Single Winding Saturable Core Magnetometer With Field Nulling

This invention relates to magnetometers--i.e., instruments for magnetic field sensing and measurement. More fully, it relates to a new type of magnetometer which uses a coil having only a single winding per axis of detection, and operates in accordance with flux-balance principles.

A magnetometer is a device which measures magnetic field and provides a signal which is related to the field intensity. Magnetometers have long been known and have many applications. There are several types of magnetometers, including those which are mechanical in nature and those which are electrical and electronic. One popular type of electronic magnetometer is called a flux-gate magnetometer. Flux-gate magnetometers operate by saturating and unsaturating a magnetic core about which a sense winding is provided. The operation of saturating and unsaturating the core is achieved with a drive signal applied to a drive winding on the same core. The drive winding alternately concentrates flux lines through the sense winding and allows them to relax. This arrangement thus serves as a gate or valve for the magnetic flux. An external magnetic field interacts with the magnetic field generated by the sense winding in a way which can be detected through electronic circuitry connected to the sense winding. The output signal of a flux-gate magnetometer is a voltage which is linearly proportional to the external magnetic field. The reader will appreciate that compasses can readily be constructed from devices having these properties. Such compasses are useful, for example, in automotive, marine, and aircraft navigation.

In compass applications, two or more magnetometers are generally employed, at least one per axis. Electronic circuitry processes the information about field changes along the two orthogonal axes to derive heading information. We may consider each axis separately as the two magnetometers are normally of the same design and construction. More than two axes are sometimes used, to increase the signal-to-noise ratio of the measurement, or to detect magnetic dip angle.

Prior magnetometers comprise two or more coil windings and a number of electronic circuits. One of the windings is driven by an electronic signal and at least one of the other windings then produces an output which depends not only on the driving signal, but also on changes in flux within the driven and sensing windings produced by movement or reorientation of those windings. For example, in U.S. Pat. No. 4,305,035, issued Dec. 8, 1981, in the name of Dedina O. Mach et al., two coils are used for sensing along each axis. Each coil pair has its own electronics and each coil pair is wound on separate, small strips of core material; the coils are positioned in a gimballed fixture for use as an electronic compass. One of the coils in each pair is a driving coil to which a driving signal is applied and the other coil is a field-sensing coil. In a two-axis compass application, each of the sensing coils senses only the component of magnetic field in one plane, with one output being representative of the field along the axis of one of the coils and the other being representative of the field along the axis of the other coil. The driving signal applied to each driving coil is of triangular waveshape, to saturate the coil core in alternating, opposite directions. An applied field has an additive effect during one-half of the excitation cycle and a subtractive effect during the other one-half cycle, producing a tendency to establish a net dc level in the sensed waveform. This dc component is removed and the signal required to do so is used as a representation of the applied field strength. As seen in FIG. 1 of that patent, the there-indicated embodiment of the invention not only requires the aforementioned multiple coils, but also (for each axis) six operational amplifiers and numerous additional components.

A flux-gate magnetometer which employs three coils and one drive winding for two-axis sensing is marketed by KVH Industries, Inc., Middleton, R.I., as part of that company's KVH PC103 flux-gate compass and related products. According to KVH literature, its sensor uses a toroidal flux-gate magnetometer with a free floating ringcore in the center. The core is made of a stainless steel bobbin wound with Permalloy (a registered trademark) brand tape. A drive field is applied to the core and the external field interaction with the drive field produces an asymmetric change of core flux. This change in core flux is detected by a secondary winding over the core and the resultant signal is then processed.

Other multiple winding flux-gate magnetometer compasses are shown, for example, in U.S. Pat. No. 3,899,834 and in U.S. Pat. No. 4,277,751.

Applications exist, however, for smaller and less expensive magnetic field sensors. The aforementioned U.S. Pat. No. 4,277,751, for example, indicates a need for magnetometers of reduced power dissipation.

It is therefore an object of the present invention to provide a very low-cost magnetometer.

At least partially in accordance with the foregoing, another object of the present invention is to provide a magnetometer using a reduced number of coil windings.

Yet another object of the present invention is to provide a magnetometer using fewer circuit components than prior art magnetometers.

The foregoing and other objects and advantages of the present invention (which will become more apparent from reading the detailed description below) are achieved using a magnetometer which requires only a single coil winding to sense the magnetic field along a single axis. The magnetometer relies on the use of a saturable core material of high permeability, and employs a force-balance feedback loop. The shape of the core depends upon the application; a toroidal core may be used for sensing current, while a cylindrical core may be used for sensing the earth's magnetic field.

