JP2021156700A - Flux gate magnetic field sensor - Google Patents

Abstract

To provide an orthogonal flux gate magnetic field sensor of a bias switching method having extremely good noise characteristics and output offset characteristics.SOLUTION: An orthogonal flux gate magnetic field sensor obtains a sensor output by detecting a pickup signal induced in a coil wound on an elongated magnetic body to which a magnetic field to be detected is applied through which an exciting current flows which is a DC bias current whose polarity is periodically reversed is superimposed on an AC current. During a stabilization period, which is a predetermined period of time beginning at the same time as respective switching of a polarity of the DC bias current and ending prior to the next switching, the detection signal is replaced by a signal that is not derived from a pickup signal and stabilized.SELECTED DRAWING: Figure 3

Description

The present invention relates to a fluxgate magnetic field sensor, and more particularly to a bias switching type fluxgate magnetic field sensor with reduced noise and output offset.

The fundamental wave type orthogonal fluxgate sensor, which is a kind of magnetic field sensor, is a small fluxgate sensor in which the exciting magnetic field of the magnetic core and the magnetic field to be measured are orthogonal to each other. By superimposing a DC bias current on an alternating current as an exciting current that generates an exciting magnetic field, noise reduction, sensitivity improvement, magnetic field detection with a fundamental wave signal, and the like are realized. In the fundamental wave orthogonal fluxgate, a large offset may occur in the output due to the magnetic anisotropy of the magnetic core, and a method called a bias switching method for suppressing this has been proposed (Patent Document 1). In the bias switching method, the positive and negative polarities (hereinafter referred to as exciting current polarities) of the above-mentioned DC bias current (and, in some cases, the DC bias current and the alternating current superimposed on the DC bias current) are switched to alternating currents in a predetermined cycle. After that, the signal detected under the polarity of each exciting current is subtracted (when only the polarity of the DC bias current is switched) or added / averaged (when the phase of the alternating current is also switched). The offset and its drift are offset by both polarities for stabilization. In this specification, switching means a switching operation at the moment of switching the positive and negative of the exciting current, bias switching means an operation of continuously switching the exciting current, and the bias switching method is It means a method of averaging the output offsets of both polarities of DC current by using bias switching. The switching cycle is the period from the switching from the positive electrode property to the negative electrode property to the next switching from the positive electrode property to the negative electrode property until the polarity is switched in the same direction next time. That is. The switching cycle is equal to the cycle of the switching signal when switching is performed by controlling the switching signal. Further, the next switching after a certain switching is performed when the direction of the polarity switching is opposite and the half cycle period of the switching cycle elapses.

When the polarity of the exciting current is switched in the bias switching method, transient noise (transient) in which magnetic noise generated when the magnetization in the magnetic core suddenly reverses and induced noise due to a sudden change in the current are mixed. Signal) is generated. Since this noise is not guaranteed to have symmetry or repeatability between the excitation poles, it leaks to the output without being offset by the averaging of the outputs of both polarities by bias switching, and the noise characteristics of the fundamental wave orthogonal fluxgate sensor. And the stability of the output offset will be deteriorated. In particular, the magnetic noise caused by the magnetic core changes depending on the magnitude of the exciting current. Therefore, for example, when the exciting current is reduced to reduce the power consumption, the noise is remarkably increased. In addition, the magnetic noise changes depending on the timing of reversing the polarity of the exciting current, and when the switching timing of the exciting current is slightly deviated due to the temperature and aging of the circuit, the sensor noise characteristics deteriorate. It also causes a large variation in the output offset of the sensor. As described above, the transient noise associated with bias switching becomes a big constraint when the sensor performance is stabilized under a wide range of conditions.

To solve such a problem, a method has been proposed in which a transient signal is cut off by completely cutting off the detection circuit at the moment of switching to open the state (Patent Document 2). Further, a method of blocking a transient signal by holding and outputting a detection signal at the time of operation immediately before switching by using a sample & hold circuit has also been proposed (Non-Patent Document 1).

Patent No. 4209114 Japanese Unexamined Patent Publication No. 2018-096690

N. Murata. Et al .. Fundamental Mode Orthogonal Fluxgate Magnetometer Applicable for Measurements of DC and Low-Frequency Magnetic Fields. IEEE Sensors Journal. Vol. 18 (7), 2018

The method of dealing with the adverse effect on noise performance due to bias switching by the prior art is to disconnect the detection circuit to cut off the transient signal due to bias switching and put it in the open state or the state where the immediately preceding signal is held. This is intended to improve noise performance. However, it is difficult to guarantee the stability of circuit operation by these methods, and it is still a problem to provide stable sensor performance. Regarding the method of opening the circuit, for example, if the detection circuit is temporarily cut off and becomes open, the feed pack loop may become unstable and the output may be disturbed depending on the conditions, resulting in sensor malfunction. There is. In the case of the method of holding the detection signal at the time of operation immediately before the bias is switched by the sample & hold circuit, the level of the held signal is held unless the operation timing of the circuit and the hold source signal are the same conditions for each sample. It fluctuates with each operation. Therefore, the level fluctuation of the held signal may be mixed in the output to cause a drift of the output offset of the sensor. As described above, in this method, the sensor output is affected by the signal state of the hold source, and a slight deviation in the hold timing becomes jitter, which causes an increase in noise and an offset fluctuation. .. In order for the fundamental wave orthogonal fluxgate sensor to operate stably over a long period of time and not only reduce sensor noise but also stabilize the output offset, it is necessary to process the transient signal associated with bias switching in a more robust manner. .. However, no such method existed.

Furthermore, even if the original signal component is output by removing the transient part in the bias switching type sensor, it is included in the induced signal obtained under the positive and negative exciting currents. It is unavoidable that the imbalance remains as an output offset. It is possible to correct this output offset by adding an operational amplifier and an appropriate voltage source to the sensor output, but it is necessary to add an operational amplifier, which complicates the circuit configuration. ..

The above problem is solved by the present invention having the following characteristics. That is, the present invention is from an elongated magnetic material to which a detected magnetic field is applied, in which an exciting current in which a DC bias current whose polarity is periodically switched between positive and negative properties is superimposed on an AC current is passed. The pickup signal from the pickup coil wound around the magnetic core is amplified by an amplifier, and the detection signal is output by detecting the amplified pickup signal, which is started at the same time as each switching of the polarity of the DC bias current. During the stabilization period, which is a predetermined period that ends before the next switching of the switching, the detection signal is stabilized by replacing it with a signal that does not originate from the pickup signal, and the magnetic field detected from the stabilized detection signal. It is characterized in that a sensor output is generated based on a magnetic field detection signal from which uncorresponding signals are removed.

The present invention stabilizes the detection signal by cutting the circuit at any point between immediately after the detector and immediately before the output circuit during the stabilization period, and grounding immediately after the cut point. be able to. The present invention stabilizes the detection signal by cutting the circuit at any point between immediately after the pickup coil and immediately before the detector during the stabilization period, and grounding immediately after the cut point. be able to. In the present invention, the circuit is cut at any point between immediately after the detector and immediately before the output circuit during the stabilization period, and the voltage of a predetermined voltage is immediately after the cut point. It can be done by connecting to the source. In that case, the voltage of the predetermined voltage source can be determined so that the output offset remaining in the sensor output is minimized.

The present invention can be configured such that the stabilization period begins with switching and ends in half the period until the next switching of said switching. The present invention can be configured to synchronously detect an amplified pickup signal with reference to a reference signal whose frequency is synchronized with that of an alternating current. The present invention may be configured such that the polarity of the alternating current is periodically reversed at the same timing as the switching of the direct current bias current. Further, the present invention can also be configured as a closed loop configuration in which the output circuit inputs a footback signal obtained by integrating the magnetic field detection signal to the pickup coil with a polarity that cancels the magnetic field detection signal, and generates a sensor output from the feedback signal.

The present invention replaces the detection signal with a signal not derived from the pickup signal during the stabilization period, which is a predetermined period that starts at the same time as each switching of the polarity of the DC bias current and ends before the next switching of the switching. By stabilizing the voltage, it is possible to effectively remove transient noise and provide a bias switching type flux gate magnetic field sensor with further reduced noise. The present invention has an effect that a bias switching type flux gate magnetic field sensor with further reduced noise can be provided by using a signal not derived from the pickup signal that replaces the detection signal as the ground potential, and by using a simple circuit configuration. Is obtained. The present invention has the effect that an arbitrary voltage can be superimposed on the sensor output by setting a signal not derived from the pickup signal that replaces the detection signal to a predetermined voltage. Then, if the predetermined voltage is a voltage that cancels the output offset, the effect that the output offset can be made substantially 0 can be obtained.

