Magnetic Field Detection Apparatus

The instant invention relates to apparatus for detecting the presence of a magnetic field disturbance. More particularly, the instant invention relates to a magnetic field detector having a pair of spaced-apart field sensors for sensing a variation in the earth's magnetic field caused by the presence of a magnetic anomaly in the vicinity of the magnetic field detector apparatus.

An effective device for measuring magnetic fields and locating magnetic anomalies is the magnetometer. The capability of magnetometers to detect the presence of magnetic anomalies is well known and has found significant use. One magnetometer that has been found to be suitable for detecting the presence of ferromagnetic bodies is the proton magnetometer. This magnetometer is based in principle upon the phenomenon of free nuclear precession.

One instrument in the prior art that utilizes proton magnetometers to detect the presence of ferromagnetic bodies is that described and shown schematically in a book, Physics and Archaeology, authored by M. J. Aitken and published in London in 1961 by Interscience Publishers, Inc. The instrument shown therein utilizes two field sensors separated by a distance and used to measure the average gradient over the distance. The sensors themselves each comprise a bottle of liquid having a coil of wire placed around the outside of the bottle. The two coils are connected in series and are supplied with a polarizing current. Since proton magnetometers operate on the free proton precession phenomenon, if the two bottles are in the same magnetic field strength, the sum of the two signals is merely twice the amplitude of one of the signals and decays to zero. If the bottles experience different fields, however, the frequencies of the signals are different; and the two signals will successively get in and out of phase thereby producing a beat frequency. Because the two sensor signals are being summed, the beat frequency, which is the frequency difference between the two component signals, will result.

An output signal comprising the sum of the two individual sensor signals is applied to a high input impedance preamplifier. The output signal from the preamplifier is then applied to the input of an amplifier circuit tuned to pass only signals of a frequency within a particular band. A final amplifier stage is provided which feeds a loudspeaker to produce an audible sound indicative of the presence or absence of a search object detectable by the instrument.

In one of the more particular details of the instrument, the bottle coils, which are connected in series, are pulsed by a common polarizing current through a single relay contact. A timing flip-flop automatically controls the relay in order to alternately connect the bottle coils to a source of polarizing current and to the input of a preamplifier. This arrangement provides "listening" periods between the polarizing periods.

The period of the beat signal indicates the strength of the disturbance, with the length of time required for the beat signal to drop to its first zero being the least when the detected object is reached. The loudspeaker is provided to make available a readout indication of the beat frequency. In the absence of a magnetic disturbance, a gradually decaying signal of a fixed frequency is heard during each listening period. In the presence of a magnetic disturbance "bleeps" or beat signals are heard, and the greater the strength of the disturbance the more beats or bleeps within the listening period that will be heard. As an alternative to an audio readout, it is suggested in the book that the rate of fall of the beat signal can be timed electrically. Also, the number of oscillations occurring during the listening period or during some portion thereof could be counted to provide a field strength readout.

Although the instrument described is an effective magnetic anomoly sensor, the design integrates the human ear into the detector loop. To obtain a high degree of accuracy in measuring the strength of the magnetic anomoly, it is desirable to measure the period of time between the beginning of a listening period and the first detectable beat signal minimum. The ear's ability to detect a minimum requires generally a higher signal-to-noise ratio than that required by the eye in watching meter deflections, thereby making an accurate determination of small field strengths difficult to obtain with the instrument. The major deficiency in the instrument is that it fails to move the "zero of signal" to the several second interval during which the dying signal can be detected. The zero of signal is a beat infinitely long from the onset of the listening period. A quite sizable search object is required to move the first beat from infinity into the listening period, thereby causing a limitation in this device.

A magnetic field measuring apparatus that utilizes nuclear magnetic resonance in the measurement of a varying magnetic field is that disclosed in U.S. Pat. No. 3,260,926. The apparatus disclosed comprises a magnetometer having an integrating circuit coupled to the output leads of a pair of polarizing coils. A controlled switch is provided which operates in a first position to hold the integrating circuit output at zero and in a second position to apply the voltage induced in the coil windings to the input of the integrating circuit, while simultaneously releasing the output of the integrating circuit.

The signal available from the integrating amplifier circuit is fed to the input terminals of a graphical type recorder. The recorder is disclosed as possibly being an oscilloscope or other form of x-y display. In any event, the resulting visual display produced will show the resonance lines plotted against magnetic field value. Alternatively, the integrating amplifier circuit output voltage could be supplied to an accurate volt meter or digital volt meter. The amplifier output voltage for a given magnetic field intensity will be related by a mathematical relation that involves not only the voltage but the component values of the integrating circuit amplifier and the physical dimensions of the magnetometer probe coil.

