Current sensor with frequency control

Abstract

The invention describes a current sensor circuit. According to one exemplary embodiment, the circuit comprises a magnetic core, a secondary winding arranged on the magnetic core, a controllable voltage source designed to apply a voltage with an adjustable magnitude and an adjustable polarity to the secondary winding, and a control circuit designed to actuate the voltage source in order to reverse the polarity of the voltage when magnetic saturation of the magnetic core is detected and to produce a modulated output signal at a sensor frequency. The circuit further comprises a controller designed to adjust the magnitude of the voltage on the basis of the sensor frequency.

Description

CURRENT SENSOR WITH FREQUENCY CONTROL TECHNICAL FIELD

[0001] The present description relates to sensor devices and methods for contactless current measurement, for example devices and methods according to the flux gate principle. BACKGROUND

[0002] So-called direct-imaging current sensors are used for contactless and thus potential-free current measurement. A direct-imaging current sensor measures the magnetic flux caused by the current to be measured, which depends on the strength of the current to be measured. The measured value of the magnetic flux is also a measure of the strength of the current to be measured. The conductor carrying the current to be measured is usually enclosed by a magnetic core (e.g., a toroidal core) that conducts the magnetic field. Hall sensors, for example, can be used as sensor elements for the magnetic flux. Current sensors constructed in this way are inexpensive but have relatively low accuracy.

[0003] Another group of current sensors is known as compensation current sensors. In these sensors, the magnetic field caused by the current to be measured in a magnetic core is compensated by a compensation winding arranged on the magnetic core, which generates an opposing magnetic field. Using a closed control loop, the compensation winding is driven with a compensation current so that the total magnetic field in the magnetic core is approximately zero. The zero point of the magnetic field can be detected using a magnetic field sensor element.

[0004] A special group of compensation current sensors are so-called flux-gate current sensors. These also have a mostly ring-shaped magnetic core with (at least) one primary winding that carries the current to be measured and (at least) one secondary winding that is part of an oscillator circuit. During operation of the oscillator circuit, the magnetic core is regularly remagnetized. For this purpose, the current flowing through the secondary winding can be reversed whenever magnetic saturation occurs in the magnetic core. Both the current through the secondary winding and the time until magnetic saturation is reached depend on the current to be measured in the primary winding. With appropriate subsequent signal processing, the influence of the magnetic core's hysteresis can be eliminated, thus achieving a very precise current measurement.

[0005] For this reason, flux-gate current sensors are also suitable for differential current measurement. Differential current measurement is a special case of total current measurement. In total current measurement, two or more primary conductors are magnetically coupled to the toroidal core, so that the magnetic fields of the currents flowing in the primary conductors constructively overlap in the toroidal core. Therefore, the current sensor measures the sum of the currents flowing in the primary conductors. If two primary windings are wound in such a way that the magnetic fields caused by the (primary) currents flowing in them destructively overlap in the toroidal core, the current sensor measures the difference between the two primary currents. Current sensors suitable for differential current measurement are frequently used in residual current circuit breakers.

[0006] An example of a differential current sensor operating according to the flux-gate principle is described in the publication US 2016 / 0033555 Al (Böttcher et al.). The inventors set out to improve existing current sensors based on the flux-gate principle, particularly with regard to unwanted nonlinearities. SUMMARY

[0007] The above-mentioned object is achieved by a circuit according to claim 1 and by the method according to claim 9. Various embodiments and further developments are the subject of the appended claims.

[0008] According to one embodiment, the circuit comprises a magnetic core, a secondary winding arranged on the magnetic core, a controllable voltage source configured to apply a voltage with an adjustable magnitude and an adjustable polarity to the secondary winding, and a control circuit configured to control the voltage source to reverse the polarity of the voltage when magnetic saturation of the magnetic core is detected, and a to generate a modulated output signal at a sensor frequency. The circuit further includes a regulator configured to adjust the magnitude of the voltage based on the sensor frequency.

[0009] Another embodiment relates to a method comprising: applying a voltage to a secondary winding wound around a magnetic core; detecting magnetic saturation of the magnetic core and reversing the polarity of the voltage as soon as saturation of the magnetic core is detected; generating a modulated output signal with a sensor frequency; regulating the sensor frequency by adjusting the magnitude of the voltage based on the sensor frequency. BRIEF DESCRIPTION OF THE ILLUSTRATIONS

[0010] Various embodiments are explained in more detail below with reference to illustrations. The illustrations are not necessarily to scale, and the invention is not limited to the aspects shown. Rather, emphasis is placed on illustrating the principles underlying the illustrated embodiments.

