Magnetic current sensor combining two measurement channels

Abstract

A current sensor with a flow valve comprising a magnetic core (13), a measuring winding (14), an excitation circuit (23) arranged to generate a digital excitation signal (Se), an acquisition circuit (17) arranged to acquire an analog measuring voltage (Ve) across the measuring winding (14) and to produce a digital measuring signal (Sm), a demagnetization control circuit (32) arranged to produce, from the digital measuring signal, a digital demagnetization signal (Sdm) intended to compensate for a magnetic flux produced by the current to be measured, a summing device (36) arranged to sum the digital excitation signal and the digital demagnetization signal so as to obtain a digital injection signal (Si), and an injection circuit (37) arranged to produce an analog excitation current from the digital injection signal and to inject the analog excitation current into the measuring winding.SUMMARY FIGURE : Fig. 2.

Description

Title of the invention: Magnetic current sensor combining two measurement channels

[0001] The invention relates to the field of magnetic current sensors such as flow valve sensors.

[0002] BACKGROUND OF THE INVENTION

[0003] A magnetic current sensor with a flux valve uses the property of a magnetic material, forming a magnetic core, to saturate from a certain level of magnetic excitation: for an increasing magnetic field, the slope of the transfer function f between the magnetic field (commonly noted B) and the magnetic induction (commonly noted H) decreases greatly from a so-called saturation value of the magnetic core corresponding to the inflection point of the curve B=f(H).

[0004] [Fig.l] schematically shows a current sensor with a flow valve of the type known from document WO-A-2020 / 002184 intended to measure a current Im flowing on a conductor. The magnetic current sensor comprises a magnetic core 1, which extends around a conductor 2 on which the current to be measured Im flows, and a measurement winding 3 around the magnetic core 1. The sensor also comprises an excitation circuit 4 arranged to generate a digital excitation signal Se, an acquisition circuit 5 comprising an analog-to-digital converter 6 and arranged to acquire an analog measurement voltage across the measurement winding 2 and to produce a digital measurement signal Sm.A demagnetization control circuit 7 is arranged to produce, from the digital measurement signal Sm and after low-pass filtering, a digital demagnetization signal Sdm which is an image signal of the current to be measured Im, also intended to compensate for a magnetic flux produced by the current to be measured Im. A summer 8 sums the digital excitation signal Se and the digital demagnetization signal Sdm so as to obtain a digital injection signal Si. An injection circuit 9, comprising a digital-to-analog converter 10, is arranged to produce an analog excitation current Ie from the digital injection signal Si and to inject the analog excitation current Ie into the measurement winding 3.

[0005] Flow valve current sensors are preferred in a number of applications. This is particularly the case for measuring a current flowing in a SSPC (Solid State Power Controller) type cut-off device intended for user equipment, or for measuring a current flowing in a PEM (Power Electronic Module) type energy conversion device connected to a phase of a motor.

[0006] However, the implementation of such a current sensor tends to limit the bandwidth of the measured signal. However, there is a growing need to measure currents over a wider bandwidth. In addition, it is also desirable to better measure the DC components, as well as the amplitude and phase of the AC components.

[0007] SUBJECT OF THE INVENTION

[0008] The object of the invention is to meet at least part of this need. Summary of the invention

[0009] To this end, according to the invention, a magnetic current sensor is provided, comprising a magnetic core which extends around a conductor on which a current to be measured flows, a measurement winding, an excitation circuit arranged to generate a digital excitation signal, an acquisition circuit comprising an analog-digital converter and arranged to acquire an analog measurement voltage at the terminals of the measurement winding and to produce a digital measurement signal, a demagnetization control circuit arranged to produce, from the digital measurement signal and after a first low-pass filter, a digital demagnetization signal intended to compensate for a magnetic flux produced by the current to be measured, a summator arranged to sum the digital excitation signal and the digital demagnetization signal so as to obtain a digital injection signal,and an injection circuit comprising a digital-analog converter and arranged to produce an analog excitation current from the digital injection signal and to inject the analog excitation current into the measurement winding. The sensor comprises a processing circuit having a first path for processing the demagnetization signal and a second path for processing the digital measurement signal which are connected to a summer to provide an image signal of the current, the second processing path being arranged to cancel in the digital measurement signal at least a part corresponding to a fundamental component of the digital excitation signal and comprising a high-pass filter strictly complementary to the first low-pass filter.