Another magnetic field measurement apparatus that utilizes nuclear magnetic resonance type magnetometers is that disclosed in U.S. Pat. No. 3,441,838. The apparatus described therein has a pair of vessels that each contain a material having gyromagnetic properties. Two pairs of coils are used with one coil of each pair being would about a different one of the vessels. One pair of coils is connected in series and coupled to the input of a linear amplifier. The other pair of coils, also serially connected, are further connected to the output of the amplifier. The output of the amplifier is fed to a frequency meter.

In accordance with the present invention, a magnetic detector of the type having a pair of magnetically polarizable liquid bodies carried in spaced-apart relation, each having a separate one of a pair of serially connected magnetic probing coils wound therearound, and including a timer alternately applying current to the coils to polarize the liquid bodies and removing the polarizing current during a preset listening interval intervening the polarizing intervals is provided. In addition to the foregoing structure, a magnetic field detector in accordance with the present invention includes a combination which comprises a high gain signal channel tuned to the frequency of the free precession signals generated in the magnetic probing coils which signal channel serves to amplify the signals. A bypassable earphone jack at the output of the high gain signal channel is provided. A means of rectifying the signal carried on the high gain signal channel at a selected bias level to produce a signal at its output which reduces to zero as the free precession signals in the coils decay is provided. A meter having a meter movement operative to follow low frequency beat signals present in the output signal of the rectifier is provided. The meter provides for the observation and timing of the length of the interval between each removal of the polarizing current and the occurrence of the first subsequent minimum value of the beat signal. The earphones allow aural observation of these beat signals in a fashion analogous to the visual meter observation. A constant field bias coil is provided for reducing the time interval between the start of the listen period and the first beat zero from infinity to a few seconds.

In accordance with the more particular aspects of the present invention, the high gain signal channel comprises a multi-stage preamplifier section having a high input impedance. The high gain signal channel further includes a bandpass amplifier section coupled to the output of the preamplifier. The bandpass amplifier has a parallel resonance network in a feedback path between the input and output of the amplifier. More particularly the parallel resonance network comprises an inductor and a capacitor.

In another more particular aspect of the present invention, the means of rectifying the beat signal from the high gain signal channel includes a rectifier circuit having rectification level control means. The rectifying means further includes a filter network that receives the rectified signal and is connected at its output to a driver amplifier. The rectifier circuit in a particular configuration is a single stage ac amplifier having a biasing network at its input which comprises a variable impedance component that shifts the operating point of the amplifier. The rectifier means acts as a peak detector to extract a low frequency beat signal imposed as an envelope on the amplified sum of the signals induced in the coils.

In another particular aspect of the present invention, at the output of the bandpass amplifier is a bypassable earphone routing jack. When the earphones are plugged into this jack, the amplifier is not operably connected with the rectifier circuit.

In other more general aspects, a magnetic field detector in accordance with the present invention comprises a cylinder formed of nonmagnetic material, which cylinder has at least one window intermediate the ends. A receptacle of polarizable liquid is carried in each end of the cylinder, with each receptacle being surrounded by a coil for polarizing the protons of the liquid contained therein. A housing is carried inside the cylinder proximate the window for enclosing a circuit board having electronic circuitry thereon which is operably connected to the polarizing coils. The cylinder also carries electromagnetic shielding that encloses the interconnections between the electronic circuitry and the coils. A meter operably connected to the electronic circuitry is carried on the cylinder for visual inspection thereof.

In a particular, the cylinder can be constructed of PVC plastic tubing. The housing for the electronic circuitry can also be constructed of PVC plastic tubing of a smaller size than that used for the cylinder.

These and other aspects of the present invention will be more apparent from the description of a particular preferred embodiment which follows below.

A more complete appreciation of the invention may be had by reference to the accompanying drawings, illustrating preferred embodiments of the invention to be described in detail, wherein:

FIG. 1 is an exploded view of a particular magnetic field detector apparatus illustrating mechanical construction details thereof;

FIG. 2 is a schematic diagram of the electronic circuitry used to implement a timer for controlling a relay to alternately polarize the liquid bodies and provide a listening period;

FIG. 3 is a schematic diagram of the electronic circuitry of the high gain signal channel that amplifies the signal from the coils during listening periods;

FIG. 4 is a schematic diagram of the electronic circuitry for rectifying the output signal from the high gain signal channel of FIG. 3 and supplying a rectified version of the signal to the meter readout;

FIG. 5 is a representation of the signal waveform to be found at the output of the high gain signal channel during the listening period when the magnetic field detector is in the presence of a magnetic anomaly; and

FIGS. 6(a)-6(e) are representations of the meter signal at various rectification levels of the waveform shown in FIG. 5.