It is a schematic circuit diagram of the conventional bias switching type fluxgate magnetic field sensor. It is a schematic circuit diagram of the fluxgate magnetic field sensor of the bias switching type which reduced the noise by the conventional circuit disconnection. It is a schematic circuit diagram of the fluxgate magnetic field sensor which concerns on 1st Embodiment of this invention. It is a schematic circuit diagram of the fluxgate magnetic field sensor which concerns on the 2nd Embodiment of this invention. It is a figure which shows the schematic structure of the fluxgate magnetic field sensor. It is explanatory drawing of the magnetization of the magnetic core of the fluxgate magnetic field sensor in the positive electrode property. It is explanatory drawing of the magnetization of the magnetic core of the fluxgate magnetic field sensor in the negative electrode property. It is a waveform diagram of the detection signal when the transient signal is removed. It is a graph of the measured value of the noise density of the detection signal when the transient signal is not removed. It is a graph of the measured value of the noise density of the detection signal when the transient signal is removed. It is a waveform diagram of the exciting current when the phase of bias switching is changed. FIG. 12A is a graph of the measured value of the sensor noise amount before removing the transient signal, and FIG. 12B is a graph of the measured value of the output offset before removing the transient signal. It is a graph of the measured value of the sensor noise amount when the transient signal is not removed and when it is removed. It is a graph of the measured value of the output offset when the transient signal is removed. It is a waveform diagram of the detection signal when the detection signal is connected to a constant voltage source. It is a graph of the measured value of the output offset when the voltage of a constant voltage source is changed.

(Structure of conventional bias switching fluxgate magnetic field sensor)
The present invention will be described below with reference to the drawings. First, the basic principle of the conventional bias switching type fluxgate magnetic field sensor as shown in Patent Document 1, which is the premise of the present invention, will be described. FIG. 1 is a schematic circuit diagram of a conventional fluxgate magnetic field sensor 100. In this circuit, synchronous detection is used as a typical detection method. In addition, a closed loop configuration is adopted in which the sensor is operated in a region where the characteristics of the magnetic material are good by passing a current that cancels the detected magnetic field through the pickup coil.

The fluxgate magnetic field sensor 100 includes a DC superimposed AC magnetometer 101, an exciting switching unit 102, a sensor unit 103, a voltage follower 104, an amplifier 105, a synchronous detector 106, a low-pass filter 107, an integrator 108, a feedback resistor 109, and an output terminal. It is composed of 110. An alternating current source generates an alternating current having a frequency of fHz, but usually also outputs a reference signal (not shown) for synchronous detection in a predetermined phase relationship with the alternating current.

The DC superimposed AC excitation unit 101 is a circuit that provides an exciting current in which a DC bias current is superimposed on an AC current. The DC superimposition AC excitation unit 101 and the magnetic field strength switching unit 102 in the subsequent stage form a bias switching type excitation circuit.

The magnetic field switching unit 102 is a circuit that periodically switches the polarity of the exciting current flowing through the sensor unit 103. As a result, the polarity of the DC bias current component is periodically switched, and the exciting current by the bias switching method is supplied to the magnetic core of the sensor unit 103. The exciting pole switching unit 102 has _{a clock having a frequency of f bs} Hz that determines the period of polarity switching, and a semiconductor switch element that switches the polarity is driven by the switching signal generated by the clock, and the exciting current is generated. The polarity is switched periodically. A clock with a frequency of f _bs Hz is typically used by dividing the frequency of the AC component of the exciting current by a fraction of an integer. As shown in FIG. 1, when the exciting magnetic pole switching unit 102 switches the output from the DC superimposed AC exciting unit 101, the phase of the AC current also periodically (at the same timing) in synchronization with the switching of the polarity of the DC bias current. As a result, the offset component due to the magnetic anisotropy appears in the opposite polarity and the sensitivity component of the magnetic field appears in the same polarity with respect to each polarity of the exciting current (DC bias current).

The exciting current is a DC bias current superimposed on an alternating current. When referring to the DC bias current component, other expressions such as DC bias current, DC component of exciting current, and DC exciting current may be used depending on the context, and when referring to AC current component, Depending on the context, other expressions such as alternating current and alternating current components of the exciting current may be used.

The sensor unit 103 is a sensor element that detects an external magnetic field to be measured, and has a magnetic core made of an elongated magnetic material to which the detected magnetic field is applied, and a pickup coil wound around the magnetic core.

The voltage follower 104 is a component for accurately transmitting the pickup signal from the sensor unit 103 to the subsequent stage by converting the impedance. The amplifier 105 is a component that amplifies the output from the voltage follower 104 to an appropriate level. In FIG. 1, the output from the voltage follower 104 passes through a capacitor to take out only the AC component, and is sent to the amplifier 105 in the subsequent stage. The voltage follower 104, the amplifier 105, and the capacitor between them are additional configurations for accurately transmitting the pickup signal at an appropriate level, and are not necessarily essential.

The synchronous detector 106 is a component that performs synchronous detection on the pickup signal from the sensor unit 103 output from the amplifier 105. The synchronous detector 106 determines the sensor output by synchronously detecting the pickup signal, which is the induced voltage induced in the pickup coil of the sensor unit 103, with reference to the reference signal whose frequency is synchronized with the AC current in the exciting current. Generate a detection signal that will be different. The synchronous detector 106 detects the pickup signal from the sensor unit 103 according to the phase relationship with the reference signal having the same frequency as the AC component of the exciting current of the DC superimposed AC excitation unit 101. ..

Here, synchronous detection can be performed by multiplying the detected signal by the sine wave when the signal is conveyed by a sine wave. Since the external magnetic field is modulated by the AC component of the exciting current, synchronous detection can be performed by multiplying the pickup signal by the sine wave, which is the AC component of the exciting current. A square wave is used for multiplication instead of a sine wave, which makes the circuit easy to configure. That is, since the square wave has an amplitude of 1, -1, the multiplication result is that the pickup signal is passed through with the same polarity when the amplitude of the square wave is 1, and the pickup signal is passed when the amplitude of the square wave is -1. Synchronous detection can be performed by reversing the polarity and passing through. Here, in order to perform gate control (control of whether or not the polarity of the pick-up signal to be passed is inverted) by the sign of the reference signal, a square wave whose frequency and phase are synchronized with the AC component of the exciting current is used as the reference signal. I'm using it.

A signal having the same frequency as the AC component of the exciting current is input to the synchronous detector 106 as a reference signal, whereby the phase relationship with the pickup signal to be detected is determined, for example, a half cycle of the reference signal ( For example, when the reference signal is a square wave and the amplitude is 1, the pickup signal is passed through with the same polarity, and the remaining half period of the reference signal (for example, when the reference signal is a square wave and the amplitude is -1) is the pickup signal. By reversing the polarity of, the detection is performed in synchronization with the AC component generated by the DC superimposition AC excitation unit 101.

The low-pass filter 107 is a component of a circuit that averages the detection signals output from the synchronous detector 106. As a result, the magnetic field detection signal indicating the magnitude of the magnetic field to be detected is extracted from the detection signal. In this example, the phase of the AC current is also inverted in synchronization with the switching of the polarity of the DC bias current, and in the pickup signal, the offset component due to magnetic anisotropy for each polarity of the DC bias current is Since it appears in the opposite polarity and the sensitivity components of the magnetic field appear in the same polarity, a magnetic field detection signal that removes the signal that does not correspond to the magnetic field detected from the detection signal can be obtained by adding (realizing by averaging) the detection signals. ..

The integrator 108 is a component of a circuit that integrates the magnetic field detection signal output from the low-pass filter 107 when the system is operated in a closed loop configuration. The output of the integrator 108 applies negative feedback by a magnetic field to the magnetic core of the sensor unit 103 through the pickup coil of the sensor unit 103 via the feedback resistor 109 described later. As a result, the magnetic field detection signal that is the input of the integrator 108 represents the deviation of this feedback, and the integrator 108 controls the magnetic field detection signal to approach zero.

The feedback resistance 109 is a resistance through which the feedback signal passes when the feedback signal output from the integrator 108 is negatively fed back and input to the pickup coil of the sensor unit 103. Due to the negative feedback, a magnetic field that cancels the external magnetic field to be measured is generated in the pickup coil of the sensor unit 103. The current flowing by the feedback signal corresponds to the external magnetic field to be measured, and the current value detected as a voltage by the feedback resistor 109 is the sensor output indicating the magnitude of the external magnetic field to be measured.

The output terminal 110 is a node that outputs a sensor output obtained by converting the current flowing through the feedback resistor 109 into a voltage representing the magnitude of an external magnetic field.