Functionally, the invention comprises an oscillator, an integrator, and a voltage-controlled current source. The oscillator uses a saturating inductor which also serves as the magnetic-field-sensing element. The inductor drive alternates between positive and negative, as follows: Briefly, the inductor is driven with a positive voltage until the current through the inductor exceeds a value which indicates that the core is saturated; at that point, the driving voltage switches to an equal-magnitude negative value. The negative drive is applied until the current through the inductor exceeds a value which indciates that the core is saturated in the opposite direction, at which time the driving voltage switches back to a positive value. With no externally applied field, zero average inductor current is achieved. An externally applied field, however, causes the core to saturate in more readily one direction and less readily in the other; this results in a change in average inductor current. A departure of the duty cycle of the driving voltage from the nominal 50% is a side-effect of the non-zero inductor current. An integrator is used to provide compensation for the change in the average current. The compensating voltage output from the integrator further provides the instrument's output signal.

The invention is distinguished by the need for only a single coil winding for each axis, by the use of flux balancing and by the architecture and simplicity of the electronics, with attendant low power requirements and low manufacturing cost.

The invention will be more fully understood by reference to the detailed description below, which should be read in conjunction with the accompanying drawings, which are incorporated by reference herein.

In the drawing,

FIG. 1 is a partially-block, partially-schematic circuit diagram of a flux-balance magnetometer according to the current invention;

FIG. 2 is an idealized representation of a typical simple B-H curve for a saturable magnetic material;

FIG. 3 are observed B-H curves for three magnetic materials which may be used as core material for the inductor of the present invention;

FIG. 4 is a diagram of a piece-wise linear approximation of a representative B-H curve, depicting the effects of an externally applied magnetic field;

FIG. 5 is a diagram showing the effect of an applied magnetic field, on the areas used in the calculation of net inductor current;

FIG. 6 is a schematic circuit diagram of a first embodiment of a magnetometer according to the present invention;

FIG. 7 is a schematic circuit diagram of a second embodiment of a magnetometer according to the present invention; and

FIG. 8 is a partially-schematic circuit, paritally-block diagram of a third embodiment of a magnetometer according to the invention, with compensating drive applied to the sense winding via an auxiliary winding.

A more complete functional description of the basic flux-balance magnetometer according to the present invention, will now be provided with reference to FIG. 1 and subsequently identified figures.

FIG. 1 shows not only a block diagram of the invention, but also a very basic implementation of certain of the blocks. Briefly, the magnetometer 10 comprises an oscillator 12, an integrator 14 and a voltage-controlled current source 16. The saturating inductor 20 of the oscillator 12 is the magnetic field-responsive sensing element.

A voltage-controlled current source 16 is provided by a single resistor 22 connected between the output of integrator 14 and one end of inductor 20 (the other end of inductor 20 being connected to ground). Thus, the current source is driven by the variable voltage at the output of the integrator; this voltage is the integrated value of the voltage at node 21. A comparator 24 controls the polarity of the drive to the inductor (and, therefore, to the integrator). Comparator 24 receives at its non-inverting input the voltage across the inductor 20; a reference voltage is applied to the inverting input of the comparator. The reference voltage is obtained by placing in series between the comparator's output and ground a bipolar voltage source (i.e., a pair of reverse-connected Zener diodes) 26 and a resistor 28, the latter being connected between ground and the bipolar voltage source. The reference voltage is that voltage at node 34 (the junction of resistor 28 and source 26). The output of the comparator 24 is also connected to the junction of source 16 (i.e., resistor 22) and inductor 20 through a series resistor 32.

In operation, inductor 20 is driven with a positive voltage until the current to the inductor exceeds the value which indicates that the inductor's core is saturated (which value is established by the reference voltage at node 34). At that time, the voltage at node 33 changes toward zero and comparator 30 switches state; thus the polarity of the driving voltage at the comparator's output changes. The comparator switches when the voltage across resistor 32 exceeds the voltage of the voltage source 26. When the core saturates in the reverse direction, a similar switching action occurs. The value of the current indicating core saturation is set by the bipolar voltage source 26 and resistor 28. (The bipolar source may be replaced by a resistor, but errors would increase since the current at which the comparator switches would be affected by asymmetry of the output voltage from the comparator 24, and the fact that the permeability of inductor 20 is not zero at saturation.) The saturation current threshold is thus given by the relationship ##EQU1## where I.sub.lim represents the saturation current, V.sub.Z represents the magnitude (unsigned) of the bipolar voltage source output (i.e., the limited swing of the output of comparator 24) and R.sub.32 represents the resistance of resistor 32. The value of the swings at the output of the comparator 24 is set by the comparator itself, and may be a function of the power supplies used. Any asymmetry of the magnitude of the positive and negative comparator output swings causes a second-order error. If the voltage source 26 is ideal, this error vanishes.