[0010] In the sensor of the invention, two signals having undergone different processing are added to obtain the image signal of the current to be measured. The first processed signal is the demagnetization signal which is provided by the flow valve circuit and incorporates a DC component but has a relatively low bandwidth due to the low-pass filtering of the demagnetization servo circuit. The digital measurement signal, which has not passed through the demagnetization servo circuit, has a relatively large bandwidth but includes at least one part corresponding to a fundamental component of the digital excitation signal. The second processing channel allows this component to be cancelled. By adding the signals coming out of the two processing channels, we therefore obtain a current image signal which benefits from the increased bandwidth of the digital measurement signal in the high frequencies while retaining the richness of the demagnetization signal in the low frequencies. We can also obtain two summed signals at the end of the filtering having a strictly constant amplitude response as a function of the frequency.

[0011] According to optional characteristics: - the second processing path comprises a first loop for extracting, from the digital measurement signal, a part corresponding to an in-phase amplitude of the fundamental component of the digital excitation signal and a second loop for extracting, from the digital measurement signal, a part corresponding to a quadrature amplitude of the fundamental component of the digital excitation signal and, preferably, the second path comprises at least one gain balancing component between the loops; - the digital excitation signal is obtained from a digital reference signal of frequency f0 and a digital reference signal of frequency 3.f0; the excitation circuit preferably comprising an excitation servo circuit which is arranged to servo the digital excitation signal so as to cancel a frequency component 3.f0 of the analog measurement voltage; the excitation servo circuit preferably comprising a first multiplier arranged to multiply the digital measurement signal by a digital demodulation signal of frequency 3.f0; the excitation servo circuit preferably further comprising an integrator, a first amplifier connected to an output of the integrator and arranged to produce a digital error signal, and a second multiplier which multiplies a raw digital excitation signal by the digital error signal to obtain the digital excitation signal; the degaussing servo circuit comprising a third multiplier arranged to multiply the digital measurement signal by a digital demodulation signal of frequency 2.f0; the degaussing servo circuit preferably further comprising a second amplifier connected to an output of the first low-pass filter and arranged to produce the digital degaussing signal; . - the sensor comprises at least one gain balancing member between the two processing channels

[0012] The invention will be better understood in light of the following description of a method of particular non-limiting implementation of the invention. Brief description of the drawings

[0013] Reference will be made to the accompanying drawings, among which:

[0014] [Fig-1] [Fig. 1] schematically represents a flow valve current sensor according to the prior art;

[0015] [Fig.2] [Fig.2] schematically represents a flow valve current sensor according to the invention. DETAILED DESCRIPTION OF THE INVENTION

[0016] With reference to [Fig.2], a flow valve current sensor according to the invention is used here to measure a current to be measured Im which flows on a conductor 11.

[0017] The flux gate current sensor comprises a transformer 12 comprising a magnetic core 13, a primary winding and a secondary winding. The primary winding is the conductor 11. The magnetic core 13 extends around the conductor 11. The secondary winding is a measuring winding 14 wound around the magnetic core 13.

[0018] The flow valve current sensor comprises a plurality of electronic components forming a digital part 15 and an analog part 16.

[0019] The digital part 15 comprises a digital processing component which is in this case an FPGA but which could be a different component: microcontroller, processor, ASIC, etc.

[0020] The flow valve current sensor further comprises an acquisition circuit 17. The acquisition circuit 17 belongs to the analog part 16.

[0021] The acquisition circuit 17 is connected to the measuring winding 14.