[0011] Figure 1 shows a block diagram of a current sensor operating according to the flux-gate principle.

[0012] Figure 2 shows in a diagram as magnetization versus magnetic field strength the idealized course of the magnetization characteristic of the magnetic core in a freely oscillating current sensor at a primary current of zero.

[0013] Figure 3 shows a current-time diagram of the idealized course of the secondary current in a freely oscillating current sensor with a primary current of zero.

[0014] Figure 4 shows in a diagram as magnetization versus magnetic field strength the idealized course of the magnetization characteristic of the magnetic core in a freely oscillating current sensor at a primary current greater than zero.

[0015] Figure 5 shows in a current-time diagram the idealized course of the secondary current in a freely oscillating current sensor with a primary current greater than zero.

[0016] Figure 6 shows a block diagram of a current sensor operating according to the flux-gate principle, particularly for measuring differential current and for detecting the strength of the primary current.

[0017] Figure 7 is a diagram illustrating the relationship between the sensor oscillation frequency and the primary current.

[0018] Figure 8 is a diagram illustrating the relationship between sensor sensitivity and sensor oscillation frequency.

[0019] Figure 9 is a block diagram of a current sensor circuit according to an embodiment with a control loop for controlling the voltage applied to the secondary winding.

[0020] Figure 10 is a general block diagram of the control circuit for controlling the voltage applied to the secondary winding.

[0021] Figure 11 is a block diagram of the control loop, using a frequency-to-voltage converter as in the example in Fig. 9.

[0022] Figure 12 is a block diagram of a digital implementation of the control loop. DETAILED DESCRIPTION

[0023] Fig. 1 shows a block diagram of an example of a current sensor operating according to the flux-gate principle. In the example shown, the current sensor is intended to measure a current flowing through a primary conductor 101, which is referred to below as primary current ip. The current sensor has a secondary conductor 102 and an, for example, unprotected, annular magnetic core 103 made of soft magnetic material. The magnetic core 103 magnetically couples the primary conductor 101 and the secondary conductor 102. The primary conductor 101 and the secondary conductor 102 can each be unwound (magnetically effective number of turns equal to 1) or wound (number of turns greater than 1), wherein the number of turns of the primary conductor 101 and the secondary conductor 102 can be the same or different. For example, the primary conductor 101 be guided straight (unwound, number of turns equal to 1) through the annular core 103 and enclosed by the annular core 103. The secondary conductor 102 can be wound and have a number of turns N>1. In this case, the secondary conductor 102 is also referred to as the secondary winding.

[0024] In the example shown, the secondary conductor 102 is fed by a controlled voltage source 104, which, for example, generates an approximately rectangular, bipolar source voltage ±UMAG and thus drives a secondary current is through the secondary conductor 102. The voltage source 104 can, for example, be implemented by means of a transistor HB bridge (DC voltage source with reversible polarity).

[0025] To measure the secondary current is, a measuring resistor 105 with a resistance value Res can be connected between the secondary conductor 102 and the voltage source 104. A voltage Ucs occurring across the measuring resistor 105 is measured by a measuring unit 106, preprocessed (e.g., digitized), and fed to a control unit 107 in the form of a current measurement signal representing the secondary current is. The control unit 107 generates a control signal CTR for the voltage source 104 from this signal. Other concepts for measuring the secondary current are known. For example, when using a transistor bridge circuit as the voltage source 104, so-called sense transistors can be used for current measurement.

[0026] The operation of the current sensor shown in Fig. 1 is explained in more detail below with reference to Figures 2 to 5. The magnetization characteristic shown in Figure 2 illustrates the ferromagnetic properties of the magnetic core 103 at a primary current of zero amperes, with the magnetic field strength H plotted on the abscissa and the magnetization M plotted on the ordinate. The magnetization characteristic exhibits an approximately rectangular hysteresis with a coercive field strength Hc and a saturation magnetization MSAT. For the magnetic field strength H, simplified according to Ampere's law, H = NÜS / IFE applies, where the parameter IFF denotes the effective magnetic path length of the magnetic field lines in the core 103 and N is the number of turns of the secondary conductor 102 when the number of turns of the primary conductor 101 is 1.