(Operation of conventional fluxgate magnetic field sensor)
Next, the operation of the conventional fluxgate magnetic field sensor 100, which is the premise of the present invention, will be described. FIG. 5 is a diagram showing a schematic structure of a fluxgate magnetic field sensor. FIG. 5A shows a schematic diagram of the sensor unit 103. An amorphous magnetic wire extends in the vertical direction, and a pickup coil is wound around it. The circumferential direction of the amorphous magnetic wire is the x direction, and the longitudinal direction is the z direction. The detected external magnetic field flows into the amorphous magnetic wire in the z direction. When the exciting current flows from the bottom to the top in the figure, the magnetic field due to the exciting current is in the x direction and is orthogonal to the direction of the detected external magnetic field. FIG. 5B shows a waveform in which a DC bias current is superimposed on an AC current output from the DC superimposed AC excitation unit 101. Since the excitation pole switching has not been performed yet, it is in a one-biased state. The in amorphous magnetic wire has uniaxial anisotropy K _u, the excitation magnetic field, anisotropy, magnetization J _s at the position where the energy of the external magnetic field is minimized so that the vibrating. The z-axis projection of this vibration induces a voltage in the pickup coil, which can be taken out as a pickup signal. The influence of an external magnetic field is modulated on this voltage, and the output can be taken out by synchronous detection. Even when there is no external magnetic field, an offset component is generated due to the influence of anisotropy, but if the polarity of the exciting pole is reversed by reversing the polarity of the DC bias current, the magnitude of the offset component due to uniaxial symmetry. The polarity of the offset will also be reversed while maintaining the above. By utilizing this property and periodically inverting the magnetic field strength and adding (averaging) the outputs, it is possible to suppress the output offset that appears in the sensor output. In FIG. 5 (c), the exciting current output from the exciting magnetic field switching unit 102 for such inversion, whose polarity is periodically switched, and the DC bias current superimposed on the alternating current, is shown. The waveform is shown.

Next, the waveform of the signal picked up from the pickup coil by the sensor unit 103 (waveform of the pickup signal) and the vector diagram of the magnetization will be shown to explain the waveform. In the following, the term "positive electrode property" refers to the case where the DC bias current is positive electrode property, and the term "negative electrode property" refers to the case where the DC bias current is negative electrode property. FIG. 6 is an explanatory diagram of the magnetization of the sensor portion of the fluxgate magnetic field sensor in the positive electrode property. FIG. 6A shows the waveform of the pickup signal. When the z-axis projection of the magnetization by the exciting current is J _s · sin (θ (t)), its time derivative is induced in the pickup coil. Waveform of the induced pickup signal includes an offset component caused by uniaxial anisotropy K _u of the amorphous magnetic wire. In (b) of FIG. 6, the external magnetic field H _ex, uniaxial anisotropy K _u of the amorphous magnetic wire, which illustrates the direction of magnetization J _s of the amorphous magnetic wire to the exciting magnetic field H (t) is shown There is. The external magnetic field H _ex can be regarded as zero because of the closed loop control. The x-axis, the excitation magnetic field _{H (t) = H ac ·} sin (2πft) + H dc (H ac AC excitation field, H _dc is the DC bias magnetic field) is applied, also different with even z-axis component Directional K _u exists. The external magnetic field H _ex , the anisotropic K _u , and the exciting magnetic field H (t) generate _{the magnetization J s of the amorphous magnetic wire.} Assuming that the angle of the magnetization J _{s from} the circumferential direction is the angle θ ₀ , the exciting magnetic field H (t) vibrates due to the AC component, so that the angle θ ₀ vibrates at the same frequency. The exciting magnetic field H (t) is shown in FIG. 6 (c). When the exciting magnetic field H (t) has a positive peak, the angle θ ₀ becomes smaller due to the influence of it, and when the exciting magnetic field H (t) has a negative peak, the influence of it becomes smaller, so that the angle θ _{0 becomes smaller.} Becomes larger. That is, when the exciting magnetic field H (t) has a positive peak at (1), the magnetization J _s is most attracted in the direction of the exciting magnetic field H (t), and when the exciting magnetic field H (t) has a negative peak at (2). Sometimes the magnetization J _s is farthest from the direction of the exciting magnetic field H (t). The components of a z-axis component of the magnetization J _s · sin (θ ₀₎ is to generate a pickup coil magnetic flux interlinking to generate a pickup signal to the pickup coil.

Next, the case where the polarity of the DC bias current is reversed will be described. FIG. 7 is an explanatory diagram of the magnetization of the sensor portion of the fluxgate magnetic field sensor in the negative electrode property. In this example, the polarity of the DC bias current is reversed, and the phase of the alternating current on which it is superimposed is also reversed. Figure 7, similarly to FIG. 6, the waveform of the pick-up signal in FIG. 7 (a), the direction of magnetization J _s of amorphous magnetic wire (b) of FIG. 7, the excitation magnetic field H in FIG. 7 (c) ( t) is shown. In FIG. 7 (c), contrary to FIG. 6 (c), when the exciting magnetic field H (t) has a positive peak, the effect of the exciting magnetic field H (t) becomes smaller, so that the angle θ ₀ becomes larger and the excitation is performed. When the magnetic field H (t) has a negative peak, the angle θ ₀ becomes smaller due to the influence of the peak. That is, when the exciting magnetic field H (t) is (1) and has a negative peak, the magnetization J _s is most attracted in the direction of the exciting magnetic field H (t), and when the exciting magnetic field H (t) is (2) and has a positive peak. Sometimes it is farthest from the direction of the exciting magnetic field H (t).

In this way, when the direction of the DC bias magnetic field H _dc is reversed, the angle θ ₀ is also reversed, and the magnetic flux interlinking with the pickup coil is also reversed. In this case, assuming an ideal state where there is no imbalance in the anisotropic K _u, offset components caused by the anisotropy K _u is manifested in the opposite polarity to the case excitation current of positive polarity, positive polarity Will be present in the same amount as. Therefore, when the polarity of the exciting current is switched, the output offset appearing at the sensor output can be set to 0 by finally averaging these offset components on the circuit. The "output offset" means the offset remaining in the output after summing (averaging) the offset when the bias is positive and the offset when the bias is negative. And. Further, the anisotropy K _u varies with temperature, even if offset due to this change, in an ideal state, the change is appearing always symmetrically, the average is not changed, at the output The output offset component remains 0. However, in practice, since the anisotropic K _u are present unbalanced, the output offset is normally may not become zero.

Regarding the sensitivity and the magnitude and polarity of the offset component, a pickup signal is induced in the pickup coil by the time change of the _{magnetization J s , and the pickup signal contains information on the external magnetic field H ex} and the offset component. Therefore, as described above, the external magnetic field depends on how the magnetization J _s moves according to the respective factors of the exciting magnetic field H (t) due to the exciting current, the anisotropic K _u , and the external magnetic field H _ex. magnitude and polarity of the offset component due to the magnitude and polarity and anisotropic K _u of sensitivity is determined for.

(Noise suppression method by temporarily disconnecting the conventional detection circuit)
Next, a bias switching type fluxgate magnetic field sensor using a conventional method for reducing noise will be described. FIG. 2 shows an outline of a fluxgate magnetic field sensor 200 that employs a method shown in Patent Document 2 that cuts off a transient signal by completely cutting off the detection circuit at the moment when the bias is switched to open the detection circuit. The circuit diagram is shown. FIG. 2 corresponds to the lower view of FIG. 4 of Patent Document 2, and the reference numerals thereof are changed.

The flux gate magnetic field sensor 200 is similar to the flux gate magnetic field sensor 100 of the basic bias switching method, and has a DC superimposed AC excitation unit 201, an exciting magnetic pole switching unit 202, a sensor unit 203, a buffer circuit (voltage follower) 204, and an amplifier 205. , Synchronous detector 206, low-pass filter 207, error amplifier (integrator) 208, feedback resistor (feedback resistor) 209, and output terminal 210. In addition, the fluxgate magnetic field sensor 200 has a configuration in which the DC superimposed AC excitation unit 201 switches between two phase shifters, and also has an additional configuration of a periodic cutting means 211 and a voltage adjusting unit 212. doing. In Patent Document 2, the configuration in which the DC superimposing AC excitation unit 201 switches between the two phase shifters is the "third switch 11c (the third switching means of claim 2)", and the periodic cutting means 211 is the "fourth". The switch 20 (fourth switching means of claim 4) and the voltage adjusting unit 212 are expressed as "second switch 23c (second switching means of claim 1)".

The configuration in which the DC superimposition AC excitation unit 201 switches between the two phase shifters (the third switch 11c of Patent Document 2) makes it possible to optimize the feedback conditions for the signals obtained in both bias excitation sections. It is a configuration, and it is said that compensation that follows a change in the induced voltage can be easily performed. Further, the voltage adjusting unit 212 (second switch 23c of Patent Document 2) is a positive electrode power supply that adjusts the current added to the output signal in the case of the positive electrode property, corresponding to the switching polarity of the exciting pole switching unit 202. In the case of negative electrode property, the connection with the negative electrode power supply that adjusts the current added to the output signal is switched, and each polarity can be performed as an independent separate process. It is said that the instability of the feedback loop caused by the transition between them can be eliminated.