FIG. 7 is a schematic diagram of the electric circuitry for producing a constant magnetic field difference or bias.

The magnetic field detector of the present invention in general involves the determination of a small difference in magnetic fields at two locations accomplished by the utilization of two sensors placed in spaced-apart relation. In the particular embodiment to be described herein, the sensors are filled with magnetic field sensitive protons embodying the theory of free nuclear precession.

Proton magnetometers typically monitor the precession of protons present in a polarized liquid about the earth's magnetic field. In a conventional proton magnetometer, a coil is wrapped around a bottle containing a proton carrying liquid. Introduction of current into the coil causes the protons to be aligned and point in the same direction, a direction which is at an angle with the direction of the earth's magnetic field. Upon the cessation of current flow in the coil, the protons undergo precession about the earth's field which induces a voltage in the coil. The voltage v induced in a coil by the precession is given by

where f is the frequency of precession in hertz, .mu. is the magnetic moment of the proton, .theta. is the angle between earth's field and the magnetic moment of the spinning electron and t is time. The relationship between the frequency of precession and the intensity of the earth's magnetic field is linear and is given by the equation

Where p is the angular momentum of the proton, .mu. is the magnetic moment of the proton, F is the intensity of the earth's magnetic field, and f is the frequency of precession.

The proton magnetometer is a device that is extremely sensitive to small changes in the earth's magnetic field. As an example, proton magnetometers are suitable for ferromagnetic magnetic field detecting applications. Although a single proton magnetometer can be utilized to monitor the total field intensity, most applications for a magnetic field change detection have as their objective the determination of spatial changes in the earth's field, and thus do not require total field measurements. Moreover, total field can change with time for a great many reasons. Therefore, a magnetic field detector which is insensitive to total field is frequently more desirable. Accordingly, in most magnetic field detector applications, magnetic gradient is more significant than is the total field intensity.

The magnetic field detector of the present invention, having as its object the determination of a magnetic gradient, employs two spaced-apart sensors. Accordingly, the magnetic field detector has the sensing protons arranged to provide a gradient magnetometer and utilizes two bodies of polarizable liquid, each of which has a coil wrapped around it. The coils are connected in series to provide an output signal that is the sum of the two individual coil signals. This voltage sum S appears at the output of the series connection as

By applying the appropriate trigonometric identity to equation (3), the voltage sum S can be rewritten as

The above expression, as will be readily appreciated, has a required maximum at t=0. The voltage sum signal has an envelope, or modulation term cos (f.sub.1 t-f.sub.2 t), that reaches a first minimum when f.sub.1 t-f.sub.2 t=1/2. Since the time to this first minimum from the point in time marked by the cessation of current flow in the coils can be designated t.sub.2, the expression above can be rewritten as f.sub.1 -f.sub.2 =1/2t.sub.2. The longer t.sub.2 is, the smaller the detectable carrier frequency difference which can be measured and, the smaller the detected field strength difference. It will be appreciated that a magnetic field detector using a pair of spaced-apart sensors following the theory of operation outlined above, that the closer the magnetic field detector comes to the object producing the magnetic anomaly, the greater the difference in the detected magnetic field experienced by each sensor. This results in a greater frequency difference in individual sensor induced voltage signals and leads to a shorter time between the cessation of current flow in the coils and the first detectable beat frequency minimum value.

Referring now to FIG. 1 of the Drawings, there is shown in an exploded view a magnetic field detector 10 constructed in accordance with the present invention. More particularly, an elongate cylinder 12 formed of nonmagnetic material is provided. At each of the ends 14 and 16 of cylinder 12, there is carried a magnetically polarizable liquid body. Illustrative of the arrangement for carrying the liquid bodies at the ends of cylinder 12 is the detailed presentation of the mechanical arrangement for so carrying the liquid body in the end 14. The mechanical arrangement is duplicated in end 16 for the other probing coil.