[0022] The acquisition circuit 17 comprises a low-pass filter 19 and an analog-to-digital converter 20 connected to an output of the low-pass filter 19. The low-pass filter 19 provides an anti-aliasing function and here has a cut-off frequency of approximately 10 MHz but other values ​​can be used.

[0023] The analog-to-digital converter 20 is here a 12-bit converter whose operation is clocked at a sampling frequency Fech. Here Fech = 100MHz. The analog-to-digital converter 20 could of course have different characteristics.

[0024] The acquisition circuit 17 acquires an analog measurement voltage Ve at the terminals of the measurement winding 14. The measurement voltage Ve is applied to the input of the low-pass filter 19.

[0025] The resulting voltage is applied to the input of the analog-to-digital converter 20 which produces a digital measurement signal Sm.

[0026] The flow valve current sensor comprises a high frequency generator 22 known in itself. The high-frequency generator 22 is integrated into the FPGA (one could replace the term "integrated" with one of the following terms: programmed, implemented, implemented, etc.)

[0027] The high-frequency generator 22 produces a digital reference signal of frequency f0 (or signal H1), a digital reference signal of frequency 2.f0 (or signal H2) and a digital reference signal of frequency 3.f0 (or signal H3).

[0028] The digital reference signal of frequency fO is here of the form:

[0029] Sir = sin(coOt).

[0030] The digital reference signal of frequency 2.f0 is here of the form:

[0031] S2r = cos(2.co0t).

[0032] The digital reference signal of frequency 3.f0 is here of the form:

[0033] S3r = sin(3.œ0t).

[0034] The flow valve current sensor further comprises an excitation circuit 23. The excitation circuit 23 is implemented in the FPGA.

[0035] The excitation circuit 23 is connected to the high-frequency generator 22 and comprises a raw excitation circuit 24 and an excitation control circuit 25.

[0036] The raw excitation circuit 24 receives the digital reference signal of frequency f0 and the digital reference signal of frequency 3.f0 and produces from these signals a digital partial excitation signal of frequency f0 and a digital partial excitation signal of frequency 3.f0.

[0037] The digital partial excitation signal of frequency fO is here of the form:

[0038] Slp = kl.sin(coOt).

[0039] The digital partial excitation signal of frequency 3.f0 is here of the form:

[0040] S3p = k2.sin(3.co0t).

[0041] The raw excitation circuit 24 adds the partial digital excitation signal of frequency f0 and the partial digital excitation signal of frequency 3.f0 to generate a raw digital excitation signal Sb. The raw digital excitation signal is of the form:

[0042] Sb = Slp + S3p = kl.sin(coOt) + k2.sin(3co0t).

[0043] The excitation servo circuit 25 is connected to the high-frequency generator 22 to receive the digital reference signal of frequency 3.f0 and is arranged to produce, from the digital reference signal of frequency 3.f0, a digital demodulation signal of frequency 3.f0.

[0044] The digital demodulation signal of frequency 3.f0 is here of the form:

[0045] S3d = k3.sin(3.co0t).

[0046] The excitation control circuit 25 comprises a first multiplier 27, a second multiplier 28, an integrator 29 and a first amplifier 30. The first amplifier 30 is connected to an output of the integrator 29.

[0047] The first multiplier 27 multiplies the digital measurement signal Sm by the digital demodulation signal of frequency 3.f0. The resulting signal is applied to the input of the integrator 29. It is noted that the first multiplier 27 and the integrator 29 play the role of a first synchronous demodulator.

[0048] The first amplifier 30 then produces a digital error signal. The second multiplier 28 multiplies the raw digital excitation signal Sb by the digital error signal to obtain a digital excitation signal Se. The digital excitation signal is of the form:

[0049] Se = kO.(kl.sin(coOt) + k2.sin(3co0t)).

[0050] The flow valve current sensor further comprises a demagnetization servo circuit 32. The demagnetization servo circuit 32 is integrated into the FPGA.

[0051] The demagnetization servo circuit 32 is connected to the high-frequency generator 22 to receive the digital reference signal of frequency 2.f0 and is arranged to produce, from the digital reference signal of frequency 2.f0, a digital demodulation signal of frequency 2.f0.