[0027] For a voltage ui induced in the secondary conductor 102, the following applies according to Faraday's law: ui= -N dO / dt = -NA dB / dt, (1) where A is the (effective) cross-sectional area of ​​the core 103, the magnetic flux in the core 103 caused by the secondary current is, B is the magnetic flux density with B = po (H+M), and po is the magnetic field constant. During the remagnetization of the core 103, which corresponds to the left or right vertical branch of the magnetization characteristic in Fig. 2, the rate of change of the magnetization dM / dt is proportional to the induced voltage ui, with the magnetic field strength H and thus also the secondary current is essentially constant, i.e., ui = -NA po dM / dt (during remagnetization) (2) In other words, the differential inductance of secondary coil 102 during magnetization reversal is very large (ideally infinitely large). As soon as the magnetization in core 103 reaches the saturation magnetization MSAT, the secondary current is increases and is limited only by the sum of the ohmic resistance of secondary conductor 102 and the ohmic resistance RM of measuring resistor 105.

[0028] The temporal profile of the secondary current is (while the primary current ip is equal to zero) is shown in Fig. 3. An (abrupt) increase in the level of the secondary current is (in the positive and negative directions) can be detected by the control unit 107, for example, using comparators. In the exemplary embodiment described here, the control unit 107 generates a signal to reverse the polarity of the voltage UMAG of the current source 104 as soon as the secondary current is exceeds a positive threshold value MSMAX or falls below a negative threshold value -is- MAX, thereby initiating the next magnetization reversal cycle. Essentially, the voltage source 104, the secondary winding 102, the current measuring circuit 106, and the control unit 107 form an oscillator circuit. Typically, the polarity reversal of the voltage source 104 occurs, as mentioned, when magnetic saturation is reached, thereby essentially forming a relaxation oscillator. The specific implementation of the oscillator circuit (e.g., implementation of the current source, current measurement, and magnetic saturation detection) is not important for the concepts described here. Various possibilities are known per se.

[0029] During the magnetization reversal process, the secondary current i can be assumed to be essentially constant and corresponds to a magnetizing current +iji or -iji. The magnitude of the magnetizing current i depends on the width of the hysteresis in the magnetization characteristic, i.e., on the coercive field strength Hc, i.e., ip ~ IFE / N HC (for a primary current ip = 0A). As soon as the magnetization in core 103 reaches the positive or negative saturation magnetization ±MSAT, the secondary current is begins to increase in the positive or negative direction, as mentioned above. Due to the symmetry of the hysteresis characteristic, the temporal variation of the secondary current is is symmetrical about a mean value.

[0030] Fig. 4 and 5 show the magnetization characteristic and the time course of the secondary current is for the case where the primary current ip is not zero. The magnetic field generated by the primary current ip is superimposed in the core 103 with the magnetic field of the secondary current is, which is expressed in Fig. 4 as a shift of the magnetization characteristic along the abscissa (compared to Fig. 2). The corresponding time course of the secondary current is is shown in Fig. 5. This is similar to the case shown in Fig. 3, in which the primary current ip is zero, but with the difference that the secondary current is no longer symmetrical to the abscissa (is=0), but symmetrical to a straight line shifted parallel to it with is = ip / k.This means that during the magnetization reversal process, the primary current ip and secondary current is have the same ratio k as the number of turns of the primary conductor 101 and the secondary conductor 102, except for a hysteresis offset equal to the magnitude of the magnetization current ±ip. The ratio k corresponds to the number of turns N of the secondary conductor 102 when the number of turns of the primary conductor is 1.

[0031] There are various approaches to determining a measured value for the primary current. A first approach involves evaluating the secondary current during the magnetization reversal process. Accordingly, the secondary current signal is or the voltage ucs at the measuring resistor 105 is sampled during the magnetization reversal process. (see Fig. 3). By sampling the signal representing the secondary current is in the first half of a period of the secondary current is (first measurement cycle), a current measurement value is[nl] = (ip / N)+ip is obtained, and in the second half of the period (second measurement cycle), a current measurement value is[n] = (ip / N)-ip. By averaging over the first and second measurement cycles, the hysteresis error caused by the magnetizing current can then be eliminated. The primary current ip at a sampling time n is calculated as follows: ip[n] = (N / 2) (is[nl]+is[n]). (3)

