog domain, and neither the auxiliary value nor the intermediate value nor the reference value are present as digital signals. After the reference value has been reached again, a first state is used as the respective sensor value for the remaining time period of the same clock cycle. Otherwise, a second state is used as the respective sensor value for the remaining time period of the same clock cycle. In other words, the sensor value thus only has two different states, and the determination is carried out, for example, using a comparator. In summary, during each clock cycle, the second state is used as the sensor value as long as the reference value is not present, and otherwise the first state. Consequently, each sensor value is present as a binary signal.Conveniently, the different states correspond to different electrical potentials, simplifying further processing. Alternatively, or in combination, the reference value, the intermediate value, and / or the auxiliary value each represent different electrical potentials.

[0020] The sensor values ​​created in this way are fed, in particular immediately, i.e. already during the respective cycle, into a common output line led to an output point by means of a feed-in circuit of the respective sensor unit. In other words, the binary sensor values ​​created in this way are fed into the output line so that a binary signal is now present at the output point, which simplifies further processing. Each of the sensor values ​​corresponds to information about the state of the converter, so that the state of the converter is determined by means of the method. The output point is in particular a component of a controller of the converter, by means of which, for example, an output power is regulated and / or controlled. In particular, the controller is suitable, in particular intended and configured, to be operated by a person.It is advisable to further evaluate the status by means of the control system and / or to adapt (other) operation of the converter depending on this.

[0021] In this method, the common clock signal is used for the different sensor units, and it is only necessary that the auxiliary values ​​of the different sensor units and / or their relationship to the respective reference value differ only relatively slightly. Thus, only those components with relatively low manufacturing tolerances are required. For the other components, comparatively large error tolerances can be selected without this having an impact on the method. This reduces manufacturing costs. Furthermore, since the binary sensor values ​​or signals based on them are fed into the output line via the feed circuit, further processing is simplified.Since the sensor values ​​only have two different states, it is possible to choose comparatively low electrical potentials representing the different states, thus increasing safety.

[0022] For example, all sensor values ​​are fed separately into the output line, with each sensor value being assigned a wire of the output line. In other words, multiple values ​​are present at the output point, each of which corresponds to one of the measured values. This enables an evaluation and / or comparatively accurate determination of the current state of the converter. Alternatively, for example, only a limited number of wires are present, and the output line is operated like a bus system. Due to the predetermined clock signal, which is the same for all sensor units, a coordinated feeding of the different sensor values ​​is possible, thus further reducing manufacturing costs.

[0023] However, it is particularly preferred to transmit only one extreme value of the sensor readings to the output point. This simplifies further processing.

[0024] In particular, the extreme value is used as the state of the converter, and / or the extreme value corresponds to a specific state of the converter, such as overheating or operation within a normal temperature range. For example, the extreme value is the sensor value that most closely approaches a predetermined value. In particular, the extreme value is the minimum or maximum of the sensor values. If the sensor values ​​correspond to the current temperature, the extreme value preferably corresponds to the maximum. In particular, the feed-in circuits have a logic circuit for transmitting only the extreme value, and by means of this logic circuit, for example, a corresponding feed is made in coordination with the other sensor values. Preferably, the sensor units are connected to the output line with an "and" connection or an "or" connection.Conveniently, all sensor values ​​are fed into the output line simultaneously, resulting in a superposition of these values. The feed-in is performed in such a way that, due to the superposition, only the extreme value is available at the output point. This further simplifies wiring, and the feed-in circuits can be created comparatively inexpensively.

[0025] For example, the feed circuits are each designed like a Schmitt trigger. However, it is particularly preferred for each feed circuit to have a switching element that is actuated based on the respective sensor value. In other words, the switching element of the respective sensor unit is actuated based on the respective sensor value. The switching elements of all sensor units are electrically connected in parallel between a reference potential, which is, for example, ground, and the output point. Since the sensor values ​​are binary signals, either the reference potential or a different electrical potential is present at the output point. Due to such a connection, depending on the arrangement, the output point for the first state or the second state of the respective sensor value is connected to the reference potential, with the temporal length of this corresponding to the measured value / sensor value that represents the extreme value.In particular, the highest temperature measured by the analog sensor can be derived from the time the output point is connected to the reference potential and is functionally related to it.