The polarizable liquid body is carried in a receptacle 18 which can be a plastic bottle. Receptacle 18 is surrounded by a coil 20 for polarizing the protons in the liquid contained within receptacle 18. Coil bobbins 22 and 24 are placed on receptacle 18, and it is between coil bobbins 20 and 24 that coil 20 is wound. Looped nylon lines 23 are each strung through a bead 35 which holds the bobbin on receptacle 18 when the two looped nylon lines are tied together at the terminal support disc 32. The assembly comprising the coil, bobbin, nylon lines and bottle is tied to the terminal support disc by stringing the nylon lines through two holes in the disc and tying these lines securely so that the disc is held firmly against the bottom of the bottle. Two bolts 28, 30 pass through the terminal support disc and are tightened by nuts 29, 31 to disc 32. These nut-bolt pairs act as terminals for the coil by securing the coil wires thereto. The bottle 18 and coil 20 are inserted within detection receptacle housing 26, which is itself inserted in the open end 14 of cylinder 12, with a cover 34 being placed over the open end.

Cylinder 12 has a first window 36 and a second window 38 intermediate the ends thereof. Window 36 is provided with an access port cover 40, and window 38 is provided with an access port cover 42. Both covers are removable. The removable access port covers provide access through the windows to the electronics which are carried on a circuit board 44 mounted inside cylinder 12. The electronic circuit board 44 is mounted in a separate one-piece housing 46 having a cut along its length allowing it to clamp onto circuit board 44. The mounting of the electronics in a separate housing secured within cylinder 12 provides a desired resistance to shock potentially encountered by impact of forces created through dropping of the apparatus or striking of the apparatus against an object. Electromagnetic shielding 48 is provided to enclose the interconnections between the electronic circuitry and the probing coils and protect against the picking up of spurious, external signals. Finally a meter 50 is provided which can be mounted on cylinder 12 or carried separately. The bias coil is an assembly coinciding in appearance and mounting principle with the meter assembly.

Cylinder 12 can be constructed of PVC plastic tubing. This material is not only nonmagnetic, but it also provides a high impact resistant, lightweight material. In addition the white plastic provides a reflective shield from solar visible and near infrared radiation. The bottle housings carried in the ends of cylinder 12 can also be PVC plastic tubing of a smaller size than that used for cylinder 12. In a preferred embodiment, the polarizable liquid bodies are distilled water.

As mentioned briefly in the general discussion, the two magnetometer probing coils are connected in series and are energized by a common polarizing current to insure that the two field sensor signals have the same phase and also to accomplish the addition of the two signals without the need of a mixing circuit. As the output terminals of the coils must alternately be connected to a source of current and disconnected therefrom and reconnected to the input of the electronics, means must be provided to automatically perform this timing function and provide "listening" periods of a proper duration. Typically, approximately four seconds is provided for both the listen and polarize phases. A longer listening period will result in greater sensitivity, while a longer polarizing time gives a stronger signal.

Referring to FIG. 2, electronic circuitry for providing timer means to alternately apply current to the coils to polarize the liquid bodies and removing the polarizing current during a preset listening interval intervening polarizing intervals is provided. In FIG. 2, a first container of liquid 52 having a coil 54 wound therearound has one lead 56 connected to the center terminal 58 of a single pole double throw (SPDT) relay 60. The other end 62 of coil 54 passes through shielding 64 and connects to a first lead 66 of a second coil 68 which is wrapped around a container of polarizable liquid 70. The second lead 72 of coil 68 is connected to shielding 64 which is, of course, grounded.

The center contact of single pole double throw relay 60 is actuated by a coil 74. Energization of coil 74 is by current flow from the power supply through resistor 76 connected in series with coil 74. The series combination of resistor 76 and coil 74 are connected to the collector of driver transistor 78 which is operable in response to a signal on its base lead to establish the current flow through coil 74. A capacitor 80 is shunted across coil 74 and driver transistor 78.

The timing cycle in which driver transistor 78 energizes coil 74 to alternately switch relay 60 between one position, in which the lead 56 is connected to the positive power supply potential, and a second position, in which lead 56 is connected to detector circuitry through lead 84, is established by a squarewave generator generally referenced by the numeral 86. In general, the squarewave generator 86 comprises an operational amplifier (op-amp) 88 connected as a comparator. In the circuit configuration shown, a resistor 90 is connected between a source of positive potential and the noninverting input of op-amp 88. A resistor 92 is connected between the output of op-amp 88 and the noninverting input. A threshold level is set at the noninverting input of op-amp 88, with a portion of the output of op-amp 88 being fed back to a noninverting input terminal. In a second feedback path from the output of op-amp 88, there is a series connection of diode 94, resistor 96 and variable resistor 98. A capacitor 100 having one lead thereof connected to ground has the second lead connected to variable resistor 98 and also connected through variable resistor 102 and fixed resistor 104 to the inverting input of op-amp 88.