The periodic cutting means 211 (fourth switch 20 of Patent Document 2) is configured to reduce noise, and cuts the circuit at a cycle earlier than the switching cycle of the exciting magnetic field switching unit 202. The periodic disconnection means 211 is for temporarily cutting off the signal of the circuit at the moment when the bias is switched, and by operating at a frequency twice the switching, the bias has the same polarity from the moment of the bias switching. The detection circuit is cut off until half of the half cycle of the switching cycle that holds the signal elapses, the signal is cut off, and the detection circuit is connected for the other half of the half cycle. It is a kind of sample circuit that returns the signal to the feedback circuit. The periodic cutting means 211 is said to have an effect that noise generated at the time of polarity switching can be removed by cutting the detection circuit for a predetermined period at the moment when the polarity switching occurs. However, in Patent Document 2, the fourth switching means corresponding to the periodic cutting means 211 is described in claim 4, which is characterized by the first and third switching means characterized by the second switching means. It is a dependent claim of claim 2. As described above, the periodic cutting means 211 (fourth switch 20 of Patent Document 2) has a configuration in which the DC superimposing AC excitation unit 201 switches between two phase shifters (third switch 11c of Patent Document 2) and voltage adjustment. Since it is premised on the part 212 (the second switch 23c of Patent Document 2), it is recognized that it does not exert a sufficient effect by itself. In fact, if the detection circuit is temporarily disconnected to open the circuit, it is considered that the feed pack loop may become unstable and the output may be disturbed, resulting in sensor malfunction depending on the conditions.

(Basic configuration of the noise reduction method of the present invention)
The present invention cuts the circuit and cuts the circuit downstream during the stabilization period, which is a predetermined period that starts at the same time as each switching of the polarity of the DC bias current and ends before the next switching of the switching. By replacing the output of the detector with a signal not derived from the pickup signal at the disconnection point, the noise is reduced and the sensor output is stabilized. As a method of replacing the output of the detector with a signal not derived from the pickup signal, the output of the detector is replaced with a signal not derived from the pickup signal immediately after the detector and immediately before the output circuit such as a low-pass filter or an integrator. By doing so, the output of the detector is directly stabilized, or the output of the detector is stabilized by replacing the pickup signal from immediately after the pickup coil to entering the detector with a signal not derived from the pickup signal. Either method can be considered. Replacing the output of the detector with a signal not derived from the pickup signal is performed by providing a switch controlled by the control signal at an appropriate replacement location. The switch is a semiconductor switch element that provides a single-pole, double-throw connection that connects the circuit downstream of the switch, which is the direction in which the detector output flows, to the circuit upstream of the switch, or derives from a pickup signal. Switch whether to connect to the signal source of the signal that does not. Therefore, when the switch connects its downstream circuit to the source of a signal that does not derive from the pickup signal, the circuit upstream of the switch will be disconnected there, and the circuit of the sensor based on the pickup signal will be at the switch. It will be disconnected. As the signal source of the signal not derived from the pickup signal, a ground or a voltage source having a predetermined voltage can be used. Replacing the output of the detector with a signal that does not originate from the pickup signal means connecting the signal source of the signal that does not originate from the pickup signal to the replacement location of the circuit and disconnecting the circuit immediately before the replacement location. Shall be done. That is, the signal source of the signal not derived from the pickup signal is connected immediately after the disconnection point of the circuit. As a result, a signal not derived from the pickup signal flows downstream of the replacement portion (cutting portion).

As a method of directly stabilizing the output of the detector by using a signal not derived from the pickup signal, first, it is conceivable to ground the output of the detector. The output of the detector is replaced with ground immediately after the detector and immediately before the output circuit including the low-pass filter in the subsequent stage (integrator if the low-pass filter is omitted in the feedback system (closed loop)). Is done. In this case, the output of the detector during the stabilization period, which starts at the same time as the switching and ends before the next switching of the switching, becomes zero (ground potential), the transient signal due to the bias switching is removed, and during that period. The output of the detector is zero, which does not affect the sensor output. Then, a stable output based on the pickup signal that does not include the transient signal other than the stabilization period is output from the detector.

Further, as a method of directly stabilizing the output of the detector by using a signal not derived from the pickup signal, it is conceivable to connect the output of the detector to a voltage source having a predetermined voltage. The output of the detector is connected to the voltage source of the predetermined voltage of the output circuit including the low-pass filter (integrator when the low-pass filter is omitted in the feedback system (closed loop)) in the subsequent stage immediately after the detector. It takes place just before. In this case, the output of the detector during the stabilization period, which starts at the same time as the switching and ends before the next switching of the switching, becomes a predetermined voltage, the transient signal due to the bias switching is removed, and the detector during that period is removed. The output of is a predetermined voltage, which affects only the DC level of the sensor output. This predetermined voltage is preferably a voltage that cancels the output offset that could not be removed by bias switching. Then, a stable output based on the pickup signal that does not include the transient signal other than the stabilization period is output from the detector.

Next, as a method of stabilizing the pickup signal until it enters the detector by using a signal not derived from it, the pickup signal is grounded at an arbitrary point between immediately after the pickup coil and immediately before the detector. Can be considered. In this case, the output of the detector during the stabilization period, which starts at the same time as the switching and ends before the next switching of the switching, becomes a nearly zero value based on the pick-up signal of zero level, and is a transient signal due to bias switching. Is removed, and the output of the detector during that period is a value of almost zero, which is an output that does not affect the sensor output. Then, a stable output based on the pickup signal that does not include the transient signal other than the stabilization period is output from the detector.

It should be noted that, at any point between immediately after the pickup coil and immediately before the detector, it is possible to stabilize the pickup signal by connecting it to a voltage source of a predetermined voltage (rather than grounding). Even if the pickup signal is replaced with a predetermined voltage and a DC component is applied, the detector output removes the influence of such a DC component from the sensor output, so that the sensor output is the same as when the pickup signal is grounded. It is the same. That is, in this case, the output offset cannot be canceled.

In this way, in order to remove the influence of the transient signal due to bias switching from the output of the detector and stabilize it, during the stabilization period, immediately after the pickup coil, such as immediately after the detector, and immediately before the output circuit. Cut the circuit at any of the above points and ground immediately after the cut point, or immediately after the detector, such as immediately after the detector during the stabilization period, from immediately after the detector to just before the output circuit. There are at least two methods of cutting the circuit at the above and connecting immediately after the cut point to a voltage source of a predetermined voltage. Hereinafter, as a first embodiment, a mode in which any one of the circuits is grounded during the stabilization period will be described, and as a second embodiment, the output of the detector during the stabilization period will be described. Will be described in the form of connecting the device to a voltage source having a predetermined voltage.

(Structure of Fluxgate Magnetic Field Sensor of First Embodiment)
The structure of the fluxgate magnetic field sensor 300 according to the first embodiment will be described. The flux gate magnetic field sensor 300 is an output circuit immediately after the pickup coil during a stabilization period, which is a predetermined period that starts at the same time as each switching of the polarity of the DC bias current and ends before the next switching of the switching. The output of the detector is replaced with a signal not derived from the pickup signal by cutting the circuit at any part of the circuit up to just before the above and grounding immediately after the cut part.

FIG. 3 is a schematic circuit diagram of the fluxgate magnetic field sensor 300. In this circuit, synchronous detection is used as a typical detection method. In addition, a closed loop configuration is adopted in which the sensor is operated in a region where the characteristics of the magnetic material are good by passing a current that cancels the detected magnetic field through the pickup coil by feedback. Alternatively, it is also possible to adopt a closed loop configuration in which a dedicated coil is separately provided without using the pickup coil in order to cancel the detected magnetic field, or an open loop configuration in which feedback is not performed.

The fluxgate magnetic field sensor 300 includes the same configuration as the conventional basic fluxgate magnetic field sensor 100, and includes a DC superimposition AC excitation unit 301, an exciting magnetic pole switching unit 302, a sensor unit 303, a voltage follower 304, an amplifier 305, and the like. It includes a synchronous detector 306, a low-pass filter 307, an integrator 308, a feedback resistor 309, and an output terminal 310. The fluxgate magnetic field sensor 300 further includes a grounded periodic stabilization circuit 311.

The DC superimposed AC excitation unit 301 is a circuit that provides an exciting current in which a DC bias current is superimposed on an AC current. The DC superimposition AC excitation unit 301 is typically a known power supply circuit. In addition to the illustrated method, an AC source and a DC source may be connected in series, or an AC source and a DC source may be added by using an adder such as an operational amplifier. The DC superimposition AC excitation unit 301 modulates an external magnetic field by passing an AC current through the sensor unit 303, and superimposes a DC bias current to form a bias switching type excitation circuit together with the exciting magnetic field switching unit 302 in the subsequent stage. do. In this case, the polarity of the alternating current on which the direct current bias current is superimposed is switched according to the switching of the polarity of the direct current bias current.