[0026] Particularly preferably, the respective sensor value is transmitted to the switching element in a galvanically isolated manner. In particular, an optocoupler is used for this purpose, and the switching element is, for example, a component of the optocoupler. As a result, the output point and also the output line are galvanically isolated from the other electrical voltages carried by the respective sensor unit and in particular from the analog sensor, thus increasing safety. It is also possible, for example, to position the respective analog sensor relatively close to any power semiconductor switches, thus increasing measurement accuracy while still increasing safety. Furthermore, the requirements for the output line are reduced due to the galvanic isolation, allowing the use of comparatively cost-effective components, thus further reducing manufacturing costs.

[0027] In a particularly preferred development, the clock signal is received by means of a galvanic isolator of the respective sensor unit. In other words, the clock signal is galvanically isolated from the other components of the sensor unit, in particular from the analog sensor, by means of the galvanic isolator of the respective sensor unit. The clock signal is preferably provided by means of a clock generator, which is, for example, a component of a controller of the converter. This simplifies the design. Due to the respective galvanic isolator, it is possible to operate the controller at a separate electrical potential with respect to the sensor units or any modules assigned to the sensor units, which increases safety. The requirements for the line for transmitting the clock signal to the sensor units are also reduced, which reduces manufacturing costs.

[0028] For example, the line providing the clock signal and the output line are implemented using a common cable, with the sensor values ​​also being expediently fed into the output line in a galvanically isolated manner, so that the respective sensor values ​​are transmitted to the respective switching element in a galvanically isolated manner. This further increases reliability and reduces manufacturing costs. Furthermore, it is possible to arrange the analog sensors at a relatively large distance from any control system. For example, each of the sensor units has one or more voltage sources, with one of the voltage sources providing the reference value and the other the auxiliary value. This increases flexibility.Particularly preferably, however, each sensor unit comprises only a single voltage source, based on which the respective reference value and the respective auxiliary value are created, wherein the reference value and the auxiliary value are provided by means of a common voltage divider. In other words, the voltage divider is additionally present, which has a plurality of ohmic resistors connected electrically in series, between which a tap is formed, wherein one of the taps is assigned to the reference value and the other to the auxiliary value. In the method, it is not the absolute values ​​of the reference value and the auxiliary value that are decisive, but their relationship to one another. In this development, the ratio is specified by means of the voltage divider and thus based on the ohmic resistors. Such ohmic resistors are comparatively inexpensive with low manufacturing tolerances, which is why manufacturing costs are further reduced.In other words, no separate voltage sources and / or capacitors are required.

[0029] Particularly preferred is negative integration as the operation by which the reference value is changed to the intermediate value, and integration as the opposite operation. It is advisable to use the same time constant for negative integration and integration. Alternatively, different time constants are assigned. This choice of operation ensures that the reference value or the intermediate value are not changed excessively, thus preventing malfunctions. Within one clock state, the electrical voltage, when used for the reference value and the measured value, drops comparatively quickly for a large measured value. In other words, the rate of decrease depends on the respective measured value.The intermediate value realized in each case, however, always occurs with the same gradient, namely around the auxiliary value, whereby the length of the time window required to reach the reference value again and thus the length of the remaining time period depend on the respective measured value / intermediate value realized in each case.

[0030] For integration and negative integration, a suitable operational amplifier is used, which is interconnected accordingly. Using these operations simplifies the interconnection of the sensor units, and they can be implemented using comparatively inexpensive components. Furthermore, it is possible to connect the operational amplifier used for integration in such a way that a limit to the reference value is achieved, thus further simplifying the interconnection. In particular, the two operational amplifiers are interconnected appropriately.