Under the influence of the reference voltage on the noninverting input of op-amp 88, the output of op-amp 88 will be at a positive voltage. However, as capacitor 100 begins charging, the voltage on the inverting input will increase. The output of op-amp 88 remains until the voltage on capacitor 100 reaches the preset threshold, whereupon the op-amp output reverses to the opposite voltage polarity. In the configuration shown, variable resistor 98 provides a pulse width control which sets the duration of the listen period and of the polarizing period within the total period of a cycle of operation. In addition, variable resistor 102 provides a pulse period control which sets the length of the operational cycle. The output of op-amp 88 is coupled to the base lead of driver transistor 78 through a resistor 106 and to ground by a phase changing capacitor 59 which prevents relay chatter. As the output polarity switches back and forth, transistor 78 is alternately driven into conduction and cut-off. This action of transistor 78 causes coil 74 to be alternately energized and de-energized, which results in a flip-flopping of relay 60.

Referring next to FIG. 3, a high gain signal channel 108 tuned to the frequency of the free precession signals generated in coils 54 and 68 for amplifying the same is presented. More particularly, high gain signal channel 108 comprises a preamplifier section 110 and a bandpass filter section 112. The signal available during the listening phase of the cycle of operation from lead 84 of relay 60 is applied to input terminal 114 of signal channel 108. An input capacitor 116 in combination with coils 54 and 68 forms a tuned circuit and provides broad band input filtering.

Preamplifier section 110 is a three stage amplifier comprising three cascaded Field Effect Transistor (FET) circuits. The signal applied to terminal 114 is coupled to the first stage through a capacitor 118. The signal is applied to the gate lead of FET 120, which is biased by resistor 122. A resistor 124 is connected to the drain of FET 120, and an RC circuit comprising capacitor 126 and resistor 128 is connected between the source lead of FET 120 and ground potential. The output of the first stage is taken from the drain lead of FET 120 and coupled through a capacitor 130 to the gate of FET 132, which is baised by a resistor 134. Resistor 136 and capacitor 138 form an RC circuit in the source lead of transistor 132. A resistor 140 is connected from the drain to a source of positive potential. The output of the second stage is taken from the drain of FET 132 and is coupled to the next stage through a coupling capacitor 142. In addition, a pair of diodes 144 and 146 are connected anode to anode between the drain lead of FET 132 and ground potential to limit the output signal level of the second stage of amplification.

The signal coupled through capacitor 142 is applied to the gate lead of FET 148 which has a resistor 150 connected from the gate lead to ground potential. An RC circuit comprising resistor 152 and capacitor 154 is connected between the source lead and ground potential. A resistor 156 connects between the drain of FET 148 and the source of voltage that supplies the first and second amplification stages through a resistor 158. The output of preamplifier 110 is coupled to the input of bandpass filter section 112 through a capacitor 160.

Bandpass filter 112 comprises an operational amplifier 162 having a parallel resonance network in the feedback path between the output and input thereof. The parallel resonance network comprising inductor 164 and capacitor 166 produces a narrow bandpass active filter which is tuned to the frequency at which the protons are precessing within the liquid bodies in containers 52 and 70. The output signal from op-amp 162 is provided to the signal channel output terminal 168 through a capacitor 170 and a variable resistor 172. At 168 is a bypassable routing jack into which a plug 211 can be inserted thereby disconnecting the amplifier and the rectifying circuit and connecting the amplifier to earphones through a shielded cable.

Referring to FIG. 4, means to rectify the signal available at output terminal 168 of high gain signal channel 108 at a selected bias level to produce a signal which reduces to zero as the free precession signals decay is shown. The means includes a rectifier circuit generally designed by the numeral 174 and an amplifier circuit generally designated by the numeral 176. More specifically, the rectifier circuit 174 is a maximum and minimum peak detector having rectification level control means for shifting the threshold level at the input of circuit 174. Rectifier 174 comprises a transistor 178 which has a biasing network at its input that comprises a variable impedance component, resistor 180, operable to shift the detection level circuit 174. The biasing network further includes a fixed resistor 182. Both resistor 180 and resistor 182 are connected to the base lead of transistor 178. A resistor 184 connected to the collector lead of transistor 178 is tied to a source of positive potential, as is variable resistor 180. A parallel combination of capacitor 186 and resistor 188 are connected between the emitter of transistor 178 and ground potential.

