A fluxgate current sensor makes use of the property of a magnetic material forming a magnetic core to become saturated as from a certain level of magnetic excitation. With reference to
With reference to
Fluxgate current sensors are preferred in certain applications. This applies in particular for measuring a current flowing in a solid state power controller (SSPC) type cut-off member for user equipment, or for measuring a current flowing in a power electronic module (PEM) type energy converter member connected to a phase of a motor.
Conventional fluxgate current sensors suffer from certain weaknesses. In particular, peak detection is relatively sensitive to external electromagnetic disturbances, which can consequently degrade the accuracy with which the current for measuring is measured. Furthermore, peak detection does not give accurate information suitable for servo-controlling on the saturation bend of the magnetic core. This is particularly problematic when the external electromagnetic disturbances are large, e.g. in the environment of a switch mode power converter. The same applies when the conventional fluxgate current sensor 1 is constrained by a difficult thermal environment causing the intrinsic characteristics of the magnetic core to drift.
Naturally, this problem of accuracy is made worse when the frequency range of the current Im for measuring increases (when the current for measuring is an alternating current), or when the operating temperature range of the fluxgate current sensor increases.
An object of the invention is to improve the accuracy of a fluxgate current sensor.
In order to achieve this object, there is provided a fluxgate current sensor comprising a magnetic core that extends around a conductor conveying a current for measuring, an excitation coil, an excitation module connected to the excitation coil in order to generate an excitation voltage across the terminals of the excitation coil, a measurement coil conveying an induced measurement current, and a measurement synchronous demodulator connected to the measurement coil for the purpose of multiplying the induced measurement current by a demodulation signal of frequency 2f0 in order to obtain an image voltage that is the image of the current for measuring. The excitation voltage is obtained from a first voltage of frequency f3 and from a second voltage of frequency 3f0. The fluxgate current sensor further includes a servo-control coil conveying an induced servo-control current, and a servo-control synchronous demodulator connected to the servo-control coil for the purpose of multiplying the induced servo-control current by a demodulation signal of frequency 3f0. The servo-control synchronous demodulator is connected to the excitation module in order to servo-control the excitation voltage in such a manner as to cancel the component of frequency 3f0 in the induced servo-control current.
The fluxgate current sensor of the invention enables the excitation voltage to be servo-controlled accurately in such a manner that the fluxgate current sensor operates at its optimum operating point, corresponding to the bend 9 in the transfer function shown in
The invention can be better understood in the light of the following description of a particular non-limiting embodiment of the invention.
Reference is made to the accompanying drawings, in which:
With reference to
The fluxgate current sensor 10 includes a transformer having a magnetic core 12 with four coils wound around the magnetic core 12. The magnetic core 12 extends around the conductor 11.
The four coils comprise an excitation coil 13, a measurement coil 14, a servo-control coil 15, and a demagnetization coil 16. The measurement coil 14 and the servo-control coil 15 are both constituted in this example by a respective pair of windings. The excitation coil 13 is used for exciting the magnetic core 12.
The fluxgate current sensor 10 also has a low frequency generator 18, a high frequency generator 19, and an excitation module 20 that is connected to the excitation coil 13.
The low frequency generator 18 transmits a signal C to two demodulators 42, 43, 34, and 35. The high frequency generator 19 transmits to the excitation module 20 a signal 1H at the frequency f0 (as sin(ω0t), with ω0=2πf0) together with a signal 3H at the frequency 3f0 (as sin(3ω0t), with ω0=2πf0).
With reference to
The first amplifier block 21 receives the signal 1H at the frequency f0, the second amplifier block 22 receives the signal 3H at the frequency 3f0, and the gain block 24 receives a gain setpoint Cg. The summing block 23 sums the outputs from the first amplifier block 21 and from the second amplifier block 22.
The excitation module 20 thus serves to generate an excitation voltage Ve from a first voltage V1 of frequency f0 and from a second voltage V2 of frequency 3f0.
The measurement coil 14 is used for measuring the current Im. The fluxgate current sensor 10 includes a measurement synchronous demodulator 32, a first lowpass filter 33, and a first unit gain block 34 which are connected in series with the measurement coil 14. The first unit gain block 34 produces a gain that can be selected to be equal to 1 or to −1. The first unit gain block 34 receives a switching signal C from the low frequency generator 18. The measurement synchronous demodulator 32 is connected to a first selector block 35. The first selector block 35 receives a switching signal C from the low frequency generator 18 and a signal 2H of frequency 2f0 from the high frequency generator 19. The first selector block 35 produces a demodulation signal of frequency 2f0 written respectively 2H− or 2H depending on whether it is in phase with or in phase opposition with the signal 2H at the frequency 2f0. The demodulator signal 2H−/2H is a cos(2ω0t) signal, where ω0=2πf1. The measurement synchronous demodulator 32 multiplies, by the demodulation signal 2H−/2H, an induced measurement current Iim flowing in the measurement coil 14 and induced by the current Im for measuring. A first proportional integral corrector 37 connected to the first unit gain block 34 produces an image voltage Vi that is the image of the current Im for measuring.