[0052] The digital demodulation signal of frequency 2.f0 is here of the form:

[0053] S2d = k4.cos(2.co0t).

[0054] The demagnetization servo circuit 32 comprises a third multiplier 33, a low-pass filter 34 and a second amplifier 35. The second amplifier 35 is connected to an output of the low-pass filter 34.

[0055] The third multiplier 33 multiplies the digital measurement signal Sm by the digital demodulation signal of frequency 2.f0. The resulting signal is applied to the input of the low-pass filter 34. The second amplifier 35 then produces a digital image signal of the current to be measured Im, which is also a digital demagnetization signal Sdm. The digital image signal, or digital demagnetization signal Sdm, is of the form

[0056] Sdm = k.Im.

[0057] It is noted that the third multiplier 33 and the low-pass filter 34 play the role of a second synchronous demodulator.

[0058] The flow gate current sensor further comprises a summer 36. The summer 36 is implemented in the FPGA.

[0059] The adder 36 sums the digital excitation signal Se and the digital demagnetization signal Sdm to obtain a digital injection signal Si.

[0060] We have:

[0061] Si = kO.(kl.sin(coOt) + k2.sin(3co0t)) + klm.

[0062] The flow valve current sensor further comprises an injection circuit 37. The injection circuit 37 belongs to the analog part 16. The injection circuit 37 is connected to the adder 36 and to the measuring winding 14.

[0063] The injection circuit 37 comprises a digital-to-analog converter 38.

[0064] The digital-to-analog converter 38 is here a 12-bit converter whose operation is clocked at the frequency Fech (Fech = 100MHz). The digital-to-analog converter 38 could of course have different characteristics.

[0065] The digital-to-analog converter 38 acquires the digital injection signal Si, produces an analog excitation current Ie from the digital injection signal Si, and injects the excitation current Ie into the measuring winding 14.

[0066] The flow valve sensor further comprises a processing circuit 40 for processing the digital image signal or digital degaussing signal Sdm and the digital measurement signal Sm.

[0067] The processing circuit 40 comprises a first processing channel having as input a multiplier 41 having a first input receiving the digital demagnetization signal Sdm, a second input receiving data from a memory 42 and an output connected to a first input of a summer 43 having an output connected to an input of a low-pass filter 44 having a cut-off frequency equal to one tenth of the sampling frequency. As will be seen later, the low-pass filter 44 provides as output a digital signal image of the measured current freed from the fundamental and harmonics of the excitation signal. The memory 42 is a non-volatile memory which contains a coefficient to be applied to the demagnetization signal Sdm by the multiplier 41 to balance the amplitude of the demagnetization signal Sdm at the input of the summer 43 with the amplitude of the digital measurement signal Sm.Indeed, the two measurement channels do not necessarily have exactly the same amplitude gains. However, to guarantee the flatness of the frequency response in the entire measurement frequency band, it is necessary that the two measurement channels have the same gain. To achieve this objective, in the "factory" adjustment phase, an AC current of constant amplitude is injected (into the wire in which the current to be measured flows) and swept from a frequency much lower than the cutoff frequency of the transition filter to a frequency much higher than this transition frequency. The value of the gain coefficient to be applied is adjusted to obtain this frequency response flatness. Once obtained, this coefficient is recorded in the memory (or could also be called a non-volatile memory register).Alternatively, it may be provided that the memory 42 contains a table containing coefficients at different temperatures and that the sensor comprises a temperature probe to be able to select from the table the coefficient to be applied for the measured temperature.

[0068] The processing circuit 40 also comprises a second processing channel comprising a servo-control circuit 50 for servo-controlling to zero the part of the measurement signal Sm corresponding to the fundamental component of the excitation signal. servo circuit 50 comprises two servo loops 50.1, 50.2 each comprising: a first multiplier 51.1, 51.2; a low-pass filter 52.1, 52.2 having an input connected to an output of the multiplier 51.1, 51.2; an amplifier 53.1, 53.2 having an input connected to an output of the low-pass filter 52.1, 52.2; a second multiplier 54.1, 54.2 having a first input connected to the output of the amplifier 53.1, 53.2.