[0032] Because the hysteresis of the magnetization characteristic has virtually no influence on the measurement result, this current measurement method is very well suited for measuring very small currents (and also for differential current measurements). The measuring range extends, depending on the circuit design, from a few milliamperes to kiloamperes. During the magnetization reversal process of the magnetic core 103, the secondary current is follows the primary current ip according to the transfer ratio N:l. The secondary current is is sampled at least once during the magnetization reversal process in order to obtain a measured value (is + ip or is - ip) for determining the primary current ip. During the magnetization reversal process, sampling can also be repeated at a defined sampling rate that is significantly higher than the oscillation frequency of the sensor fsENsoR.The secondary current is is approximately constant during the magnetization reversal process and before magnetic saturation occurs in the core 103 and is equal to (ip / N) ± ip. This idealized approach can be particularly useful when the hysteresis characteristic of the magnetic core 103 is approximately rectangular.

[0033] The procedure explained above with reference to Figs. 1 to 5 can also be used for differential current measurement with only minor modifications to the current sensor circuit shown in Fig. 1. A differential current measurement is used, for example, in residual current circuit breakers. For this purpose, the primary winding 101 is divided into a first partial winding 101a and (at least) a second partial winding 101b. Accordingly, the primary current flowing through the first partial winding 101a is designated ipi and the primary current flowing through the second partial winding 101b is designated ip2. The partial windings 101a and 101b can each consist of only a single turn and be oriented such that the currents ip generated by the primary currents a and ipb at least partially compensate (destructively superimpose) each other, and only the net primary current ipi-ip2 (effective primary current) generates a corresponding net magnetic field in the core 103. The net magnetic field is in turn superimposed by the magnetic field of the secondary current is. A correspondingly modified sensor is shown in Figure 6. Apart from the fact that the example in Fig. 6 has two primary windings 101a, 101b through which an effective primary current ip = ipi-ip2 flows, the example in Fig. 6 is the same as the example in Fig. 1. Depending on the orientation of the primary conductors 101a and 102b, a total current measurement can be performed instead of a differential current measurement. More than two primary conductors can also be provided, so that theoretically a sum of any number of currents is measured. The primary current in this case is ip = ipi ± ip2 ± ips, etc.In the case of differential current measurements, the term Aip is used below for the effective primary current (differential current). Whether currents are added or subtracted depends on the orientation of the partial windings 101a, 101b.

[0034] In the example shown in Fig. 6, the two primary windings 101a and 101b are connected upstream and downstream of a load 601, respectively, so that the difference, i.e., the net primary current ip i -ip2, is only nonzero if, for example, a leakage current (fault current) flows in the area of ​​the load 601, which then exactly corresponds to the differential current Aip = ipi - ip2. The differential current Aip is calculated from sampled values ​​of the secondary current analogously to Equation 3 as follows: Aip[n] = N (is[n-1] + is[n]) / 2. (4)

[0035] According to a second approach to determining a measured value for the primary current, the times of the magnetization reversals are evaluated. A magnetization reversal process begins with the polarity reversal of the voltage source 4 and ends when magnetic saturation is reached (which triggers the next polarity reversal of the voltage source 104). One cycle comprises two magnetization reversals, e.g.: from the positive saturation of the core until negative saturation is reached and (after the polarity reversal of the voltage source 104) back until positive saturation is reached. In Figs. 3 and 5, these times are denoted by Δt+ and Δt-, where Δt+ is the time until magnetic saturation is reached in the positive direction (i.e., +MSAT) and Δt- is the time until Reaching magnetic saturation in the negative direction (ie -MSAT).