The output of rectifier circuit 174 is taken from the collector of transistor 178 and applied to a filtering network comprising shunt capacitor 190, resistor 192 and parallel shunt capacitors 194, 196. Together, the rectifier circuit and filter network act as an envelope demodulator.

After filtering, the rectified signal is supplied to amplifier circuit 176 which comprises an operational amplifier (op-amp) 198 having the signal from circuit 174 applied to its noninverting input through a resistor 200. A feedback resistor 202 connects to the inverting input of op-amp 198. In addition, a resistor 204 is connected between the inverting input and a source of positive potential. The output of amplifier circuit 176 is applied through a resistor 206 to a meter 208 having a diode 210 shunted across it.

In order to better understand the manner in which circuits 174 and 176 operate to provide a rectification of the output signal from signal channel 108, reference will now be made to FIGS. 5 and 6. Specifically, in FIG. 5, there is shown a representation of the unrectified signal available from high gain signal channel 108. The signal is the sum of the two individual voltage signals induced in the coils 54 and 68 and has the waveform shown when one of the liquid bodies is experiencing an external magnetic field of an intensity greater than the intensity of the magnetic field that the other coil is experiencing. In accordance with equation (4) above, the output signal shown is characterized by a beat modulation frequency that creates an envelope for the carrier waveform. As shown, the beat modulation frequency damps the carrier waveform. Damping depends upon the difference in the magnitudes of the magnetic field intensities experienced by the two coils.

In order to provide a signal suitable for display on a meter movement, the signal of FIG. 5 must be rectified. If the detection level is set by variable resistor 180 to at a level corresponding to level (a) shown in FIG. 5, then only signals having an amplitude greater than level (a) will result in an output signal from circuit 176 to meter 208. For example, referring to FIG. 6(a), the meter signal shown therein will be provided to meter 208. If, however, the bias on the input of circuit 174 is lowered to a level corresponding to level (b) shown in FIG. 5, then the signal applied to meter 208 will appear as that shown in FIG. 6(b). Accordingly, as the detection level is shifted lower to levels corresponding to (c) and (d), then meter signals as shown in FIGS. 6(c) and 6(d) will be obtained.

In order to understand how the magnetic field bias coil operates, reference will be made to FIG. 7. It is readily seen that this is a common resistance bridge with a bias coil 224 in the bridge arm. In the other four arms are two fixed resistances 222 and 223 in two of the arms, and in the other two arms are the two sections of a resistance pot's total resistance, this resistance lying on either side of the wiper arm. This bias coil is mounted in a holder 50 resembling that of the meter holder. It is clamped to the cylinder 12 near one of the two bottle coils. A characteristic of this bridge circuit is that when the switch 225 is on, and the wiper arm of the pot is on either side of mid-position, current will flow in the bias coil, the current increasing as the wiper arm moves progressively away from mid-position. At mid-position the bias coil current is zero. The current characteristic of the wiper arm on one side of mid-position produces a magnetic field through the coil opposite in direction to the field produced with the wiper arm on the other side of mid-position. For a fixed position of the wiper arm and a fixed position of the magnetometer in space, the bias coil's field will add vectorially to the earth's field at the closest bottle and give a fixed resultant field vector thereby producing a fixed, non-zero difference or bias in field between the two bottles. This, in turn, will produce a fixed beat frequency in the output signal, which, as will be explained, is a more sensitive no-search-object signal from which to check signal changes when a search object presents itself.

In operation during the performance of a search for magnetic anomalies, the magnetic field detector apparatus described herein operates as a gradient magnetometer to sense a small difference in the magnetic fields existing at two spaced-apart locations, which difference is the result of the presence of a magnetic anomaly or search object in the vicinity of one of the spaced-apart magnetic coils. The magnetic field detector, through the use of the two coils and measuring of the difference in the magnetic fields experienced by each, eliminates variations in the earth's magnetic field which occur over time. This cancellation of variations in the earth's magnetic field, of course, results from the fact that the variations are commonly experienced by the two coils; and in accordance with equation (4), the equal shift in frequencies f.sub.1 and f.sub.2 is algebraically subtracted out of the modulation term. In addition to these long-term magnetic field drifts, magnetic noise such as that from nearby power lines is also eliminated.