Thus, the frequency of oscillation is determined by the driving voltage, the current limit, and the characteristics of the inductor core. The long-term time-average voltage across an inductor must, as is well known, be zero. This implies that the current through the inductor 20, with the feedback loop broken (i.e., no current through resistor 22), is given by ##EQU2## where V.sub.switch represents the output voltage of the comparator 24, I.sub.L represents the current through inductor 20. The coupled brackets [and] indicate the average over a relatively long time of the bracketed variable or expression.

As indicated above, an externally applied magnetic field will "help" the core saturate in one direction and "hinder" it in the other. This has the result of causing the average inductor current to change and depart from zero. When the feedback loop is broken at resistor 22 and the average inductor current is non-zero, then the average voltage must appear across resistor 32. Thus, the latter is acting as a current sense resistor. In the presence of an external field, therefore, the driving voltage applied to the inductor 20 is lower on one-half cycle than the other, as a direct consequence of the fact that resistor 32 has a non-zero value. This, in turn, causes the duty cycle of the driving voltage to depart from fifty percent.

Since the time-average value of the voltage across the inductor (i.e., [V.sub.L ]) is zero, the node 33 at the "top" (i.e., ungrounded) end of the inductor 20 can be used as a current-summing node; this leads to an analysis similar to that used for assuming a "virtual ground" at the inverting input of operational amplifier circuits. Thus, the output of the integrator 14 is connected to a first end of resistor 22, while the other end thereof is tied to the summing node 33. The other connections to node 33 are the non-inverting input of comparator 24 (which draws virtually no current) and one end of resistor 32. Consequently, resistor 22 acts like a voltage-controlled current source and with the feedback loop closed, the integrator 14 functions to replace the average current flowing through resistor 32 with a direct current via resistor 22.

Assuming an ideal integrator, the average current through resistor 32 becomes zero, and the oscillator returns to a fifty-percent duty cycle. The current through resistor 32 is the current required through the turns of the inductor to cancel exactly the external field. The extent to which this operation departs from the ideal (i.e., the extent of the cancellation) is limited by the finite gain of a realizable integrator. Permeability changes in the core due to temperature, for example, do not (to any practical extent) affect the current through resistor 22. Neither do they change the integrator output voltage.

The voltage output of the integrator is the output signal of the instrument, at node 36. The ripple on the output voltage is dependent upon the integrator time constant and the oscillator frequency.

The feedback loop employed in FIG. 1 is of the force-balance type. In addition to making possible the single-winding topology, it also linearizes the response to magnetic field variations and extends the input range of the instrument.

For a more detailed explanation of how the application of an external field affects the operation of the oscillator, the reader is now referred to the simple B-H curve 50 shown in FIG. 2. There, the y-axis represents the magnetic flux density, B, which may be measured in volt-seconds per square meter. Restricting attention to the region within the windings of a cylindrical inductor, the flux density will increase in proportion to the integral of the applied voltage: ##EQU3## where A is the cross-sectional area of the cylinder, in square meters. The x-axis represents magnetic field intensity, H, which may be measured in ampere-turns per meter. In this case, the "meter" represents the measure of the "effective magnetic path length." This is easiest to visualize as the circumference of a toroidal core of small cross-section. For a cylindrical core, part of the path is in the core and part is in the surrounding air. The field intensity is proportional to the number of turns in the inductor for a given current, since the value of the current is identical for all turns. That is: ##EQU4## where I.sub.L represents the inductor current, N.sub.t represents the number of turns and m represents the effective magnetic path length.

Magnetometer operation is sensitive to hysteresis effects in the inductor core, so a core material must be selected for the proper hysteresis qualities. In general, the smaller the remnant magnetization 52 of the core material, the smaller the offset current which may be contributed by hysteresis.

FIG. 3 shows typical B-H curves for several core materials: supermalloy--curve 54; Orthonol (a registered trademark of Spang & Company, Butler, Pa.)--curve 56; and Silectron (a registered trademark of Allegheny Ludlum Steel Corporation, Pittsburgh, Pa.)--curve 58. Each material is characterized by its saturation flux density (B.sub.sat), permeability (.mu.), and the shape of the B-H curve (particularly the shape of the "corners"). Supermalloy may be called a "square-loop material", whereas Silectron may be called "round-loop."

