Method for operating a multi-frequency metal detector and multi-frequency metal detector

Abstract

A metal detector (1) has a drive coil (L61) and at least one detection coil (L62, L63) that detect fluctuations in a magnetic field generated by the drive coil, caused by metallic particles present in an inspected object. A multi-frequency transmitter unit (10) has a converter (4) with a plurality of drive switches (S41, S42; S43, S44) driven by a drive controller (2). The drive switches alternately conduct a drive current through the drive coil to generate an electromagnetic field with two or more different frequency components. A waveform of the drive current is determined, as is at least one pulse-modulated (PXM) signal corresponding to the determined waveform. The determined PXM-signal is determined online or is stored in a memory module (231; 232). The determined PXM-signal is generated and applied to control the drive switches. The drive current can be applied to the drive coil through an admittance unit (5).

Claims

1. A method for operating a metal detector which has a drive coil that produces an electromagnetic field in a product, at least one detection coil that is arranged to detect fluctuations in a magnetic field caused by the presence of metallic particles in the product, and a multi-frequency transmitter unit having a converter with a plurality of drive switches that are driven by a drive controller according to operating instructions, with the drive switches alternately conducting a drive current (i.sub.D) through the drive coil so that the generated electromagnetic field exhibits two or more different frequency components (f.sub.D1, f.sub.D2); the method comprising the steps of: determining a waveform of the drive current (i.sub.D) for the at least two different frequency components (f.sub.D1, f.sub.D2); determining at least one pulse modulated (“PXM”) signal (s.sub.PXM) that is a modulated pulse sequence which corresponds to the determined waveform of the drive current (i.sub.D), the at least one PXM signal the at least one PXM signal being either a pulse-width modulated signal or a pulse-density modulated signal; and controlling the drive switches by generating and applying the at least one determined PXM-signal (s.sub.PXM) to the drive switches, wherein the at least one determined PXM-signal (s.sub.PXM) is determined online or is stored in at least one module of a memory according to the operating instructions provided.

2. The method of claim 1, wherein the step of determining the waveform of the drive current (i.sub.D) comprises the step of: superposing current components relating to the at least two different frequency components (t.sub.D1, f.sub.D2), including sinusoidal frequency components (t.sub.D1, f.sub.D2) that are odd and/or even harmonics.

3. The method of claim 2, further comprising the step of: determining the waveform of the drive current (i.sub.D) at least for a cycle duration of a frequency component (f.sub.D1) with a lowest frequency present in the drive current (i.sub.D) and generating the drive current (i.sub.D) by sequentially repeating the determined waveform of the drive current (i.sub.D).

4. The method of claim 1, wherein: the step of determining the PXM-signal (s.sub.PXM) is achieved by approximating a triangular or trapezoidal signal to the determined waveform of the drive current (i.sub.D) such that the maxima and minima of the determined waveform of the drive current (i.sub.D) and the maxima and minima of the triangular signal correspond to one another and/or coincide and defining switching angles (α1, α2, . . . ) sequentially for falling and rising edges of the PXM-signal (s.sub.PXM) at the maxima and minima of the triangular or trapezoidal signal.

5. The method of claim 1, wherein the step of determining two or more PXM-signals (s.sub.PXM) each with different sets of frequency components (f.sub.D1; f.sub.D2) and storing the two or more PXM-signals (s.sub.PXM) in the at least one module of the memory, which are selectable for generating and applying one of the stored PXM-signals (s.sub.PXM).

6. The method of claim 1, wherein the step of applying the at least one determined PXM-signal (s.sub.PXM) to the drive switches is achieved by either: applying the PXM-signal to a first and a second drive switch of the plurality of drive switches, wherein the first and second drive switches form a half-bridge circuit that is connected on one side to a first voltage potential (V.sub.D) and on the other side to a second voltage potential (V.sub.S), which are connected at a centre tap of the half-bridge circuit, and which are controlled such that the first end of the drive coil, which is connected to the centre tap, is alternately connected to the first voltage potential (V.sub.D) and to the second voltage potential (V.sub.S), or applying the PXM-signal to the plurality of drive switches, which are arranged as a bridge circuit comprising a first and a second branch that are connected on one side to a first voltage potential (V.sub.D) and on the other side to a second voltage potential such as ground potential and that comprise each a first or second centre tap, respectively, connected to a first and a second end of the drive coil with a first pair of the drive switches arranged in the first branch and connected to one another at the first centre tap and with a second pair of the drive switches arranged in the second branch and connected to one another at the second centre tap and which are controlled such that the first and second end of the drive coil are alternately connected to first voltage potential (V.sub.D) and to the second voltage potential.