When the magnetic field detector circuitry is energized, as by the connection of a battery power source to the +v leads to the circuitry, timer 86 is activated and begins supplying a squarewave output to switch relay 60 between alternate positions wherein the coils 54 and 68 are first energized by a current from the power supply and subsequently connected to the input of the high gain signal channel 108. Because the high gain signal channel has a bandpass amplifier section which is to be tuned to the frequency at which the protons are precessing within the liquid bodies contained in receptacles 52 and 70, and because the frequency of precession is dependent primarily upon the surrounding magnetic field of the earth, it is necessary to adjust the bandpass amplifier for the particular location where the search is being conducted. Since there are in effect two tuning circuits in the high gain signal channel, tuning of the signal channel to the proper detection frequency is accomplished by adjustment of the value of input capacitor 116 and adjustment of the value of capacitor 166 in the parallel resonant circuit of the bandpass filter 112. Changes in the center frequency of the bandpass amplifier can also be adjusted over a considerable frequency range without changing the capacitor by varying the core position of the inductor 164.

With the signal channel 108 properly tuned, and the coils 54 and 68 each experiencing a different magnetic field intensity, an output signal characterized by a beat modulation frequency will be available at the output terminal 168 of high gain signal channel 108. The carrier waveform of this signal has a frequency of between 1 kHz and 3 kHz. The beat modulation frequency typically ranges between approximately 1/10 Hz to 10 Hz. More generally expressed, the beat modulation frequency will be 1/2t Hz, where t is the time from the beginning of the listening period (i.e., marked by the cessation of current flow in the probe coils) to the first detectable beat minimum. The time t increases as the detectable carrier frequency difference decreases, which corresponds to a decrease in the detectable field strength difference. Therefore, in order to detect small or distant objects with the magnetic field detector, it is necessary to observe the output signal over a long period of time. Observation of the output signal is by observation and timing of the deflections of meter 208, which deflections show the interval between the first carrier frequency maximum and the next subsequent minimum. The output signal from the high gain signal channel, if not directed to the earphones 212, is, of course, applied to meter 208 through the rectification circuitry shown in FIG. 4. In order to properly time the deflections of the meter 208, it is necessary to adjust the rectification circuitry for proper operation. For example, if the meter signal is noisy, a larger capacitance value for capacitor 196 could be needed to provide additional filtering of the rectified proton carrier frequency. However, a balancing must be made between the extent of filtering and the resulting time constant, as a long time constant will affect the sharpness of the beat signal minimum. It is also necessary to adjust variable resistor 172 at the output of the high gain signal channel 108 to control the amplitude of the output signal therefrom. Accordingly, resistor 172 acts as a volume control. Finally and most importantly, the rectification level must be set by adjustment of variable resistor 180. With reference to the plots of FIG. 6, it is apparent that a desirable setting for the rectification level control is one that corresponds to the (d) level in FIG. 5. This level gives the sharpest minimum, as shown in the corresponding plot of FIG. 6 (d). Other settings, either higher or lower, result in a minimum unnecessarily high on the meter readout or a flat minimum.

When the magnetometer is being used in favorable surroundings free of stray magnetic fields, for instance in open country, magnetic field gradients are very low. This being the case, only a gradually dying precession signal is heard in place of the rising and falling beat signals characteristic of a non-zero field gradient. Here the field gradient is the field difference between the two detector positions. A zero field gradient is equivalent to saying, in signal terms, that the beat minimum occurs infinitely long from the start of the signal.

Rewriting a previous expression

in general form as

where n is the sequence number of the minimum, and differentiating equation (2) to obtain the gradient of the magnetic field .DELTA.F

one finds that the combination of (5) and (6) gives

The longer the signal is observed, that is the larger t is, the smaller the observable gradient. Since the signal gradually dies regardless of whether a beat is superimposed on it, there is an upper limit on how long one can observe the signal and still be certain a beat has been detected at the end of the dying signal. For example, if t is equal 3 seconds, field differences from zero gammas to 3.9 gammas are indistinguishable. This result is obtained by substituting infinitely and 3 for t in equation (7) corresponding physically to the "dead" time during which no signal can be heard, this time extending from 3 seconds to infinite seconds.