[0031] The converter comprises a plurality of sensor units, each having an analog sensor, and an output line led to an output point. Preferably, the output line is led to a controller of the converter, and the output point is expediently a component of the controller. Suitably, the controller queries the electrical potential applied to the output line. The converter is operated according to a method for determining a state of a converter. According to the method, a clock signal having a plurality of identical clock pulses with two different clock pulse states is sent to each sensor unit, and during one of the clock pulse states of each clock pulse, a predetermined reference value is changed at each sensor unit by the respective current measured value to a respective intermediate value using a predetermined operation.During the other clock state of the same clock, the respective intermediate value of each sensor unit is changed by a respective auxiliary value using an opposite operation up to the reference value. During the respective remaining time period of the same clock, a first state and otherwise a second state is used as the respective sensor value. The sensor values ​​are fed into a common output line connected to an output point by means of a feed circuit. The sensor units are preferably identical in construction and / or serve in particular to determine a temperature. For this purpose, each analog sensor suitably has a temperature-dependent resistor which is connected, for example, by means of an operational amplifier such that an electrical voltage is provided by each analog sensor, said voltage being dependent on the temperature acting on the temperature-dependent resistor.Suitably, the converter, preferably the controller, has a clock generator by means of which the clock signal is provided during operation.

[0032] In particular, the converter comprises several modules, each of which is assigned one of the analog sensors. Each module expediently comprises two power semiconductor switches electrically connected in series, thereby forming a bridge arm. The converter preferably comprises a DC link, which, during operation, preferably carries a DC voltage greater than 200 V, and the modules are electrically connected in parallel between the two electrical potentials.

[0033] In particular, an operational amplifier is used to perform the operation and the opposite operation, and inverse integration, for example, is used as the operation. In particular, the interconnection of the two operational amplifiers ensures limitation to the reference value, for which purpose a diode is preferably used, by means of which, in particular, the output of one of the operational amplifiers, hereinafter referred to as the first operational amplifier, is connected to the inverting input of the remaining operational amplifier, hereinafter referred to as the second operational amplifier. The output of the second operational amplifier is preferably also connected to its inverting input via a capacitance, such as a capacitor.In addition, its output is connected to the non-inverted input of the first operational amplifier, by means of which the integration takes place, whereas the second operational amplifier is used for the negative integration.

[0034] Furthermore, the invention relates to a sensor unit for carrying out the method.

[0035] In this case, a part of the method is carried out by means of the sensor unit. The sensor unit comprises an analog sensor, by means of which, for example, an electrical voltage is provided that is proportional to a physically realized value. In particular, the analog sensor is suitable for measuring a temperature, and in particular is provided and configured for this purpose. Preferably, the analog sensor comprises a temperature-dependent resistor, which is connected in particular by means of an operational amplifier such that an electrical voltage is provided by the analog sensor during operation, wherein the electrical voltage is dependent, in particular proportional, to a prevailing temperature.

[0036] The further developments and advantages explained in connection with the process can also be transferred to the converter / sensor unit and to each other and vice versa.

[0037] In the following, exemplary embodiments of the invention are explained in more detail with reference to a drawing. In the drawings:

[0038] Fig. 1 shows a schematic diagram of a converter having several sensor units, Fig. 2 shows a circuit of the sensor units,

[0039] Fig. 3 shows a circuit diagram of one of the identical sensor units, Fig. 4 shows a method for determining a state of the converter,

[0040] Fig. 5 the time course of a measured value, a sensor value and a

[0041] clock signal, and

[0042] Fig. 6 according to Fig. 3 an alternative embodiment of the sensor unit.

[0043] Corresponding parts are provided with the same reference numerals in all figures.

[0044] Figure 1 shows a simplified schematic of a converter 2 in the form of an inverter. The converter 2 has a DC voltage connection 4, which has two different connections 6, between which an intermediate circuit voltage 8 of 400 V is present during operation. The two connections 6 are each electrically connected to one another via a module 10, which is electrically connected in parallel between the connections 6.

[0045] The modules 10 each have two power semiconductor switches 12 electrically connected in series, so that each module 10 provides a bridge branch, and the converter 2 has a bridge circuit, namely a B6 circuit. Between the power semiconductor switches 12 of each module 10, a tap is formed, which is connected to an AC voltage connection 14 of the converter 2. The power semiconductor switches 12 are actuated by means of a driver circuit (not shown in detail), which in turn is actuated by a controller 16, which regulates the converter 2 depending on external specifications. The controller 16 also monitors the operating parameters of the converter 2, and the converter is shut down in the event of overheating, namely when the operating temperature of the power semiconductor switches 12 exceeds a limit value, such as 90°C.