[0069] The multiplier 51.1 has a first input connected to an output of a summer 45 arranged at the input of the second processing channel 40 and a second input connected to the frequency generator 22 to receive an amplitude signal in phase of the digital reference signal of frequency f0. The multiplier 54.1 has a second input connected to the frequency generator 22 to receive the amplitude signal in phase of the digital reference signal of frequency f0 and an output connected to a first input of the summer 45.

[0070] The multiplier 51.2 has a first input connected to the output of the adder 45 and a second input connected to the frequency generator 22 to receive an amplitude signal in quadrature of the digital reference signal of frequency f0. The multiplier 54.2 has a second input connected to the frequency generator 22 to receive the amplitude signal in quadrature of the digital reference signal of frequency f0 and an output connected to a second input of the adder 45.

[0071] It is noted that the multiplier 51.1 and the low-pass filter 52.1 form a synchronous demodulator of the amplitude signal in phase of the digital reference signal of frequency f0 and that the multiplier 51.2 and the low-pass filter 52.2 form a synchronous demodulator of the amplitude signal in quadrature of the digital reference signal of frequency f0.

[0072] Each amplifier 53.1, 53.2 is associated with a memory 55.1, 55.2 containing an adjustable gain coefficient to equalize the gains of the two processing channels for values ​​close to the cutoff frequency of the filters.

[0073] Memories 55.1 and 55.2 contain the values ​​of the amplitude of the H1 component (in I and in Q) of the excitation signal of the magnetic core 13. These values ​​will be used to inject into the signal restored at the output of the analog / digital converter a signal having these values ​​in phase opposition so as to cancel the H1 excitation component in the restored signal. Indeed, the excitation of the magnetic core 13 by the H1 component is found at the direct output of the analog / digital converter and it is therefore necessary to remove it so as not to distort the measurement. In doing so, it is necessary not to remove a component which would come from the outside (i.e. from the current to be measured) and which would be close to the H1 excitation component, which would be done by the two loops 50.1 and 50.2 for controlling H1 to 0 (these loops are not capable of distinguishing an internal signal resulting from the excitation of an external signal coming from the current to be measured). The stored values ​​are obtained during an initialization phase, without external current (the current Im is zero), by recording the amplitude values ​​I and Q of the excitation component H1 in memories 55.1 and 55.2 with coefficients determined to dose the cancellation values ​​I and Q of H1 including exactly the values ​​received at the ADC output, components coming from the excitation.

[0074] The adder 45 of the second processing channel has a third input receiving the digital measurement signal Sm. The output of the adder 45 is also connected to an input of a high-pass filter 46 having an output (forming the output of the second processing channel) connected to a second input of the adder 43. The high-pass filter 46 is the strict complement of the low-pass filter 34.

[0075] The operation of the flow valve current sensor is now described.

[0076] The excitation circuit 23 generates a digital excitation signal Se, which is transformed into an excitation current Ie and injected into the measuring winding 14.

[0077] The measurement voltage Ve is acquired and then digitized to produce a digital measurement signal Sm.

[0078] At the output of the second amplifier 35, a digital image signal of the current to be measured Im is obtained. This digital image signal is used to estimate the current to be measured Im after passing through the processing circuit 40.

[0079] The digital image signal is also a digital demagnetization signal Sdm which makes it possible to compensate for the magnetic flux produced by the current to be measured Im.

[0080] Thanks to the adder 36 and by paralleling, the digital demagnetization signal Sdm and the digital excitation signal Se are easily added. The excitation current therefore allows both to excite the magnetic core 13 and to demagnetize the magnetic core 13. The demagnetization current is therefore integrated into the excitation current le.