[0036] The time intervals Δt+ and Δt- (see Figure 3) are not constant, but depend on the magnitude of the primary current. Furthermore, the remagnetization rate dM / dt, according to Equation 2, increases the higher the amplitude UMAG of the voltage ±UMAG generated by the voltage source 104. From Equation 2, it follows: dM / dt = - ui / (NA-po) = - (UMAG - Rcs-is) / (NA-po). (5) Consequently, the oscillation frequency of the secondary current is higher, the higher the amplitude UMAG of the voltage ±UMAG generated by the voltage source 104. The oscillation frequency FSENSOR of the sensor is calculated by integrating Equation 5: fsENsoR = l / (At+ + At-), where (6a) At+ = (po-AM-NA) / (UMAG+(ip / Ni H )-Rcs), (6b) At- = (po-AM-NA) / (UMAG-(ip / N+iji)-RRcs). (6c) Here, AM denotes the magnetization swing during a magnetization reversal (e.g., AM = 2 MSAT) and ip / N±iji denotes the secondary current during the respective magnetization reversal process. From equations 6a to 6c, it can be seen that the oscillation frequency fsENsoR of the sensor depends on the primary current itself, as well as on the voltage amplitude UMAG of the voltage generated by the voltage source 104 and on the magnetization swing AM. A third approach to determining a measured value for the primary current can therefore consist of evaluating the duty cycle of the oscillator circuit. Fig. 5 also shows the signal SpMw(t), which is a binary signal and indicates the polarity of the voltage source 104. The signal SpMw(t) can easily be generated from the secondary current signal is(t) using a comparator circuit.The signal SpwM(t) changes from a low level to a high level when negative saturation is detected in the magnetic core 103 (is<-isMAx), and from a high level to a low level when positive saturation is detected in the magnetic core 103 (is>+isMAx). The switching conditions is<-isMAx and is>+isMAx can be easily detected, for example, using comparators. The signal SPWM(1) is then the output signal of the. Comparator circuit used to detect magnetic saturation (see Fig. 9, saturation detection 108). The signal SPWM(1) is a pulse-width modulated (PWM) signal with a variable frequency and a duty cycle D=At+ / (At++At-) that depends on the effective primary current.

[0037] In many applications, it is desirable to output the measured primary current information as a PWM signal. Since the duty cycle of a PWM signal is usually specified as a percentage, the sensor sensitivity can be specified as a percentage per ampere. However, studies have shown that the sensor sensitivity is not constant but varies with the oscillation frequency fsENsoR, which depends on the primary current. These relationships are illustrated in Figs. 7 and 8.

[0038] Fig. 7 uses an example to illustrate the relationship between the sensor frequency fsENsoR and the primary current. Fig. 8 shows the change in sensor sensitivity (in percent per ampere) as a function of the sensor frequency fsENsoR. Fig. 8 shows that the sensitivity decreases with increasing sensor frequency. It is only practical to assume the sensitivity to be constant within a relatively small measuring range, because with a small change in the primary current, the sensor frequency changes only slightly (see Fig. 7; in the range between -0.5A and 0.5A, the frequency is approximately constant) and, consequently, the sensitivity can also be assumed to be constant. For larger measuring ranges, these assumptions are no longer applicable.

[0039] With regard to the relationships shown in Figures 7 and 8, it can be summarized that a nonlinear relationship exists between duty cycle and current. The exemplary embodiments discussed below aim to reduce this nonlinearity by maintaining the sensor frequency at a constant value using a control loop.

[0040] The example in Fig. 9 illustrates an exemplary implementation of the differential current sensor from Fig. 6. The signal processing contained in the control unit 107 allows current measurement based on both the secondary current and the duty cycle of the PWM signal. The primary conductors 101a, 101b, the magnetic core 103, and the secondary winding 102 are the same as in the example in Fig. 6. The controllable Voltage source 104 is implemented, for example, as an H-bridge with four semiconductor switches Si, S2, S3, and S4. During operation of the circuit, either the switches Si and S4 are on while the switches S2 and S3 are off, or vice versa (the switches S2 and S3 are on while the switches Si and S4 are off). By switching the H-bridge (e.g., switching Si and S4 off, switching S2 and S3 on), the polarity of the voltage applied to the secondary winding 102 is reversed. The secondary current is flowing through the secondary winding 102 also flows through the measuring resistor Res, and the voltage drop ucs across the resistor Res (current measurement signal) represents the secondary current is (ucs=Rcs is).

[0041] The current measurement signal ucs is fed to the current measurement circuit 106. In the example shown, this circuit is designed to perform analog preprocessing of the current measurement signal ucs and to digitize the preprocessed signal ucs'. The resulting digital signal is denoted by ucs[n] in the example from Fig. 8. The analog preprocessing can, for example, comprise one of the following: amplification, level adjustment, filtering, or the like. In the example shown, the digital signal ucs[n] is fed to a microcontroller 115, which is designed to calculate a measured value for the differential current Aip (e.g., according to equation 4 or according to equation 3 if no differential current is measured). Instead of a microcontroller, another digital computing unit (programmable or hard-wired) can also be used. The analog / digital converter of the current measurement circuit 106 can also be part of the microcontroller 115.