[0046] The converter 2 has three identical sensor units 18, each of which is integrated into one of the modules 10. The sensor units 18 are each a component of a circuit board or a part of a circuit board to which, for example, one or both power semiconductor switches 12 are also attached. The sensor units 18 are spaced apart from one another.

[0047] Figure 2 shows the interconnection of the sensor units 18. Each sensor unit 18 has a galvanic isolator 20 in the form of a first optocoupler, which is electrically contacted with a line 22 and is operated by means of the latter. In other words, a signal is provided via the line 22, based on which signal the respective galvanic isolator 20 is operated, so that the rest of the sensor unit 18 is galvanically isolated from the line 22. The galvanic isolators 20 are electrically connected in parallel between two wires of the line 20. The line 22 is operated by a clock generator 24, which is a component of the controller 16. Each sensor unit 18 further has a second optocoupler 26, which, like the first optocouplers, each has a switching element 28 that is operated by means of a light-emitting diode 30 of the second optocoupler 26.The switching elements 28 are electrically connected in parallel to one another and between a reference potential 32, namely ground, and an output point 34, for which an output line 36 is used, which thus comprises two wires. In summary, the output line 36 is routed to the output point 34, which is a component of the controller 16, by means of which the electrical potential present at the output point 34 is interrogated during operation.

[0048] The output line 36 and the line 22 are implemented using a common cable, whereby the lines and also the controller 16 are galvanically isolated from the other electrical potentials carried by the rest of the sensor units 18 and also by the modules 10 thanks to the galvanic isolators 20 and the second optocouplers 26. Thus, the requirements for the cable and the controller 16 are reduced, even if parts of the sensor units 18 are not galvanically isolated from the electrical voltages carried by the power semiconductor switches 12.

[0049] Figure 3 shows a circuit diagram of one of the identical sensor units 18. Each sensor unit 18 has an analog sensor 38, each of which includes a temperature-dependent resistor 40. This resistor is connected in parallel to the middle of three first auxiliary resistors 42 connected in series. These are ohmic resistors whose resistance values ​​can differ. The series connection of the first auxiliary resistors 42 is connected between two electrical potentials, in this example between the reference potential 32 and an auxiliary electrical potential 44, which is specified by a voltage source (not shown in detail).

[0050] The non-inverting input of a further operational amplifier 46 is connected to one of the terminals of the temperature-dependent resistor 40, which is connected to two second auxiliary resistors 48 to form a non-inverting amplifier circuit. For this purpose, the inverting input of the further operational amplifier 46 is connected to the center tap of the electrical series connection of the second auxiliary resistors 48, which is connected between the output of the further operational amplifier 46 and the reference potential 32. Due to this connection, a measured value 50 is present at the output of the further operational amplifier 46 during operation, namely an electrical potential that is functionally related to the current value of the electrical resistance of the temperature-dependent resistor 40 and thus to the temperature acting on it.The values ​​of the auxiliary resistors 42, 48 are selected such that essentially a linear amplification / conversion occurs in a temperature range between 60°C and 100°C. Alternatively, or in combination with this, the values ​​of the auxiliary resistors 42, 48 are selected such that the change in the measured value 50 is as large as possible between 60°C and 100°C. The measured value 50 is an analog value, so the sensor is an analog sensor 38.

[0051] The output of the further operational amplifier 46 and thus the measured value 50 are connected to the galvanic isolator 20, namely its associated switching element 52, which conducts electrical current when an electrical voltage is applied to the LED 54 of the galvanic isolator 20, which is provided by the clock generator 24. The switching element 52 is connected directly to two further resistors 56, one of which is connected to the reference potential 32. The remaining further resistor 56 is connected to the inverting input of a second operational amplifier 58 and, via a capacitor 60, to the output of the second operational amplifier 58. The output of the second operational amplifier 58 is further connected to the non-inverting input of a first operational amplifier 62, the output of which is connected to the inverting input of the second operational amplifier 58 via a diode 64.