7. The method of claim 1, wherein the step of applying the at least one determined PXM-signal (s.sub.PXM) to the drive switches is achieved by applying the PXM-signal (s.sub.PXM) via a drive unit that comprises signal drive elements and/or inverters individually to control inputs of the drive switches.

8. The method of claim 1, further comprising the step of: guiding the drive current to the drive coil, directly or via an admittance unit to the drive coil, wherein the admittance unit together with the drive coil form a plurality of resonant circuits that are active in the two or more different frequency components (f.sub.D1, f.sub.D2), wherein each resonant circuit has coil current (i.sub.L61) is larger than the drive current (i.sub.D).

9. The method of claim 8, further comprising the steps of using at least a first branch with a first capacitor and a first inductor in the admittance unit that together with drive coil form a first resonant circuit of the plurality of resonant circuits and a second branch with a second capacitor and a second inductor in the admittance unit that together with drive coil form a second resonant circuit of the plurality of resonant circuits.

10. The method of claim 8, comprising the steps of: selecting the PXM-signal (s.sub.PXM) with a set of frequency components (f.sub.D1; f.sub.D2); and activating the resonant circuits in the admittance unit that corresponds to the set of frequency components (f.sub.D1; f.sub.D2) of the selected PXM-signal (s.sub.PXM).

11. The method of claim 1, further comprising the step of: determining the waveform of the drive current (i.sub.D) at least for a cycle duration of a frequency component (f.sub.D1) with a lowest frequency present in the drive current (i.sub.D) and generating the drive current (i.sub.Dd) by sequentially repeating the determined waveform of the drive current (i.sub.D).

12. A device for detecting metal in an object being inspected, comprising: a drive coil that produces an electromagnetic field in the object; at least one detection coil, arranged to detect fluctuations in the electromagnetic field, caused by metallic particles present in the object being inspected; a multi-frequency transmitter unit, comprising: a drive controller, provided with a memory unit that stores a set of operating instructions and data for at least one pulse modulated (“PXM”) signal (s.sub.PXM) that corresponds to a pre-determined waveform; and a converter having a plurality of drive switches driven by the drive controller according to the set of operating instructions such that the drive switches alternately conduct a drive current (i.sub.D) through the drive coil to generate the electromagnetic field which exhibits at least two different frequency components (f.sub.D1, f.sub.D2) based on the pre-determined waveform.

13. The device of claim 12, wherein: the set of operating instructions and data stored in the drive controller comprises instructions for doing at least one of the following: applying the at least one PXM-signal (s.sub.PXM) to a first and a second of the plurality of drive switches, which form a half-bridge circuit that is connected on one side to a first voltage potential (V.sub.D) and on the other side to a second voltage potential (V.sub.S), which are connected at a center tap of the half-bridge circuit, and which are controllable such that the first end of the drive coil, which is connected to the center tap, is alternately connectable to the first voltage potential (V.sub.D) and to the second voltage potential (V.sub.S), or applying the at least one PXM signal (s.sub.PXM) to the plurality of drive switches, which are configured as a bridge circuit comprising a first and a second branch, each branch being connected on one side to a first voltage potential (V.sub.D) and on the other side to a second voltage potential such as ground potential, the bridge circuit also comprising a first or second center tap, respectively, connected to a first and a second end of the drive coil with a first pair of the drive switches arranged in the first branch and connected to one another at the first center tap and with a second pair of the drive switches arranged in the second branch and connected to one another at the second center tap and which are controllable such that the first and second end of the drive coil are alternately connectable to first voltage potential and to the second voltage potential.