[0042] For example, with very small (differential) currents, quantization errors that arise during the digitization of the possibly preprocessed current measurement signal ucs can have a significant impact on the measurement result. A somewhat more precise measurement is made possible by the above-mentioned evaluation of the times Δt+ and Δt- (see also Fig. 5), whereby the duty cycle D is used in particular as the measured value for the (differential) current. The duty cycle D corresponds to the ratio Δt+ / (Δt+ + Δt-) and is usually specified as a percentage. In the example from Fig. 3, where the primary current or the primary current difference is zero, the duty cycle is 0.5 or 50 percent. In the example from Fig. 5, where the primary current or the primary current difference is positive, the duty cycle is less than 50 percent. Depending on the definition of the current direction, the duty cycle can also be greater than 50 percent for positive currents. In some cases, The ratio R=At+ / At- is also evaluated, which is practically equivalent to the duty cycle evaluation, since D=R / (1+R). In both cases, a division must be performed to calculate R or D if the evaluation is performed digitally, as in the example shown (although this is not necessarily the case). In yet other embodiments, the time difference AT = At- - At+ can be used as a measure of the effective primary current.

[0043] The control circuit 107 includes the aforementioned circuit 108 for saturation detection. This can, for example, contain one or more comparators to detect the increase in the secondary current as a result of the magnetic saturation of the core 103 (see Figs. 3 and 5). The circuit 108 is designed to generate a PWM signal SpwM(t), where, for example, a rising edge represents the detection of magnetic saturation of positive polarity (+MSAT) and a falling edge represents the detection of magnetic saturation of negative polarity (-MSAT) (see hysteresis curve in Fig. 4). The PWM signal SPWM(I) is fed to the switch controller 109, which is designed to reverse the polarity of the voltage UMAG applied to the secondary winding 102 at each edge, as described above. For this purpose, the switch controller 109 can, for example, have a plurality of flip-flops. As mentioned, the frequency fsENsoR= (At- + At )" 1 of the PWM signal SPWM(I) is not necessarily constant, but varies with the effective primary current (see Fig. 7). In the example shown, the times Δt- and Δt+ are measured using circuit 110. This circuit can, for example, have one or more counters to determine the times Δt- and Δt+. Time measurement using counters is known per se and will therefore not be discussed in detail here. The counter readings, which represent the desired times, are fed to microcontroller 115, which can calculate a measured value for the effective primary current based on them (among other things). Circuit 110 can also be part of the microcontroller.

[0044] In the example shown, the microcontroller 115 would in principle be able to calculate the frequency fsENsoR from the measured times Δt- and Δt+. However, in the example shown, a frequency-to-voltage converter 111 (f / U converter) is used. This converter is designed to generate a voltage signal UF ZU that is proportional to the sensor frequency FSENSOR. This voltage signal UF can be used to regulate the voltage UMAG such that the sensor frequency remains approximately constant and corresponds to a reference frequency fREF (fsENsoR = fREF). In the example shown in Fig. 9, the reference voltage UREF represents the reference frequency fREF (UREF ~ fREF). The voltage difference UF-UREF is fed to the controller 20, which provides the voltage UMAG at a controller output (UF-UREF ~ fsENsoR-fREF). The controller 20 can, for example, be a proportional controller (P controller), a proportional-integral controller (PI controller) or the like. In an alternative example, the controller 20 can be designed to increase or decrease the voltage UMAG in discrete steps AU (depending on whether the difference UF-UREF is positive or negative) until the absolute value of the difference |UF-UREF| is less than a threshold value Us.

[0045] Fig. 10 illustrates the control loop for controlling the sensor frequency using a block diagram. The controlled system is the flux-gate sensor 10 (as shown, for example, in Fig. 9). The controlled system receives the voltage UMAG as input and delivers the sensor frequency as output. The sensor frequency is the controlled process variable. The controller 20 receives the control error Af=fREF-fsENSOR as input and generates the voltage UMAG for the controlled system as the manipulated variable. The control error is the deviation of the actual sensor frequency fsENsoR from the setpoint fREF. Depending on the transfer function of the controller 20, the control error can be very small (ideally zero).

[0046] The control loop shown in Fig. 10 merely illustrates the abstract model of the control system. The control loop in Fig. 11 is closer to the practical implementation. In the example shown, the frequency fsENsoR of the PWM signal SPWM(Ü is converted into a voltage UF (voltage-to-frequency converter 111, see also Fig. 9). The voltage UF (which represents the frequency FSENSOR) is subtracted from a reference voltage UREF (which represents the setpoint fREF). The difference AU then represents the control error AF, which is fed to the controller 20 as an input signal. The control loop in Fig. 11 essentially corresponds to the embodiment in Fig. 9.