[0052] Each sensor unit 18 further includes a voltage divider 66 comprising three resistors 68 connected in series, and connected between the reference potential 32 and the auxiliary potential 44. The non-inverting input of the second operational amplifier is connected to one of the center taps of the voltage divider 66, and the inverting input of the first operational amplifier 20 is connected to the remaining center tap.

[0053] The output of the first operational amplifier 62 is connected to the non-inverting input of an additional operational amplifier 70, which is connected as a voltage follower. For this purpose, the output of the additional operational amplifier 70 is connected to its inverting input. The output of the additional operational amplifier 70 is also connected to the LED 30 of the respective second optocoupler 26, which in turn is connected to the reference potential 32 via an additional resistor 72. The values ​​and configurations of the additional resistors 56, the resistors 68, and the additional resistor 72 can vary and are adapted in particular to the respective operational amplifiers 58, 62, 70 used.

[0054] Figure 4 illustrates a method 74 for determining a state of the converter 2, which is performed using the sensor units 18 of the controller 16 and other components of the converter 2, such as the line 22 and the output line 36. Thus, the converter 2 is operated according to the method 74. In a first step 75, a clock signal 76 is provided by the clock generator 24 and transmitted to each of the sensor units 18 via the line 20.

[0055] The clock signal 76, whose temporal profile is shown in Figure 5, has two different clock states 78, between which the clock signal 76 alternates. The clock signal 76 is divided into successive clock states 80, which are similar to one another and which directly follow one another. Each clock 80 begins by assuming one of the clock states 78. During the course of the respective clock 80, this is changed to the other clock state 78, and the clock 80 ends with the renewed change to the original clock state 78. Thus, the clock signal 76 is a binary signal and, in particular, a square-wave signal. The two clock states 78 are realized using different electrical potentials.The clock signal 76 is received essentially simultaneously by means of the galvanic isolator 20 of each sensor unit 18, and is configured such that upon receiving the first clock state 78 of each clock 80, the respective switching element 52 of the galvanic isolator 20 is electrically conductive, so that the respective measured value 50 is passed to the second operational amplifier 58.

[0056] By means of this, a second work step 82 is carried out in which a reference value 84, which is predetermined and provided by the voltage divider 66, is reduced by the respective measured value 50 to a respective intermediate value 86 by means of negative integration, as also shown in Figure 5. Thus, the reference value 84 is reduced at the beginning of each clock 80 for the duration of one of the clock states 78, as long as this clock state 78 lasts. When this clock state 78 ends, the respective intermediate value 86 is reached.

[0057] Here, as the measured value 50 increases, i.e., as temperatures increase, the respective intermediate value 56 is further reduced, with the gradient being steeper due to the constant duration of these clock states 78. In summary, during one of the clock states 78 of each clock 80, the predetermined reference value 84 for each sensor unit 18 is changed by the respective current measured value 50 by means of a predetermined operation, namely by means of inverse integration, to the respective intermediate value 86. The time constant for negative integration is determined based on one of the further resistors 56 and the capacitance 60. In particular, the first and second work steps 75, 82 are performed simultaneously.

[0058] As soon as the other clock state 78 is assumed by the clock signal 76, a third operating step 86 is performed. In this step, the switching element 52 of the galvanic isolator 20 is no longer current-carrying, so that initially no further operation of the second operational amplifier 58 takes place. However, the first operational amplifier 62 changes the respective intermediate value 86 by means of the opposite operation, namely by integrating by a respective auxiliary value 90, until the reference value 84 is reached again. As soon as the reference value 84 has been reached, a further increase is prevented by means of the diode 64, so that the reference value 84 represents the maximum of the integration during each clock cycle 80.