In bias switching, it is necessary to periodically switch the polarity of the DC bias current. Here, the AC current on which the DC bias current is superimposed can be configured to switch the polarity according to the change of the polarity of the DC bias current, or to not switch the polarity. In the above example, the polarities of both AC current and DC bias current are switched. In this case, if addition processing (averaging processing) is performed after synchronous detection, signals that do not correspond to the detected magnetic field are removed. The sensor output can be obtained.

Alternatively, in order not to switch the polarity of the AC current together with the polarity of the DC bias current, the DC superimposed AC excitation unit 301 applies an exciting current obtained by superimposing the DC bias current whose polarity is periodically switched on the AC current. It can also be configured to provide. In this case, the exciting magnetic field switching unit 302 described later becomes unnecessary. Further, in order to obtain the magnetic field detection signal, it is necessary to take the difference in the detection number at each polarity of the DC bias current after the synchronous detection. For that purpose, for example, it is preferable to perform the addition process (averaging process) after inverting the polarity of the detection signal at the polarity of one of the DC bias currents.

The magnetic field switching unit 302 is a circuit that periodically switches the polarity of the exciting current flowing through the sensor unit 303. As a result, the polarity of the DC bias current component is periodically switched, and the exciting current by the bias switching method is supplied to the magnetic core of the sensor unit 303. The exciting magnetic field switching unit 302 has _{a clock having a frequency of f bs} Hz that determines the period of polarity switching, and the switching signal that switches the polarity is driven by the switching signal generated by the clock, and the exciting current is periodic. The polarity is switched to. A clock with a frequency of f _bs Hz is typically used by dividing the frequency of the AC component of the exciting current by a fraction of an integer. As shown in FIG. 3, when the exciting magnetic pole switching unit 302 switches the output from the DC superimposed AC exciting unit 301, the phase of the AC current is also inverted in synchronization with the switching of the polarity of the DC bias current, and the DC For each polarity of the bias current, the offset component due to magnetic anisotropy appears in the opposite polarity, and the sensitivity component of the magnetic field appears in the same polarity. The frequency of the clock is smaller than the frequency of the alternating current generated by the direct current superimposing alternating current excitation unit 301. Further, the clock typically uses a timing signal whose frequency is divided by a fraction of an integer with respect to the AC exciting current generated by the DC superimposing AC exciting unit 301, and the phase is arbitrarily shifted as needed. NS. For example, when the frequency of the AC component of the exciting current from the DC superimposed AC exciting unit 301 is 80 kHz, the switching frequency f _bs of the exciting magnetic pole switching unit 302 is set to about 5 kHz, and the polarity of the DC bias current is about 16 sine waves. Can be configured to repeat the switching of.

In the bias switching method, it is important to periodically switch the polarity of the DC bias current component, and at that time, both a configuration in which the phases of the AC components are simultaneously inverted and a configuration in which the phases are not inverted are possible. When the exciting magnetic pole switching unit 302 is used, the polarity of the entire exciting current, which is an alternating current on which the direct current bias current is superimposed, is switched, so that the phase of the alternating current component can be inverted in synchronization with the switching of the polarity of the direct current bias current component. It will be. In this case, in the pickup signal, the offset component due to magnetic anisotropy appears with the opposite polarity and the sensitivity component of the magnetic field appears with the same polarity with respect to each polarity of the DC bias current.

Alternatively, if only the polarity of the DC bias current component is periodically switched and the phases of the AC components are not inverted at the same time, the magnetic field type switching unit 302 is unnecessary. In this case, the DC superimposed AC excitation unit 301 is configured to provide an exciting current in which a DC bias current whose polarity is periodically switched is superimposed on the AC current. If the phase of the AC current component is not inverted regardless of the polarity switching of the DC bias current component, the offset component due to magnetic anisotropy has the same polarity with respect to each polarity of the DC bias current in the pickup signal. Appears in, and the sensitivity component of the magnetic field appears in the opposite polarity.

The sensor unit 303 is a sensor element that detects an external magnetic field to be measured, and has a magnetic core made of an elongated magnetic material to which the detected magnetic field is applied, and a pickup coil wound around the elongated magnetic material. The sensor unit 303 typically has a structure in which a pickup coil is wound around an amorphous magnetic wire bent into a U shape. The amorphous magnetic wire has a non-magnetic dwarf composition in which magnetic anisotropy is unlikely to occur. The shape of the amorphous magnetic wire is not limited to the U-shape, and may be a type II in which the tip is short-circuited with a conducting wire or the like. The magnitude of the external magnetic field is determined by the detection voltage (pickup signal) detected by the pickup coil of the changing magnetic flux generated by the exciting current flowing through the amorphous magnetic wire, including the effect of magnetization by the external magnetic field generated in the amorphous magnetic wire. Measure the magnetic flux. Amorphous magnetic wire operates even in the shape of a single wire I type, but its sensitivity is lower than that of U type or II type, which has two magnetic materials and can obtain high sensitivity. Further, in order to increase the sensitivity, a plurality of shapes such as U-type, II-type, and I-type can be used in combination. The sensor unit 303 has a configuration of an orthogonal flux gate in which an exciting current is directly passed through an amorphous magnetic wire. Low noise is achieved by reducing the accompanying magnetic noise of the magnetic material.

The voltage follower 304 is a component for accurately transmitting the pickup signal from the sensor unit 303 to the subsequent stage by converting the impedance. Since a sufficient current cannot be taken out from the voltage induced in the pickup coil, a voltage drop may occur depending on the impedance of the subsequent stage and the sensitivity may decrease. The voltage follower 304 prevents such a decrease in sensitivity. The voltage follower 304 is typically an amplifier circuit composed of an operational amplifier and having an amplification factor of 1. The amplifier 305 is a component that amplifies the output from the voltage follower 304 to an appropriate level.

In FIG. 3, the output from the voltage follower 304 passes through a capacitor to take out only the AC component, and is sent to the amplifier 305 in the subsequent stage. As a result, unnecessary DC components (such as the voltage due to the feedback signal) included in the pickup voltage are removed. The voltage follower 304 and the capacitor are additional configurations for accurately transmitting the pickup signal at an appropriate level, and are not necessarily essential in the present invention.

The synchronous detector 306 is a component that performs synchronous detection on the pickup signal from the sensor unit 303 output from the amplifier 305. The synchronous detector 306 synchronously detects the pickup signal, which is the induced voltage induced in the pickup coil of the sensor unit 303, by referring to the phase of the reference signal whose frequency is synchronized with the AC current in the excitation current, thereby outputting the sensor. Generates a detection signal that will determine. The detection signal output from the synchronous detector 306 is transmitted to the grounded periodic stabilization circuit 311 in the subsequent stage. The synchronous detector 306 detects the pickup signal from the sensor unit 303 according to the phase relationship with the reference signal having the same frequency as the AC component of the exciting current of the DC superimposed AC excitation unit 301. .. A signal having the same frequency as the AC current of the DC superimposed AC excitation unit 301 is input to the synchronous detector 306 as a reference signal, whereby the phase relationship with the pickup signal to be detected is determined. Typically, when the reference signal is a square wave gate control signal synchronized with the AC component of the exciting current in the DC superimposed AC excitation unit 301, this drives an internal switch and the gate control signal becomes negative (LOW). ), The polarity of the pickup signal is inverted and turned back. This makes it possible to obtain a meaningful output consisting of components synchronized with the reference signal. The synchronous detector 306 is typically composed of elements such as an analog switch. In order to obtain a detection signal that determines the sensor output from the pickup signal, other means such as peak value detection can be used without using synchronous detection.

The low-pass filter 307 is a component of a circuit that averages the stabilized detection signal, which is an output from the grounded periodic stabilization circuit 311 arranged immediately after the synchronous detector 306. As a result, the magnetic field detection signal representing the magnitude of the magnetic field detected from the detection signal is extracted. In this example, the phase of the AC current is also inverted in synchronization with the switching of the polarity of the DC bias current, and in the pickup signal, the offset component due to magnetic anisotropy for each polarity of the DC bias current is Since it appears in the opposite polarity and the sensitivity components of the magnetic field appear in the same polarity, by adding the detection signals (realized by averaging), a magnetic field detection signal is obtained by removing signals that do not correspond to the magnetic field detected from the detection signal. Can be done. Specifically, the low-pass filter 307 smoothes the waveform of the signal representing the magnitude of the magnetic field that is the pulsating current from the detection signal to make it DC, and removes the offset due to the DC bias current whose polarity is switched by addition. .. When operating the system in a closed loop configuration, the low-pass filter 307 can be omitted.