[0047] Fig. 12 illustrates another implementation of the control loop for controlling the sensor frequency FSENSOR. In the example shown, the period of the PWM signal SPWM(I) is measured by determining the times Δt+ and Δt- (e.g., by means of the (counter) circuit 110). The measured times Δt+ and Δt- can be added digitally, e.g., by the microcontroller 115 (see Fig. 9). The sum Δt+ + Δt- is equal to the period TSENSOR = FSENSOR' 1 . In the example shown, the microcontroller 115 calculates a control error AT=fRi i -1 -TSENSOR, which is supplied to the controller 20 as an input signal. In one embodiment, the circuit 110 is also part of the microcontroller 115. Thus, the entire control loop can be implemented in the microcontroller 115.

Claims

PATENT CLAIMS 1. A current sensor circuit comprising: a magnetic core (103), a secondary winding (102) arranged (103) on the magnetic core, a controllable voltage source (104) configured to apply a voltage (UMAG) with an adjustable magnitude and an adjustable polarity to the secondary winding (102); a control circuit (107) configured to control the voltage source (104) to reverse the polarity of the voltage (UMAG) when magnetic saturation of the magnetic core (103) is detected and to generate a modulated output signal (SPWM) with a sensor frequency (fsENsoR), a controller (20) configured to adjust the magnitude of the voltage (UMAG) based on the sensor frequency (fsENsoR).

2. The current sensor circuit according to claim 1, wherein the magnetic core (103) is designed to be magnetically coupled to a primary conductor (101) during operation, and wherein a duty cycle of the output signal (SPWM) serves as a measure of a primary current (ip) flowing through the primary conductor (101).

3. The current sensor circuit according to claim 1, wherein the magnetic core (103) is designed to be magnetically coupled to two or more primary conductors (101a, 101b) during operation, and wherein a duty cycle of the output signal (SPWM) serves as a measure of an effective primary current, in particular of a differential current (ip2-ipi).

4. The current sensor circuit according to one of claims 1 to 3, wherein the regulator (20) is designed to regulate the magnitude of the voltage (UMAG) such that an existing deviation of the sensor frequency (fsENsoR) from a desired value is reduced.

5. The current sensor circuit according to 4, wherein the controller (20) is designed to adjust the magnitude of the voltage (UMAG) until the sensor frequency (FSENSOR) substantially corresponds to the setpoint value.

6. The current sensor circuit according to one of claims 1 to 5, further comprising: a frequency-to-voltage converter (111) configured to generate a signal (UF) representing the sensor frequency (FSENSOR), wherein the controller (20) is configured to adjust the magnitude of the voltage (UMAG) based on the signal (UF) representing the sensor frequency (FSENSOR).

7. The current sensor circuit according to one of claims 1 to 6, wherein the control circuit (107) is configured to detect magnetic saturation of the magnetic core (103) by detecting that a current flowing through the secondary winding (102) reaches or exceeds a threshold value.

8. The current sensor circuit according to claim 7, wherein the output signal (SPWM) is based on a current measurement signal (ucs) representing the current flowing through the secondary winding (102), which changes its level when magnetic saturation of the magnetic core (103) is detected, and wherein the current measurement signal (ucs) also comprises the sensor frequency (FSENSOR).

9. A method comprising: Applying a voltage (UMAG) to a secondary winding (102) wound around a magnetic core (103), Detecting a magnetic saturation of the magnetic core (103) and reversing the polarity of the voltage (UMAG) as soon as a saturation of the magnetic core (103) is detected; Generating a modulated output signal (SPWM) with a sensor frequency (fsENSOR); Control the sensor frequency (FSENSOR) by adjusting the voltage magnitude (UMAG) based on the sensor frequency (fsENsoR).

10. The method according to claim 9, wherein the magnetic core (103) is magnetically coupled to at least one primary conductor (101; 101a, 101b), and wherein a duty cycle of the output signal (SPWM) is used as a measure of an effective primary current (ip, ipi- Ip?).

11. The method according to claim 9 or 10, further comprising: Generating a signal (UF) representing the sensor frequency (fsENsoR) and using this signal (UF) to control the sensor frequency (fsENsoR).

12. The method according to claim 9 or 10, further comprising: regularly determining a period (TSENSOR) of the output signal (SPWM) and using the determined period (TSENSOR) for controlling the sensor frequency (fsENSOR).

Images