[0059] In summary, in method 74, negative integration is used as the operation to change the reference value 84 and integration is used as the opposite operation to change the intermediate value 86, for which the two operational amplifiers 58, 62 are used. The time at which the reference value 84 is reached depends on the intermediate value 86 and thus on the respective measured value 50 due to the constant auxiliary value 90. The time is also dependent on the ratio of the reference value 84 to the respective auxiliary value 90, which is specified by means of the voltage divider 66, since the reference value 84 and the auxiliary value 90 are provided by means of the common voltage divider 66. However, the time is independent of the level of the auxiliary potential.Thus, using all sensor units 18, provided the resistors 68 between the individual sensor units 18 have only small manufacturing tolerances, the reference value 84 is always reached again at the same time for the same measured value 50. The other components of the sensor units 18, however, can be selected with comparatively high manufacturing tolerances, which is why the manufacturing costs of the sensor unit 18 are comparatively low.

[0060] In a fourth work step 92, which is carried out essentially identically to the second and third work steps 82, 88, a sensor value 94 is created by means of each sensor unit 18, for which the additional operational amplifier 70 is also used. The respective sensor value 94, the temporal profile of which is shown in Figure 5, has a first state 96 and a second state 98, so that the sensor value 94 is binary. Here, after the reference value 84 has been reached again in the respective clock cycle 80, the first state 96 is used for the sensor value 94 and otherwise the second state 98. Thus, the second state 98 is used during integration and negative integration.In summary, the sensor value 94 is thus synchronous with the clock pulses 80, although the length of time at which the different states 78, 96, 98 are assumed do not match each other. However, the sensor value 94 also changes between the two states 96, 98 once during each clock pulse 80 and at its end.

[0061] As the measured value 50 increases, the time period that the first state 96 is assumed during each clock cycle 80 is shortened, and at the beginning of each clock cycle 80, the sensor value 94 is equal to the second state 98. In summary, there is thus a corresponding sensor value 94 for each of the sensor units 18, and these sensor values ​​94 are binary signals.

[0062] In a subsequent fifth work step 100, the respective sensor value 94 is fed into the output line 36 by means of each sensor unit 18, namely a respective feed circuit 102 comprising the respective second optocoupler 26. When the respective sensor value 94 has the second state 98, the LED 30 of the respective second optocoupler 26 is controlled such that the switching element 28 of the second optocoupler 26 is electrically conductive. Thus, the respective sensor value 94 is transmitted to the switching element 28 in a galvanically isolated manner. In summary, the switching element 28 of the respective feed circuit 102 of the respective sensor unit 18 is actuated based on the respective sensor value 94. The fourth and fifth work steps 92, 100 preferably take place simultaneously.

[0063] In addition, the output point 34 is connected to the reference potential 32 by means of the switching elements 28 of the second optocouplers 26, and this continues as long as at least one of the sensor values ​​94 has the second state 98. Consequently, the electrical potential at the output point 34 is determined only by means of the sensor value 94 for which the second state 98 is assumed for the longest time, the length of which corresponds to the respective measured value 50.

[0064] In summary, only the sensor value 94 determines the electrical potential at output point 34, which represents the maximum of all sensor values ​​94, i.e., an extreme value, and which corresponds to the highest measured value 50. Consequently, only the extreme value of the sensor values ​​94 is transmitted to output point 34, i.e., the highest temperature. As long as this is below a critical value, such as 90°, the converter 2 is operated unchanged by the controller 16. If, however, the critical value is exceeded, the condition of the converter 2 is assumed to be overheating, and the converter 2 is shut down or at least operated at reduced power. Consequently, the condition of the converter 2 is determined using method 74.

[0065] Figure 6 shows a modification of one of the sensor units 18, wherein the analog sensor 38 with the temperature-dependent resistor 40 remains unchanged. The other components of the analog sensor 38, such as the additional operational amplifier 46 and the auxiliary resistors 42, 48, are still present, but are not explicitly shown. The galvanic isolator 20 is modified, and it continues to actuate the switching element 52, which, however, is not a component of the galvanic isolator 20 but separate from it. The switching element 52 is also connected to a further voltage source 104.

[0066] The connection of the first and second operational amplifiers 62, 58 with one of the additional resistors 56, the capacitor 60, and the diode 64 remains unchanged. However, the reference value 84 is now provided via ground, and the auxiliary value 90 is provided via an additional voltage source 106.