[0081] The saturation of the magnetic core 13 causes an asymmetry of the measurement voltage Ve, said measurement voltage Ve being constituted by the sum of a frequency component f0 in sin(co0t), corresponding to the fundamental or H1, and a frequency component 2.f0 in cos(2co0t), corresponding to the harmonic component of order 2 or H2.

[0082] It is noted that the amplitude of the partial digital excitation signal of frequency f0 and that of the partial digital excitation signal of frequency 3.f0 are adjusted to obtain an excitation current Ie whose frequency component 3.f0 (or harmonic component of order 3 or H3) is in phase with the frequency component f0.

[0083] The third-order harmonic component of the measurement voltage Ve, obtained after synchronous demodulation via the first synchronous demodulator, is positive. In the case of saturation of the magnetic core 13, the 3rd order harmonic component is attenuated more strongly than the fundamental and the 3rd order harmonic component after synchronous demodulation becomes negative, because the 3rd order harmonic goes into phase opposition with the fundamental.

[0084] Thus, when the excitation current Ic is such that the magnetic core 13 approaches saturation, the ratio of the amplitude of the first-order harmonic component (i.e. the fundamental) and the amplitude of the third-order harmonic component of the measurement voltage Ve evolves until the third-order harmonic component is cancelled, then until the third-order harmonic component is phase-reversed. The operating point corresponding to the cancellation of the third-order harmonic component of the measurement voltage Ve is therefore an optimal operating point of the flux-valve current sensor.

[0085] At the output of the first amplifier 30, a digital error signal is obtained.

[0086] The digital error signal makes it possible to control the excitation current Ic. The control consists of controlling the excitation current Ic to cancel the third order harmonic component of the measurement voltage Ve. The flux valve current sensor thus operates permanently at the optimal operating point. This provides maximum gain from the asymmetries introduced by the current to be measured Im and detectable on the measurement voltage Ve. It is noted that this control by synchronous demodulation of the third order harmonic component is not very sensitive to external electromagnetic disturbances, because all the signals of a frequency different from the frequency 3.f0 generate intermodulation products whose components are filtered by the low-pass filter 34 connected to the output of the third multiplier 33. It is also noted that this control does not need to be very fast, because the variations coming from external parameters (temperature, aging) are relatively slow.

[0087] The demagnetization servo circuit 32 allows the second order harmonic component (the even order harmonic components in general) to be canceled by a zero servo.

[0088] Furthermore, it is understood that the second processing channel (or low-frequency measurement channel which also carries the DC component of the measurement signal) recovers a measurement signal in a relatively direct manner by implementing a current transformer principle and is associated with the first processing channel (or high-frequency measurement channel) by adding these two channels which are adjusted by strictly complementary filters, one high-pass (the filter 46 of the second processing channel) and the other low-pass (the filter 34 of the first processing channel). In the second processing channel, the gain coefficient adjustable by programming which is stored in the memory 42 makes it possible to equalize the gains of the two digital measurement signal processing channels Sm for values ​​close to the filter cut-off frequency (memories 55.1 and 55.2). The digital processing ensures strict linearity of the amplitude response so that the amplitude response is rigorously constant, even at the junction of the cut-off frequencies of the high-pass and low-pass filters. The high-pass filter 46 and the low-pass filter 34 have cut-off frequencies, attenuation slopes as a function of frequency and responses, which are identical. The two signals summed at the end of the filtering have a strictly constant amplitude response as a function of frequency.