With reference to
With reference to
The servo-control coil 15 conveying an induced servo-control current Iia is used to control the excitation voltage Ve. The fluxgate current sensor 10 has a servo-control synchronous demodulator 40, a second lowpass filter 41, and a second unit gain block 42, which are connected in series with the servo-control coil 15. The second unit gain block 42 produces gain that can be selected to be equal to 1 or to −1. The second unit gain block 42 receives a DC signal C from the low frequency generator 18. The servo-control synchronous demodulator 40 is connected to a second selector block 43. The second selector block 43 receives a switching signal C from the low frequency generator 18 and a signal 3H of frequency 3f0 from the high frequency generator. The second selector block 43 produces a demodulation signal 3H−/3H of frequency 3f0 that is in phase or in phase opposition. The demodulation signal 3H−/3H is a sin(3ω0t) signal, where ω0=2πf0. The servo-control synchronous demodulator 40 multiplies the servo-control induced current Iia by the demodulation signal 3H−/3H. A second proportional integral corrector 44 is connected between the second unit gain block 42 and the excitation module 20. The excitation voltage Ve is thus servo-controlled so as to cancel the component of frequency 3f0 in the induced servo-control current Iia.
The servo-control synchronous demodulator 40 is similar to the measurement synchronous demodulator 32 shown in
The demagnetization coil 16 is connected to the output of the first proportional integral corrector 37 via a voltage-to-current converter 45. A demagnetization current Id flows in the demagnetization coil 16 in order to demagnetize the magnetic core 12.
The image voltage Vi of the current Im for measuring at the output from the first proportional integral corrector 37 is applied as input to the voltage-to-current converter 45. The voltage-to-current converter 45 produces the demagnetization current Id from the image voltage Vi of the current Im for measuring. The current output by the first unit gain block 34 forms a current setpoint that is applied as input to the first proportional corrector 37. The demagnetization current Id is thus servo-controlled from the image voltage Vi of the current Im for measuring so as to compensate the magnetic flux produced by the current Im for measuring.
The characteristics of the excitation voltage Ve are described below in greater detail.
With reference to
The third order harmonic component of the induced servo-control current Iia flowing in the servo-control coil 15, as obtained after synchronous demodulation via the servo-control synchronous demodulator 40, is positive. When the magnetic core 12 saturates, the third order harmonic component is attenuated more strongly than the fundamental, and the third order harmonic component after synchronous demodulation becomes negative since the third order harmonic goes into phase opposition with the fundamental.
Thus, when the excitation voltage Ve is such that the magnetic core 12 is approaching saturation, the ratio of the amplitude of the first order harmonic component to the amplitude of the third order harmonic component of the induced servo-control current Iia varies until the third order harmonic component is cancelled and then until the phase of the third order harmonic component is inverted. The operating point corresponding to cancellation of the third order harmonic component of the induced servo-control current Iia is thus an optimum operating point for the fluxgate current sensor 10. This optimum operating point corresponds to the bend 9 in the curve for the transfer function shown in
Servo-controlling the excitation voltage Ve thus consists in controlling the excitation voltage Ve before it is applied to the terminals of the excitation coil 13 so as to cancel the third order harmonic component of the induced servo-control current Iia. The fluxgate current sensor 10 thus operates continuously on its optimum operating point. This obtains maximum gain for the asymmetries introduced by the current Im for measuring and that are detectable in the induced measurement current Iim. It should be observed that this servo-control by synchronous demodulation of the third order harmonic component is relatively insensitive to external electromagnetic disturbances, since all signals of frequency other than the frequency 3f0 generate intermodulation products having components that are filtered by the first lowpass filter 33 connected to the output from the measurement synchronous demodulator 32. It should also be observed that this servo-control has no need to be very fast since variations that come from external parameters (temperature, aging) are relatively slow.
As can be seen in
The advantage of using such an excitation voltage Ve constituted by such a first voltage V1 and such a second voltage V2 can be seen from the following.
With reference to
X(t)=(4E/π)[sin(ωt)+(1/3)sin(3ωt)+(1/5)sin(5ωt)+ . . . ]
with ω=2π/T.
The spectral decomposition of a triangular signal 51 of amplitude E and of period T is:
X(t)=(8E/π2)[sin(ωt)−(1/32)sin(3ωt)+(1/52)sin(5ωt)− . . . ]
with ω=2π/T.
The triangular signal 51 thus has a negative third order harmonic component.
With reference to
The phenomenon of the magnetic core 12 saturating causes the waveform of the induced servo-control current Iia to approximate to the waveform of a squarewave signal. On approaching saturation, the amplitude of the third order harmonic component of the induced servo-control current Iia thus tends towards zero, and then becomes negative, to match the decomposition of a squarewave signal. Control of the excitation voltage Ve is thus defined by servo-control on the point where the third order harmonic component of the induced servo-control current Iia cancels.