[0067] The further and additional voltage sources 104, 106 are DC voltage sources. The output of the first operational amplifier 62 is connected to the second optocoupler 26, which actuates the switching element 28 (not shown in detail), which is separate from the second optocoupler 26.

[0068] The invention is not limited to the exemplary embodiments described above. Rather, other variants of the invention can also be derived therefrom by those skilled in the art without departing from the subject matter of the invention. In particular, all individual features described in connection with the individual exemplary embodiments can also be combined with one another in other ways without departing from the subject matter of the invention. List of reference symbols

[0069] 2 inverters

[0070] 4 DC voltage connection

[0071] 6 Connection

[0072] 8 DC link voltage

[0073] 10 Module

[0074] 12 power semiconductor switches

[0075] 14 AC voltage connection

[0076] 16 Control

[0077] 18 Sensor unit

[0078] 20 galvanic isolators

[0079] 22 Line

[0080] 24 clocks

[0081] 26 second optocoupler

[0082] 28 Switching element

[0083] 30 LEDs

[0084] 32 Reference potential

[0085] 34 Output point

[0086] 36 Output line

[0087] 38 analog sensor

[0088] 40 temperature-dependent resistance

[0089] 42 first auxiliary resistor

[0090] 44 Auxiliary potential

[0091] 46 additional operational amplifiers

[0092] 48 second auxiliary resistor

[0093] 50 measured values

[0094] 52 switching element

[0095] 54 LEDs

[0096] 56 further resistance

[0097] 58 second operational amplifier

[0098] 60 capacity

[0099] 62 first operational amplifier diode

[0100] voltage divider

[0101] resistance additional operational amplifier additional resistance

[0102] Procedure first step

[0103] Clock signal

[0104] Clock state

[0105] Cycle second work step

[0106] Reference value

[0107] Intermediate value third work step

[0108] Auxiliary value fourth work step

[0109] Sensor value first state second state fifth work step

[0110] Feed circuit additional voltage source additional voltage source

Claims

Claims Method (74) for determining a state of a converter (2) based on a plurality of sensor values ​​(94), each of which corresponds to a measured value (50) created by means of an analog sensor (38) of a respective sensor unit (18), in which - a clock signal (76) comprising a plurality of identical clocks (80) with two different clock states (78) is sent to each sensor unit (18), - during one of the clock states (78) of each clock (80), a predetermined reference value (84) is changed by the respective current measured value (50) to a respective intermediate value (56) at each sensor unit (18) by means of a predetermined operation, - during the other clock state (78) of the respective same clock (80), the respective intermediate value (86) is changed by a respective auxiliary value (90) by means of an opposite operation up to the reference value (84) for each sensor unit (18), - during the respective remaining time period of the same clock cycle (80), a first state (96) and otherwise a second state (98) is used as the respective sensor value (94), - the sensor values ​​(94) are fed into a common output line (36) directed to an output point (34) by means of a feed circuit (102). Method (74) according to claim 1, characterized in that only one extreme value of the sensor values ​​(94) is transmitted to the output point (34). Method (74) according to claim 2, characterized in that, based on the respective sensor value (94), a switching element (28) of the respective sensor unit (18) is actuated, which is electrically connected in parallel between a reference potential (32) and the output point (34). Method (74) according to claim 3, characterized in that the respective sensor value (94) is transmitted to the switching element (28) in a galvanically isolated manner. Method (74) according to one of claims 1 to 4, characterized in that the clock signal (76) is received by means of a galvanic isolator (20) of the respective sensor unit (18). Method (74) according to one of claims 1 to 5, characterized in that the reference value (84) and the auxiliary value (90) are provided by means of a common voltage divider (66). Method (74) according to one of claims 1 to 6, characterized in that negative integration is used as the operation and integration is used as the opposite operation.Converter (2) comprising a plurality of sensor units (18), each having an analog sensor (38), and an output line (36) directed towards an output point (34), and which is operated according to a method (74) according to one of claims 1 to 7.

Images