Alternatively, when only the polarity of the DC bias current component is periodically switched and the phases of the AC components are not inverted at the same time, that is, the DC superimposing AC exciting unit 301 periodically switches the polarity without using the exciting magnetic pole switching unit 302. In the case of providing an exciting current in which the DC bias current is superposed on the AC current, the phase of the AC current component is not inverted regardless of the polarity switching of the DC bias current component. In this case, in the pickup signal, the offset component due to magnetic anisotropy appears with the same polarity and the sensitivity component of the magnetic field appears with the opposite polarity with respect to each polarity of the DC bias current. Therefore, in order to obtain the magnetic field detection signal, it is necessary to obtain the difference between the detection signals for each polarity of the DC bias current. Specifically, after inverting (-1 times) the detection signal at one polarity of the DC bias current, it is preferable to execute an addition process (averaging process) together with the detection signal at the other polarity of the DC bias current. Thereby, the magnetic field detection signal obtained by removing the signal corresponding to the magnetic field detected from the detection signal can be obtained.

The integrator 308 is a component of a circuit that integrates the step waveform output from the low-pass filter 307 when the system is operated in a closed loop configuration. If the low-pass filter 307 is omitted, the integrator 308 integrates the output of the synchronous detector 306. The output of the integrator 308 applies negative feedback by a magnetic field to the magnetic core of the sensor unit 303 through the pickup coil of the sensor unit 303 via the feedback resistor 309 described later. As a result, the magnetic field detection signal that is the input of the integrator 308 represents the deviation of this feedback, and the integrator 308 controls the magnetic field detection signal to approach zero.

The feedback resistor 309 is a resistor through which the feedback signal passes when the feedback signal output from the integrator 308 is negatively fed back and input to the sensor unit 303. For negative feedback, the footback signal integrated with the magnetic field detection signal is input to the pickup coil with a polarity that cancels the magnetic field detection signal. Due to the negative feedback, a magnetic field that cancels the external magnetic field to be measured is generated in the pickup coil of the sensor unit 303. The current flowing by the feedback signal corresponds to the external magnetic field to be measured, and the current value detected as a voltage by the feedback resistor 309 is the sensor output indicating the magnitude of the external magnetic field to be measured.

Since the above example has a closed loop configuration, a feedback signal that generates a magnetic field that cancels the detected external magnetic field that has flowed into the magnetic core is input to the pickup coil, and the magnitude of the feedback signal is used as the sensor output. be. Therefore, the low-pass filter 307, the integrator 308, and the feedback resistor 309 configure an output circuit that generates a sensor output based on the magnetic field detection signal obtained by removing the signal that does not correspond to the magnetic field detected from the detection signal.

In the case of the open loop configuration, the integrator 308 and the feedback resistor 309 are unnecessary because the output of the low-pass filter 307 represents the external magnetic field to be measured. Then, the sensor output can be obtained from the output of the low-pass filter 307.

The output terminal 310 is a node that outputs a sensor output obtained by converting the current flowing through the feedback resistor 309 into a voltage representing the magnitude of an external magnetic field. As a result, the magnitude of the current that generates the magnetic field that cancels the external magnetic field of the measurement target is output as a measure quantity that represents the external magnetic field of the measurement target.

The grounded periodic stabilization circuit 311 is synchronized for a stabilization period, which is a predetermined period that starts at the same time as the switching and ends before the next switching of the switching (before the half-cycle period of the switching cycle). The configuration is such that the detection signal of the output of the detector 306 is stabilized by replacing it with the ground potential. The stabilized detection signal is output to the low-pass filter 307 (integrator 308 in the case of a configuration in which the low-pass filter 307 is not used) in the subsequent stage. In the example of FIG. 3, the grounded periodic stabilization circuit 311 is connected to the subsequent stage of the synchronous detector 306. The grounded periodic stabilization circuit 311 typically includes a semiconductor switch element that operates by a control signal, which is controlled by the control signal and disconnects and disconnects the circuit immediately after the synchronous detector 306. By grounding immediately after the location, the output of the synchronous detector 306 is replaced with a ground potential signal.

The stabilization period begins at the same time that the polarity of the DC bias current is switched by switching, and before the switching of the polarity of the DC bias current by the next switching to the opposite polarity (the DC bias current retains either polarity). It can be a predetermined period that ends (before the elapse of the half-cycle period of the switching cycle). By setting the stabilization period to include a period in which a transient signal is mixed in the pickup signal due to bias switching, the output of the detector during that period is grounded, and the influence of the transient signal on the output can be eliminated. If the stabilization period is too short, the transient signal may not be completely removed, and if the stabilization period is too long, the amount of the original pickup signal passing through will be small, and the S / N ratio may deteriorate. Therefore, it is preferable to set an appropriate period so that such a situation does not occur. Thus, the stabilization period can be an appropriate period of less than half the switching period from one bias switching to the next, where the DC bias currents retain the same polarity, for example. It can be started from a certain switching and ended in a half period of the period from the switching to the next switching (half cycle of the switching cycle). In this case, the stabilization period will be repeated in half the switching cycle. Specifically, the stabilization period is started at the same time that the DC bias current is switched to either polarity, and the next switching after that switching (half cycle of the switching cycle in which the DC bias current holds the same polarity). The stabilization period will end at the same time that half of the period up to is passed. The control signal for controlling the stabilization period at such timing is synchronized with the switching signal for bias switching, and is the midpoint between the timing of each polarity switching of the switching signal and the timing of each polarity switching. It can be a square wave whose frequency is twice that of the switching signal, such that the polarity is switched in. In this case, the switching signal can be generated by dividing the output of the oscillation circuit of the control signal by two, and the overall circuit can be simplified.

The grounding point is preferably immediately after the synchronous detector 306 (immediately downstream) as in the grounded periodic stabilization circuit 311 shown in FIG. 3, but immediately before (immediately upstream) the synchronous detector 306. That is, even if the pickup signal is grounded immediately after the amplifier 305, almost the same operation is performed. Furthermore, the pickup signal or the detection signal is grounded by grounding any part from immediately after the sensor unit 303 (pickup coil) to immediately before the low-pass filter 307 (integrator 308 when the low-pass filter 307 is omitted). It can also be replaced with. That is, the grounded periodic stabilization circuit 311 can be arranged at an arbitrary position from immediately after the sensor unit 303 to immediately before the low-pass filter 307 (integrator 308 when the low-pass filter 307 is omitted).

(Operation of Fluxgate Magnetic Field Sensor of First Embodiment)
From now on, the operation of the fluxgate magnetic field sensor 300 to ground the output of the synchronous detector 306 during the stabilization period will be described while showing the waveform of the detection signal with the passage of time. FIG. 8 is a waveform diagram of the detection signal. In FIG. 8, in the flux gate magnetic field sensor 300 according to the first embodiment, the grounded periodic stabilization circuit 311 is operated at a period (twice the frequency) of the period of the switching signal to perform synchronous detection. The waveform of the output of the synchronous detector 306 when the output (detection signal) of the device 306 is grounded and the transient signal is removed is shown in black. At this time, the control signal for controlling the grounded periodic stabilization circuit 311 has a frequency of the switching signal in a phase relationship such that the zero point of the control signal comes to the zero point of the switching signal of the exciting magnetic pole switching unit 302. The signal is doubled. In FIG. 8, for comparison, the waveform of the output of the synchronous detector 306 when the transient signal is not removed without grounding the detection signal is shown in gray. Here, since the polarity of the DC bias current is switched by bias switching every 8 cycles of the alternating current in the exciting current, the output waveforms are on the positive side and the negative side every 8 cycles of the pickup signal corresponding to the alternating current. It is switched with. Then, the output (detection signal) of the synchronous detector 306 is switched between grounded and ungrounded by the grounded periodic stabilization circuit 311 every four cycles of the pickup signal corresponding to the alternating current. When grounded, the detection signal is set to almost 0, and when not grounded, the detection signal passes directly to the output. In the waveform of the detection signal when it is left ungrounded, which is shown in gray in the figure, a transient signal waveform with a large amplitude appears at the moment of switching, and then it is transient to the detection signal for about 3 cycles of the pickup signal corresponding to the alternating current. The influence of the signal remains, and the level is slightly higher. On the other hand, in the waveform of the detection signal when the detection signal is periodically grounded shown in black, the detection signal should be grounded for half the period from the moment of switching to the next switching. It can be confirmed that the transient signal and the signal corresponding to the pickup signal corresponding to the AC current of about three cycles affected by the transient signal are removed from the output. Then, the remaining half period from that time to the next switching corresponds to the pickup signal corresponding to the four-cycle alternating current that is not affected by the transient signal by making the detection signal ungrounded. The signal is being output.

Although the influence of the transient signal is eliminated by such a configuration, since the waveform level of half of the detection signal is zero, the final sensor output is when the detection signal is not grounded periodically. Compared with, the voltage is halved. In this way, lowering the level of the original signal has a slight adverse effect on the noise characteristics, but the noise reduction effect of removing the transient signal is much higher, so the noise is greatly reduced as a whole. .. Further, since the noise characteristics can be kept good even if the exciting current is reduced, the power consumption can be reduced.