This particular regulation point can be found by calculation.
The first voltage V1 presents the following spectral decomposition, in sine functions:
X1(t)=(4E1/π)[sin(ω0t)+(1/3)sin(3ω0t)+(1/5)sin(5ω0t)+ . . . ]
The second voltage V2 presents the following spectral decomposition in sine functions:
X2(t)=(4E2/π)[sin(ω0t)+(1/3)sin(9ω0t)+(1/5)sin(15ω0t)+ . . . ]
The third order harmonic component of the sum of the spectral decomposition of the first voltage V1 plus the second spectral decomposition of the second voltage V2 is equal to:
(4/π)[(E1/3)sin(3ω0t)+E2 sin(3ω0t)]
The third order harmonic component is thus zero when E1=−3E2.
The servo-control of the excitation voltage Ve consists in controlling the amplitude of the second voltage V2. Action is thus taken on the amplitude of the second voltage V2 in order to control the excitation voltage Ve so that the component of frequency 3f0 of the induced servo-control current Iia cancels.
Prior to the magnetic core 12 saturating, the looked-for control of the excitation voltage Ve thus needs to include a relative level of excitation by the third order harmonic component that gives a result that is slightly negative after demodulation by the servo-control synchronous demodulator 40 so as to make it possible, on saturation, to cross through zero and then become negative. Adjustment should be made on this particular point in order to find the best gain compromise for the transfer function that corresponds to the optimum operating point and to the bend 9 in the transfer function shown in
It should be observed that the signal 2H of frequency 2f0 and the signal 3H of frequency 3f0 are periodically phase inverted, specifically at a frequency of 10 kilohertz (kHz). The shifts in phase opposition are averaged by the first proportional integral corrector 37 and by the second proportional integral corrector 44, and they are thus cancelled. This function of phase inversion at 10 kHz serves to compensate for unbalances between the measurement and servo-control channels, as results from imperfections in the analog components used. This serves to improve the accuracy with which the fluxgate current sensor 10 is set on the optimum operating point, and thus the overall accuracy of the fluxgate current sensor 10 itself.
With reference to
With reference to
The fluxgate current sensor 10 presents certain advantages over existing fluxgate current sensors.
The improvement in the regulation of the operating point by means of the demodulation by the servo-control synchronous demodulator 40 serves to obtain better noise immunity.
Furthermore, the use of synchronous demodulators makes it possible to operate at high frequency, thereby enabling the current Im for measuring to have a large bandwidth, while conserving very good immunity to external electromagnetic disturbances.
Since measurement is always performed in the optimum operating point corresponding to the saturation bend 9, the sensitivity of the fluxgate current sensor 10 is constant over the temperature range. The accuracy of the fluxgate current sensor 10 is thus good over a large temperature range.
The fluxgate current sensor 10 is insensitive to manufacturing drifts in certain characteristics of the magnetic core 12. In particular, the magnetic permeability characteristics of conventional magnetic cores, which are not intended for use in their saturation zone, can be very dissimilar. Having accurate servo-control on the operating point serves to correct for such dispersion. It is thus possible to use standard magnetic cores, thereby reducing industrial constraints on manufacturing the magnetic core, and thus on the cost of the fluxgate current sensor 10. By way of example, it is thus possible to use a ferrite magnetic core of the kind conventionally used in power supply converters and in certain radio functions.
It should also be observed that the fluxgate current sensor 10 is of an architecture that can be transposed to digital solutions in order to improve the reproducibility and thus the manufacture of the fluxgate current sensor 10. Thus, the excitation voltage Ve may be produced by a fast analog-to-digital converter. Similarly, the measurement and servo-control synchronous demodulators 32 and 40 may be implemented by means of a digital computer (e.g. a field programmable gate array (FPGA). Demagnetization may also be done via a digital-to-analog converter.
It should be observed that it would be possible to use not one but rather two magnetic cores. The second magnetic core would receive the same excitation as the measurement circuit, but in phase opposition, in order to cancel the flux directly in the conductor conveying the current for measuring.
Naturally, the invention is not limited to the embodiment described but covers any variant coming within the ambit of the invention as defined by the claims.
Number | Date | Country | Kind |
---|---|---|---|
16 62794 | Dec 2016 | FR | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP2017/083671 | 12/19/2017 | WO | 00 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2018/115032 | 6/28/2018 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
3435337 | Inouye et al. | Mar 1969 | A |
20080252289 | Lenglet | Oct 2008 | A1 |
20150108967 | Barczyk | Apr 2015 | A1 |
Number | Date | Country |
---|---|---|
0137347 | Apr 1985 | EP |
1746430 | Jan 2007 | EP |
2942880 | Sep 2010 | FR |
WO 91013366 | Sep 1991 | WO |
WO 2014047644 | Mar 2014 | WO |
WO 2016016038 | Feb 2016 | WO |
Number | Date | Country | |
---|---|---|---|
20190317135 A1 | Oct 2019 | US |