Therefore, there is considerable advantage in moving the first beat minimum for zero gradient from infinitely far out to, say, 3 seconds out, where the movement of the beat along the time scale can be detected and quantified. This movement or effect is accomplished within the instrument by means of a constant magnetic field bias coil. If, for instance, 0.4 seconds is the accuracy with which the particular operator can reproduce the time to the first beat because of reaction time, stopwatch errors and so on, then when this observer moves from one position to another, he can detect a movement of the first beat minimum down the time scale to no greater accuracy than 0.4 seconds. Assume, therefore, that t is equal to 3 seconds at the first position of detection and t is equal to 2.6 seconds at the second position. This would correspond to a field difference of 11.75/3 or 3.9 gammas at the first position and 11.75/2.6 or 4.5 gammas at the second, a difference of 0.6 gammas between the two positions. Thus, search objects producing a mere 0.6 gamma signal can now be detected whereas before search objects required 3.9 gammas to be detected.

In the present design the beat minimum can be detected significantly longer than 3 seconds out. Thus, an order of magnitude improvement in minimum detectable signal is easily effected, and this constitutes the improvement over the prior art.

Certain operators create a magnetic field bias by tipping the instrument so that one bottle is closer to the earth. Because there are trace magnetic components in the earth a field difference and consequent beats are detected. The earth field suffers from lack of uniformty over time and distance rendering this an unsuitable means for creating a bias.

The signal minima are more easily detected than signal maxima because minima are sharper. This is because the amplifier goes into oscillation during the maxima while it drops out of oscillation during the minima.

If more than one beat minimum is distinguishable during the detection period, the time interval to the last distinguishable beat minimum should be taken. Equation (7) will show that the highest accuracy is given by this practice if the operator's uncertainty in reproducing the time t is more or less fixed. In addition there is a faculty, here called psychocadence, present in nearly all observers to a surprisingly high degree, which allows an observer to continue a cadence mentally after the sensation of the cadence, here the aural or visual beat signal, has died away. This ability can easily be made accurate to 2% over a very long time interval immediately following the brief time interval during which the cadence was sensed. Additionally, in guessing the instant of the last chosen minimum in the described fashion, the mind has effectively integrated the cadence information thereby producing an improvement over the case where the observer depends on the visual or aural sensation as to the instant at which the last chosen minimum occurs. Here "visual" applies to the meter mode and "aural" applies to the earphone mode.

When the ambient magnetic field gradient begins subtracting from the magnetic field gradient of the bias coil, the first minimum will begin moving out to infinity again and will again be difficult to hear. The current in the coil is reversed at this point so that the first minimum is easily heard again.

If the field of the coil is aligned parallel or antiparallel to the earth field, angular deviations of the coil field from the earth field will cause minimum errors. For instance, shifting the beat minimum from infinity to 3 seconds requires a field difference of 3.9 gammas from one detector to the other. If the coil is mounted nearer one bottle than the other, its effect on just the one bottle can be considered because of the very sharp third power falloff of coil field with respect to distance from the coil. If .theta. is the angle between the earth field and the coil field, the algebraic addition of the coil field vector to the earth field vector falls off as cos .theta., which in turn is approximately equal 1-.theta..sup.2 /2 for small angles. Here .theta. is in radians. If .theta. varies as much as plus or minus one-third radian, or approximately 20 degrees, cos .theta. will vary from 1 to 0.94. This 6% variation in the 3.9 gammas, or 0.24 gammas, is quite acceptable when compared to the error limits of the magnetometer from operator reaction time, and so on.

As should now be apparent, during the search for magnetic field changes, what is actually detected is the time interval from the beginning of the listen period to the first beat minimum or any subsequent beat minimum as shown on the meter or detected by the earphones. This time interval can be suitably measured through the use of a stopwatch actuated by the magnetic field detector operator upon his observation of the deflections of the meter movement or upon aural detection using earphones. This manner of measuring the time interval integrates the human eye or ear into the loop of detection. Under certain circumstances the meter technique of measuring the time interval to the first beat minimum is believed superior to the provision of an audio indicator in that a visually indicated minimum under off-scale-maximum conditions allows a smaller signal-to-noise ratio for a given sensitivity than is characteristic of an aural first minimum detection.

The foregoing description of the invention has been directed to a particular preferred embodiment for purposes of explanation and illustration. It will be apparent, however, to those skilled in this art that several modifications and changes in the magnetic field detector apparatus may be made without departing from the scope and spirit of the invention. For example, the rectified output signal can be fed to a chart recorder, rather than to a meter, to provide a permanent record for later study and interpretation. In addition, if a chart recorder is used the time base thereof can be readily correlated with the various positions of the magnetic field detector apparatus when performing a search in the field. These, and other modifications of the invention, will be apparent to those skilled in this art. It is the applicant's intention in the following claims to cover all such equivalent modifications and variations as fall within the scope of the invention.