The effect of noise reduction by removing the transient signal will be described. FIG. 9 is a graph of actual measurement values of the noise density of the detection signal when the transient signal is not removed in the three types of exciting currents. From the graph, it is observed that the noise increases significantly when the exciting current is reduced. This is because the amplitude of the alternating current in the exciting current is reduced, which reduces the sensor sensitivity, and the noise component caused by the difference in the saturation state of the magnetization when the magnetization is reversed due to the reversal of the exciting current. It is due to the difference. FIG. 10 is a graph of actual measurement values of the noise density of the detection signal when the transient signal is removed in the three types of exciting currents. In this case, since the transient signal that becomes a noise source is removed, it is possible to always maintain good and equivalent noise characteristics without depending on the exciting current. Further, since the exciting current can be reduced without changing the noise performance, the power consumption of the sensor can be reduced to about one-fifth. This is a huge advantage for applications with limited power resources.

Next, by using a configuration for an evaluation experiment in which the magnitude of the transient signal (that is, the magnitude of noise) is changed by changing the timing (phase) of the bias switching of the exciting current, the fluxgate magnetic field sensor 300 It was confirmed that the performance of removing noise and the robustness of the output offset were improved. FIG. 11 is a waveform diagram of the exciting current when the timing (phase) of bias switching is changed. 11 (A), (B), (C), and (D) show output current waveforms when the bias switching phases are 0 degrees, 96 degrees, 160 degrees, and 256 degrees, respectively. There is. When the timing of bias switching is changed in this way, the timing at which the alternating current of the exciting current is switched changes accordingly. As a result, the sensor noise before removing the transient signal changes.

FIG. 12A shows a graph of the measured value of the sensor noise amount when the timing (phase) of bias switching of the sensor before removing the transient signal is changed. The horizontal axis of the graph is the bias switching phase, which is the switching phase from the state (A) in FIG. 11 to the states (B), (C), (D), and again (A) in appropriate angular steps. Is changing. Then, the sensor noise at each bias switching phase is measured, and the noise density at 1 Hz is plotted on the vertical axis in the graph. From the graph, noise in the range of 10 to 75 pT / Hz ^1/2 is generated, but when the phase of bias switching is 256 degrees, the noise is ^{about 75 pT / Hz 1/2,} which is very large. Understood.

FIG. 12B shows a graph of the measured value of the output offset of the sensor before removing the transient signal. In FIG. 12B, the phase of the bias switching is changed stepwise with time in a step smaller than the phase step shown in FIG. 12A, and the output is recorded for about 10 seconds in each state. be. The horizontal axis is time, but since the phase of bias switching is changed in a constant step with time, the phase of bias switching also corresponds to the horizontal axis. Since this measurement is performed in an environment where the external magnetic field can be ignored, this sensor output corresponds to the output offset. From the graph, it can be confirmed that there is an output offset in the range of about −8.5 to 3 nT. As described above, FIG. 12B shows that the output offset fluctuates by 10 nT or more by changing the phase of the bias switching of the exciting current. The width of the line in the graph of the output offset where the phase of the via switching is around 250 degrees is wide, but this is because the noise is large, so there are many small fluctuations in the output offset, and the width looks thick. .. If the phase of bias switching shifts due to the temperature and drift of the circuit section of the sensor over time, an increase in noise and a large change in output offset will occur. It can also be a constraint on the adjustment of circuit operating conditions.

Although the output of the sensor before removing the transient signal shown in FIG. 12 contains a large amount of noise, the sensor is subjected to periodic grounding of the detection signal by the grounding type periodic stabilization circuit 311. By removing the transient signal, performance such as noise and output offset can be stabilized. FIG. 13 is a graph of the measured values of the sensor noise amount when the phase of bias switching is changed, when the transient signal is not removed and when it is removed. It can be confirmed that the influence of the transient signal is eliminated by the periodic grounding of the detection signal by the grounded periodic stabilization circuit 311, the noise is extremely small regardless of the phase of switching the bias switching, and the fluctuation is also extremely small. ..

FIG. 14 shows a graph of the measured value of the output offset when the transient signal is removed. The horizontal axis is time, but since the phase of bias switching is changed in a constant step with time, the phase of bias switching also corresponds to the horizontal axis. However, the scale of the phase adjustment step does not completely match the horizontal axis of FIG. From the graph, it can be understood that the fluctuation of the output offset is within 2 nT and is extremely stable even if the phase of the bias switching is changed. It can also be understood that there are no places where the line width of the graph is wide and the noise is extremely small. In this way, the robustness of the sensor circuit against temperature and drift over time is greatly improved. In addition, the degree of freedom in making adjustments to find the optimum operating conditions for the sensor can be increased.

In this way, the transient noise portion of the detection signal is derived from the pickup signal during the stabilization period, which is a predetermined period that starts at the same time as each switching of the polarity of the DC bias current and ends before the next switching of the switching. With the configuration of replacing with a non-existing signal and stabilizing it, the transient noise associated with bias switching can be replaced with a signal of another stable signal source not derived from the pickup signal. In this case, the feedback loop is not left disconnected when the transient noise is removed, and the transient noise portion is always replaced by a constant voltage. In addition, there is no jitter or the like due to the hold of the signal that exists when the sample and hold switch is used in the sensor output. As described above, in the present invention, it is possible to remove the transient noise caused by the bias switching very robustly.

Further, during the stabilization period, the pickup signal that replaces the transient noise portion of the detection signal can be the ground potential. In this case, the stabilized detection signal is averaged (addition / subtraction processing) according to the process of the bias switching method and becomes the sensor output. Here, since the ground potential is the reference potential of the circuit, if the output offset is not taken into consideration, the result of the averaging is 0, and the transient noise does not affect the sensor output. Therefore, the averaging result of only the original signal component appears in the sensor output, and it becomes possible to provide a sensor by the bias switching method with further reduced noise.

(Structure of Fluxgate Magnetic Field Sensor of Second Embodiment)
Next, the structure of the fluxgate magnetic field sensor 400 according to the second embodiment will be described. FIG. 4 is a schematic circuit diagram of the fluxgate magnetic field sensor 400. The fluxgate magnetic field sensor 400 includes the same configuration as the fluxgate magnetic field sensor 300 according to the first embodiment, and includes a DC superimposition AC excitation unit 401, an exciting magnetic pole switching unit 402, a sensor unit 403, a voltage follower 404, and an amplifier. It includes a 405, a synchronous detector 406, a low-pass filter 407, an integrator 408, a feedback resistor 409, and an output terminal 410. The fluxgate magnetic field sensor 400 has a constant voltage type periodic stabilization circuit 411 instead of the ground type periodic stabilization circuit 311 in the fluxgate magnetic field sensor 300.

The constant voltage type periodic stabilization circuit 411 starts at the same time as the bias switching and ends until the next bias switching of the bias switching (up to the half cycle period of the switching cycle) during the stabilization period, which is a predetermined period. , The configuration is such that the detection signal of the output of the synchronous detector 406 is stabilized by replacing it with a voltage source of a predetermined voltage. The constant voltage periodic stabilization circuit 411 differs from the grounded periodic stabilization circuit 311 in that its output is periodically connected to a constant voltage source (rather than grounded), except that it is grounded. It has the same configuration as the type periodic stabilization circuit 311. The constant voltage type periodic stabilization circuit 411 is connected to the subsequent stage of the synchronous detector 406. The constant voltage type periodic stabilization circuit 411 typically includes a semiconductor switch element that operates by a control signal, which is controlled by the control signal and disconnects the circuit immediately after the synchronous detector 406. By connecting a constant voltage source immediately after the disconnection point, the output of the synchronous detector 406 is replaced with a signal of a predetermined voltage. As the constant voltage source, a configuration in which the power supply is divided by a resistor, a constant voltage circuit using a dedicated IC, or the like can be used.

The stabilization period of the constant voltage type periodic stabilization circuit 411 can also be determined in the same manner as that of the grounded type periodic stabilization circuit 311. That is, the stabilization period starts at the same time as the switching of the polarity of the DC bias current by switching, and before the switching of the polarity of the DC bias current by the next switching to the opposite polarity (the DC bias current retains either polarity). It can be a predetermined period that ends (before the elapse of the half-cycle period of the switching cycle). By including the period in which the transient signal is mixed in the pickup signal due to bias switching, the output of the detector during that period becomes a predetermined voltage from the constant voltage source, and the influence of the transient signal on the output is eliminated. can do. It is preferable that the connection point to the constant voltage source is immediately after the synchronous detector 406 as in the constant voltage type periodic stabilization circuit 411 shown in FIG. 4, but the low-pass filter is immediately after the synchronous detector 406. It can be any place between immediately before 407 (integrator 408 when the low-pass filter 407 is omitted). The constant voltage type periodic stabilization circuit 411 is arranged at an arbitrary position between the sensor unit and the detector like the ground type periodic stabilization circuit 311, and the pickup signal is periodically used as a constant voltage source there. It is also possible to connect. In this case, the noise characteristics are improved by removing the transient signal, but the output offset, which will be described later, cannot be removed.