[0089] Furthermore, the processing circuit 40 ensures the cancellation of the fundamental component of the excitation by addition in amplitude and complementary phase of the reference signal having the frequency f0 from the internal generator 22. The precise dosage is ensured by the double loop 50.1, 50.2 controlled by the cancellation of the fundamental component. Indeed, the fundamental component resulting from the excitation of the flow valve appears as a “useful” signal in the bandwidth of the first processing channel. To cancel it, it must be measured, its amplitude recorded and reinjected exactly in opposition to that received. However, the exact phase of this fundamental component in the first processing channel is not controlled because of the previous filtering.The second processing channel makes it possible to simultaneously measure the amplitude I (in phase) and the amplitude Q (in quadrature) of the fundamental component of the digital measurement signal Sm, then to high-pass filter these two components before subtracting, from the digital demagnetization signal leaving the multiplier 41, these two filtered components (by the filter 46) to obtain a resulting signal which, after filtering by the filter 44, will constitute the image signal Sim. It is recalled that the amplitude values ​​of the signal H1 received by the analog / digital converter 20 and to be removed from the measured signal are quantified in the initialization phase of the system and that these values ​​are recorded in the memories 55.1 and 55.2 configured in the factory. The aim is not to "erase" from the signal to be measured any continuous component which would be identical to that of the excitation signal and would appear from outside the system during the operating phases.

[0090] The other “parasitic” spectral components (second and third order harmonics), likely to be captured by the current transformer, are in principle cancelled in the excitation signal by the excitation control circuit and the demagnetization control circuit. In fact: - the demagnetization control circuit 32 injects a direct current into the excitation signal to compensate for the magnetic imbalance produced by the current to be measured. The absence of magnetic imbalance results in an absence of even harmonics. - the excitation control circuit 25 cancels the harmonic component of order 3 by a zero control to be on the inflection point of the curve B=f(H) of the magnetic core 13.

[0091] It is noted that the measuring winding 14 is used both to excite the magnetic core 13, to control this excitation, to measure the current to be measured Im, and to compensate for the magnetic flux produced in the magnetic core 13 by the current to be measured Im (demagnetization).

[0092] The pooling of these functions on a single measurement winding 14 makes it possible to simplify, reduce the cost and mass, and facilitate the manufacture of the flow valve current sensor 10.

[0093] It should also be noted that the measurement processing chain is mainly digital: a single FPGA makes it possible to implement a large part of this processing chain. The digitalization of the processing chain increases its reliability and robustness (particularly in terms of temperature), reduces its cost, facilitates its manufacture and improves its industrial reproducibility.

[0094] The improvement in the regulation of the operating point thanks to the demodulation by the first synchronous demodulator also makes it possible to obtain better noise immunity.

[0095] In addition, the use of synchronous demodulators makes it possible to operate at high frequency, which allows a significant bandwidth of the current to be measured Im while maintaining very good immunity to external electromagnetic disturbances.

[0096] Since the measurement is always made in the optimal operating point corresponding to the saturation elbow, the sensitivity of the flow valve current sensor is constant in the temperature range. The accuracy of the flow valve current sensor is therefore good in a large temperature range.

[0097] As indicated earlier, the flow valve current sensor measures a current to be measured Im which flows on a conductor 11.

[0098] The bandwidth is ultimately limited by the sampling frequency and the non-aliasing criterion (Shannon). For example, sampling at 1000s / s and a simple anti-aliasing filter, it is possible to expect a bandwidth of up to 10MHz.

[0099] It is noted that to ensure its proper operation, the transformer 12 is loaded by the excitation source resistance (it is therefore not “no-load”).

[0100] Of course, the invention is not limited to the embodiment described but encompasses any variant falling within the scope of the invention as defined by the claims.

[0101] The structure of the sensor may be different.

[0102] It has been indicated that the flow valve current sensor comprises a part digital and an analog part, and that the analog-to-digital converter and the digital-to-analog converter belong to the analog part. Of course, one could consider that these components are digital components, and even that they are directly integrated into the FPGA (or into a microcontroller or other digital component), so that they would then belong to the digital part.

[0103] When the current to be measured is large, a major difficulty arises. The demagnetization current, which makes it possible to compensate for the magnetic flux produced in the magnetic core 13 by the current to be measured Im, must also be very large. The value of the demagnetization current must be equal to that of the current to be measured divided by the transformation ratio of the transformer 12, associated with the characteristics of the primary winding and the secondary winding. However, the number of turns of the measuring winding 14 is limited by its size and by the resulting inductance. The higher the inductance, the more the rate of change of the demagnetization current (from the digital demagnetization signal Sdm) will be limited, and therefore the more the bandwidth of the flux gate current sensor will be limited. Thus, increasing the number of turns of the secondary winding is not an effective solution.Increasing the demagnetization current is also not an effective solution. Indeed, this involves generating a very high demagnetization voltage under a high demagnetization current, which is complex to achieve with standard components. To solve this difficulty, a current divider bridge can be implemented to reduce the value of the current to be measured.