14. The device of claim 12, further comprising: an admittance unit that connects the converter to the drive coil, such that the admittance unit, together with the drive coil, forms at least two resonant circuits, each with a resonant frequency that is tuned at least approximately to one of the at least two different frequency components (t.sub.D1, f.sub.D2).

15. The device of claim 14, wherein the admittance unit comprises: a first branch with a first capacitor and a first inductor in the admittance unit that together with drive coil forms a first of the resonant circuits, and a second branch with a second capacitor and a second inductor in the admittance unit that together with drive coil forms a second of the resonant circuits.

16. The device of claim 14, further comprising: means for activating and deactivating the resonant circuits, individually, in the admittance unit.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) Detailed aspects and examples of the invention are described below with reference to the drawings, wherein:

(2) FIG. 1a shows an inventive metal detector 1, which comprises a converter 4 with four drive switches S41, S42, S43, S44 that for a bridge and that are controlled by means of a selectable pulse-width or pulse-density modulated signal s.sub.PXM, below called PXM-signal, provided by a drive controller 2 and that provide a drive current i.sub.D via an admittance unit 5 to a drive coil L61;

(3) FIG. 1b shows the metal detector 1 of FIG. 1a, which comprises a converter 4 with two drive switches S41, S42 that form a half-bridge and that are controlled by means of a selectable PXM-signal s.sub.PXM provided by the drive controller 2 and that provide a drive current i.sub.D via the admittance unit 5 to the drive coil L61;

(4) FIG. 2 shows the metal detector 1 of FIG. 1a equipped with a drive controller 2, which provides a first PXM-signal s.sub.PXM1 used for controlling drive switches S41, S42 and a corresponding second PXM-signal s.sub.PXM2 used for controlling drive switches S43, S44;

(5) FIG. 3 shows the metal detector 1 of FIG. 1a equipped with a drive controller 2 that allows selecting one of a plurality of stored PXM-signals s.sub.PXM, each provided with a specific set of frequency components, and with an admittance unit 5 that allows selectively connecting at least one of a plurality of branches, which comprise each at least one capacitor C51; C52; C5n and at least one inductor L51; L52; L5n, to the drive coil L61 in order to create resonant circuits tuned to the frequency components of a selected PXM-signal s.sub.PXM;

(6) FIG. 4 shows a mathematically determined waveform of the drive current i.sub.D or i(ωt) that includes four frequency components ω, 3ω, 7ω and 17ω as well as the related PXM-signal s.sub.PXM determined by approximation of a triangular signal i.sub.DA to the determined waveform of the drive current i(ωt); and

(7) FIG. 5 shows the coil current i.sub.L61 in the drive coil L61, the currents i.sub.L51, i.sub.L52 in the branches of the admittance unit 5 and the drive current i.sub.D delivered by the drive switches S41, S42, S43, S44 in the converter unit 4 for the complete frequency spectrum of the metal detectors of FIG. 1a, FIG. 1b, FIG. 2 and FIG. 3 indicating that the drive current i.sub.D is significantly lower than the coil current i.sub.L61 if the frequency components f.sub.D1, f.sub.D2 of the drive current i.sub.D are set to the resonant frequencies f.sub.RES1, f.sub.RES2 of the admittance circuit 5.

DESCRIPTION OF EMBODIMENTS

(8) FIG. 1a shows a first embodiment of an inventive metal detector 1 that comprises a transmitter unit 10 and a receiver unit 11 and a balanced coil system 6 with a drive coil L61 connected to the output of the transmitter unit 10 and two detection coils L62 and L63 connected on one end to ground potential and with the other end to an input stage 7 of the receiver unit 11. In the input stage 7 the input signal is typically amplified and filtered and then forwarded to a phase detector 8. The phase detector 8 allows distinguishing between the phases of the signal components of different origin and obtaining information about the observed product and contaminants, if present. A typical phase detector, e.g. a frequency mixer or analogue multiplier circuit, generates two independent voltage signals which represent the in-phase and quadrature component provided by the input stage 7, and a reference signal fm provided by the transmitter unit 10. The output signal of the phase detector 8 is further processed in a control unit 9, which is preferably equipped with a signal processor, input output devices, a keyboard and a display. By means of the control unit 9 the user can control the operation of the metal detector 1. In particular the user can select operating conditions of the metal detector, particularly the applied drive currents and operating frequencies as described below. The receiver unit 11 may include further features as generally known from conventional metal detectors.