(Operation of Fluxgate Magnetic Field Sensor of Second Embodiment)
From now on, the operation of the fluxgate magnetic field sensor 400 connecting the output of the synchronous detector 306 to the constant voltage source during the stabilization period will be described while showing the waveform of the detection signal over time. FIG. 15 is a waveform when the detection signal is periodically connected to a constant voltage source. In FIG. 15, in the flux gate magnetic field sensor 400 according to the second embodiment, the constant voltage type periodic stabilization circuit 411 is operated at a cycle of half the cycle of the switching signal, and the output of the synchronous detector 406 ( The waveform of the output of the synchronous detector 406 when the transient signal is removed by replacing the detector signal) with a constant voltage source is shown in black. In FIG. 15, for comparison, the waveform of the output of the synchronous detector 406 when the detection signal is not periodically replaced with the signal of the constant voltage source and the transient signal is not removed is shown in gray. In FIG. 15, first, the waveform when the voltage is zero (that is, the same as when grounded) is shown as the constant voltage source to be connected, and further, when the voltage of the constant voltage source is set to the adjustment voltage 1501. The change in waveform is indicated by the dashed line indicated by the arrow. When the detection signal is periodically replaced with a signal from a constant voltage source at the adjustment voltage 1501, the level of the detection signal is clamped at the adjustment voltage 1501 during the stabilization period. As a result, first, the transient signal is removed from the sensor output as in the first embodiment. Here, preferably, the adjustment voltage 1501 is set so that the output offset is minimized. For that purpose, the adjustment voltage 1501 may be adjusted while measuring the sensor output so that the output offset becomes almost zero. As a result, the output offset can be substantially removed from the sensor output.

The adjustment voltage 1501 can be set by the following procedure. FIG. 16 is a graph of the measured value of the output offset when the voltage of the constant voltage source is changed. FIG. 16 shows the sensor output when the voltage of the constant voltage source connected to the switch of the constant voltage type periodic stabilization circuit 411 is changed in the range of 0 to 2 mV in 0.5 mV steps. The horizontal axis is the time from the start of measurement. Since the voltage of the voltage source is 0 mV in 0 to 20 seconds of the graph, this is equivalent to the case where the detection output is grounded, and the output offset (about 8 nT) at this time is that of the magnetic material or circuit in this sensor. This is the output offset that is inevitably caused by the asymmetry. In the range of 0 to 250 seconds in the graph, the voltage of the constant voltage source is increased from 0 to 2 mV, then decreased from 2 mV to -2 mV, and then returned to 0 mV. In this way, the output offset of the sensor can be adjusted in proportion to the voltage of the constant voltage source that gives the adjustment voltage 1501. After 320 seconds in the graph, the output offset of the sensor is set to almost 0 by finely adjusting the voltage of the constant voltage source. In this way, by appropriately adjusting the voltage of the constant voltage source connected to the detection signal periodically, it is possible not only to remove the transient signal but also to reduce the output offset to almost zero.

As described above, a pickup signal that replaces the detection signal during a stabilization period, which is a predetermined period that begins at the same time as each switching of the polarity of the DC bias current and ends prior to the next switching of that switching. The underived signal can be a predetermined voltage. In this case, since the predetermined voltage that replaces the transient noise portion becomes the same signal regardless of the polarity of the DC bias current, it does not cancel out by averaging (addition / subtraction processing) and appears as it is in the sensor output. become. When the signal in which the transient noise is replaced is the ground potential, the sensor output obtained by averaging the portion of the signal replaced by the ground potential is 0 when the offset is not taken into consideration. Then, by substituting the transient noise with a predetermined adjustment voltage, it is possible to superimpose the adjustment predetermined voltage on the output in addition to the output of the original signal component. Here, the sensor output obtained by averaging the portion of the signal replaced by the ground potential is not 0 when the offset is taken into consideration. Then, when the offset is taken into consideration, by adjusting the predetermined adjustment voltage, it is possible to superimpose a voltage for correcting the output offset caused by the imbalance of each induced signal obtained by the exciting currents of the positive and negative polarities. .. By adjusting the predetermined adjustment voltage to a voltage that cancels the output offset, the output offset can be made almost zero.

Since the present invention improves the noise characteristics and output offset stability of the fluxgate magnetic field sensor and contributes to the reduction of power consumption, high reliability and resource reduction are required for aerospace applications and the like. It can be suitably applied to the magnetic field sensor used.

100: Flux gate magnetic field sensor 101: DC superimposed AC excitation unit 102: Exciting magnetic pole switching unit 103: Sensor unit 104: Voltage follower 105: Amplifier 106: Synchronous detector 107: Low pass filter 108: Integrator 109: Feedback resistance 110: Output Terminal 200: Flux gate magnetic field sensor 201: DC superimposed AC exciting unit 202: Exciting magnetic pole switching unit 203: Sensor unit 205: Amplifier 206: Synchronous detector 207: Low pass filter 210: Output terminal 211: Periodic cutting means 212: Voltage adjustment Unit 300: Flux gate magnetic field sensor 301: DC superimposed AC excitation unit 302: Exciting magnetic pole switching unit 303: Sensor unit 304: Voltage follower 305: Amplifier 306: Synchronous detector 307: Low pass filter 308: Integrator 309: Feedback resistance 310: Output terminal 311: Ground type periodic stabilization circuit 400: Flux gate magnetic field sensor 401: DC superimposed AC excitation unit 402: Exciting magnetic pole switching unit 403: Sensor unit 404: Voltage follower 405: Amplifier 406: Synchronous detector 407: Low pass filter 408: Integrator 409: Feedback resistance 410: Output terminal 411: Constant voltage type periodic stabilization circuit 1501: Adjusting voltage H: Exciting magnetic field H _dc : DC bias magnetic field H _ex : External magnetic field J _s : Magnetizing _Ku : Heterogeneous Gender f: Frequency f _bs : Switching frequency θ ₀ : Angle

Claims (9)

A magnetic core made of an elongated magnetic material to which a detected magnetic field is applied,
The pickup coil wound around the magnetic core and
A DC superimposed AC excitation unit that supplies an exciting current obtained by superimposing a DC bias current whose polarity is periodically switched between positive and negative electrodes on an AC current to the magnetic core.
An amplifier that amplifies the pickup signal from the pickup coil and
A detector that outputs a detection signal by detecting the amplified pickup signal, and
During the stabilization period, which is a predetermined period that starts at the same time as the switching of each of the polarities of the DC bias current and ends before the next switching of the switching, the detection signal becomes a signal not derived from the pickup signal. A periodic stabilization circuit that stabilizes by replacing,
A fluxgate magnetic field sensor comprising an output circuit that generates a sensor output based on a magnetic field detection signal that removes a signal that does not correspond to the detected magnetic field from the stabilized detection signal. The periodic stabilization circuit cuts the circuit at any point between immediately after the detector and immediately before the output circuit during the stabilization period, and grounds immediately after the cut point. , The fluxgate magnetic field sensor according to claim 1. The periodic stabilization circuit cuts the circuit at any point between immediately after the pickup coil and immediately before the detector during the stabilization period, and grounds immediately after the cut point. , The fluxgate magnetic field sensor according to claim 1. The periodic stabilization circuit cuts the circuit at any point between immediately after the detector and immediately before the output circuit during the stabilization period, and immediately after the cut point is a voltage of a predetermined voltage. The flux gate magnetic field sensor according to claim 1, which is connected to a source. The fluxgate magnetic field sensor according to claim 4, wherein the voltage of the predetermined voltage source is determined so that the output offset remaining in the sensor output is minimized. The fluxgate magnetic field sensor according to any one of claims 1 to 5, wherein the stabilization period starts from the switching and ends in a half period of the period until the next switching of the switching. .. The detector is any one of claims 1 to 6, wherein the amplified pickup signal is synchronously detected with reference to a reference signal whose frequency is synchronized with the AC current from the DC superimposed AC excitation unit. Fluxgate magnetic field sensor according to. The fluxgate magnetic field sensor according to any one of claims 1 to 7, wherein the alternating current has a polarity that is periodically reversed at the same timing as the switching of the direct current bias current. The output circuit is any one of claims 1 to 8, wherein a footback signal obtained by integrating the magnetic field detection signal is input to the pickup coil with a polarity that cancels the magnetic field detection signal, and the sensor output is generated from the feedback signal. The flux gate magnetic field sensor according to item 1.

Images