[0104] Loops 50.1 and 50.2 may have a different structure.

[0105] It is possible to use other gain balancing components between the loops 50.1, 50.2 than the coefficients of the memories 55.1, 55.2.

[0106] It is possible to use another gain balancing device between the two processing channels than the coefficients of the memory 42 associated with the multiplier 4L

[0107] The excitation servo circuit 25 and the demagnetization servo circuit 32 may have a structure different from that described. The same applies to the two processing channels.

Claims

Claims

1. Magnetic current sensor, comprising a magnetic core (13) which extends around a conductor (11) on which a current to be measured (Im) flows, a measurement winding (14), an excitation circuit (23) arranged to generate a digital excitation signal (Se), an acquisition circuit (17) comprising an analog-digital converter (20) and arranged to acquire an analog measurement voltage (Ve) at the terminals of the measurement winding (14) and to produce a digital measurement signal (Sm), a demagnetization control circuit (32) arranged to produce, from the digital measurement signal and after a first low-pass filter (34), a digital demagnetization signal (Sdm) intended to compensate for a magnetic flux produced by the current to be measured, a summator (36) arranged to sum the digital excitation signal and the digital demagnetization signal so as to obtain an injection signal digital (If),and an injection circuit (37) comprising a digital-to-analog converter (38) and arranged to produce an analog excitation current from the digital injection signal and to inject the analog excitation current into the measurement winding, characterized in that the sensor comprises a processing circuit (40) having a first path for processing the demagnetization signal and a second path for processing the digital measurement signal which are connected to a summer (43) to provide an image signal of the current, the second processing path being arranged to cancel in the digital measurement signal at least a part corresponding to a fundamental component of the digital excitation signal and comprising a high-pass filter (46) strictly complementary to the first low-pass filter (34).,

2. Sensor according to claim 1, in which the second processing path comprises a first loop (50.1) for extracting, from the digital measurement signal, a part corresponding to an in-phase amplitude of the fundamental component of the digital excitation signal and a second loop (50.2) for extracting, from the digital measurement signal, a part corresponding to a quadrature amplitude of the fundamental component of the digital excitation signal.

3. The sensor of claim 2, wherein the second path comprises at least one gain balancing component between the loops.

4. Sensor according to claim 1, wherein the digital excitation signal is obtained from a digital reference signal of frequency f0 and a digital reference signal of frequency 3.f0.

5. Sensor according to claim 4, in which the excitation circuit comprises an excitation servo circuit (25) which is arranged to servo the digital excitation signal so as to cancel a frequency component 3.f0 of the analog measurement voltage.

6. Sensor according to claim 5, in which the excitation servo circuit comprises a first multiplier (27) arranged to multiply the digital measurement signal by a digital demodulation signal of frequency 3.f0.

7. A sensor according to claim 6, wherein the excitation servo circuit further comprises an integrator (29), a first amplifier (30) connected to an output of the integrator and arranged to produce a digital error signal, and a second multiplier (28) which multiplies a raw digital excitation signal (Seb) by the digital error signal to obtain the digital excitation signal.

8. Sensor according to claim 7, in which the demagnetization servo circuit comprises a third multiplier (33) arranged to multiply the digital measurement signal by a digital demodulation signal of frequency 2.f0.

9. A sensor according to claim 8, wherein the degaussing servo circuit further comprises a second amplifier (35) connected to an output of the first low-pass filter (34) and arranged to produce the digital degaussing signal.

10. Sensor according to claim 1, comprising at least one gain balancing member between the two processing channels

Images