To further simplify the explanation, an approximation will now be used for the B-H curve, ignoring hysteresis and offering a piece-wise linear model for the curve's shape. This approximation is shown in FIG. 4. There, for values of H greater than H.sub.sat, B increases at a rate .mu..sub.sat, which is non-zero and substantially smaller than .mu.. The permeability, .mu., may be defined as B.sub.sat /H.sub.sat. In the presence of an external field, H.sub.ext, the oscillator will act as if the B-H curve 60 has shifted off of the origin along the H-axis (e.g., to the position of curve 62). This movement is equivalent to that which would be produced by a second winding with a constant current, or by a current source in parallel with the single winding. This leads to an assymmetry in the current waveform, I.sub.L (t).

A graphical explanation can also be given and will be discussed with reference to FIG. 5. Assuming R.sub.32 is small and the permeability at saturation is zero, the "operating point" moves at a constant rate along the B-axis between -B.sub.lim and +B.sub.lim, changing directions at the current limits, or "corners", along with the sign of the voltage, V, since the saturation permeability is zero. In the presence of an external magnetic field, the operating point spends more time on one side of the B-axis than the other side; the resulting areas used to calculate net inductor current (i.e., q and r shown on FIG. 5) are therefore unequal. This implies an average value of inductor current exactly equal and opposite to that induced by the applied field.

Another way of thinking about the operation is that the operating point of the oscillator can be "averaged" and represented as a single point on the B-H curve. Since the apparent B-H curve of the core is symmetrical about the H-axis for all values of H.sub.ext, then the average B will always be zero. This is the same as stating that the time average voltage across the inductor is zero, as was done earlier. The "average" H will be equal and opposite to H.sub.ext. Therefore, the time average value of the current I.sub.L must be proportional to H.sub.ext. The saturation "knees" or "corners" may be thought of as a way to reset the constant of the integral, C, ##EQU5## by forcing B to a known value, B.sub.sat. Note that the sensitivity of B.sub.sat to changes in the current limits is inversely proportional to .mu..sub.sat.

The average inductor current over one-half cycle is ##EQU6## which is the sum of the two areas q and r in FIG. 5: ##EQU7##

FIG. 6 provides a complete schematic diagram of a simple embodiment of the invention, showing all component values. Amplifiers 24 and 72 may be any of a large number of commercially available operational amplifiers; optimally, they may both be the same type of amplifier on the same substrate, in the same package, which will minimize cost and size. The integator is provided by a common non-inverting integrator circuit of a type well-known to circuit designers. The Zener diodes 26A and 26B in source 26 are rated for 6.2 volt zener voltage. Capacitor 84 is used to guarantee oscillation by preventing the comparator from getting trapped in a stable state in which both of its inputs and its output are all at zero; with the addition of capacitor 84, both inputs will not go to zero at the same time when the comparator output switches through zero. With such an instrument, a sensitivity of approximatley 10 mV/mGauss has been observed.

A slightly more elaborate implementation is shown in FIG. 7. There, element 24 is a actually a true voltage comparator rather than an ordinary operational amplifier. The Zener diodes previously shown in the bipolar voltage source 26 have been replaced by resistor 92. Note that the first stage of the integrator (i.e., element 94 and associated components) inverts the output relative to its input, and that the second stage of the integrator (i.e., element 106 and associated components) re-inverts, to properly phase the integrator output. The inductor comprises approximately 50 turns of wire on a model 50086-2F toroidal core from Magnetics, Inc. of Butler, Pa. Note that the inherent hysteresis of the comparator obviates the need for capacitor 84.

Having thus described the invention and multiple embodiments thereof, it is our intention that various alterations, modifications and improvements which are not expressly stated herein are nevertheless suggested and will readily occur to those skilled in the art to which the invention pertains. For example, embodiments of the invention need not be limited to the use of single-winding inductor assemblies. The same principles may be used in devices of two or more windings. As shown in FIG. 8, for example, the compensation current could be supplied by a circuit 110 which (directly or indirectly) senses the deviation of the inductor current from zero and supplies a compensating drive via another (auxiliary) winding 112 instead of via resistor 22. Thus the foregoing discussion is intended to be illustrative only and not limiting. To the extent they are obvious herefrom, those alterations, modifications and improvements are intended as part of the invention, which is limited only by the following claims and equivalents thereto.