(9) The transmitter unit 10 is a multi-frequency transmitter that is designed to provide a drive signal with a plurality of frequencies, e.g. two to eight frequencies, providing good sensitivity in a broad range of products and contaminants. The transmitter unit 10 comprises a drive controller 2, a drive unit 3, a converter 4 and preferably an admittance unit 5, which forwards a drive current i.sub.D provided by the converter 4 to the drive coil L61.

(10) In this embodiment of the invention the drive controller 2 comprises a memory unit 23 with one memory module 231 in which data that relates to the state of the drive switches S41, S42, S43, S44 for every clock cycle for a long period of the lowest operating frequency, e.g. switching angles α1, α2, . . . of a pulse-width modulated signal or PXM-signal s.sub.PXM, are stored at related addresses. Determination of a pulse-width modulated signal s.sub.PXM is discussed below with reference to FIG. 4. As outlined above, any pulse sequence PXM-signal s.sub.PXM that corresponds to the drive current when applied to the drive switches S41, S42, S43, S44 can be used. Preferably, a pulse-width modulated signals or sequence or a pulse-density modulated signal or sequence is applied. Hence, instead of using the acronym PWM for pulse-width modulated signals and PDM for pulse-density modulated signals, the acronym PXM is used, which stands for a modulated pulse sequence that corresponds to the drive current.

(11) After the system has been reset by reset signal rs issued by the control unit 9, the memory module 231 is sequentially addressed by the address counter 22 with address signal ad such that the data of the PXM-signal s.sub.PXM is sequentially read out of the memory module 231 and applied via the drive unit 3 to the drive switches S41, S42, S43, S44. The PXM-signal s.sub.PXM is routed via drive elements 31 and 311 to the input of drive switch S41, via drive elements 32 and 321 to the input of drive switch S42, via drive elements 31′ and 312 to the input of drive switch S44 and via drive elements 32′ and 322 to the input of drive switch S43. The drive elements 32 and 32′ are inverters which ensure that the drive switches S42 and S43 always open, when the drive switches S41 and S44 are closed and that the drive switches S42 and S43 always closed, when the drive switches S41 and S44 are open. In this way an alternating current is flowing through the drive coil L61 while a short-circuit is avoided. For the sake of simplicity of the drawing, the elements 31, 31′ and 32, 32′ have been duplicated. The output of element 31 could however be connected to the inputs of elements 311 and 312 and the output of element 32 could be connected to the inputs of elements 321 and 322 without requiring the elements 31′ and 32′.

(12) In order to obtain phase coherent operation of the metal detector a clock unit 21 is provided, which delivers reference signals fm to the address counter 22, the memory unit 23 and the phase detector 8.

(13) Since data of the PXM-signal s.sub.PXM are preferably stored for only one period of the lowest operating frequency, data are repetitively read out from the memory module 231. The segment of the drive current i.sub.D shown in FIG. 4 it is therefore sequentially and repetitively produced until the user terminates operation or changes settings. The address counter 22 therefore counts from the lowest address number to the highest address number and restarts at the lowest address number.

(14) The drive switches S41, S42; S43, S44 are arranged in a full wave bridge circuit or H-bridge comprising a first branch and a second branch that are connected on one side to a drive voltage V.sub.D and on the other side to ground potential. The first branch comprises a first centre tap connected to the first end of the drive coil L61. The second branch comprises a second centre tap connected to the second end of the drive coil L61. A first pair of the drive switches S41, S42 is arranged in the first branch of the bridge and connected to one another at the first centre tap. A second pair of the drive switches S43, S44 is arranged in the second branch and connected to one another at the second centre tap. As described above, by applying the PXM-signal s.sub.PXM to the drive switches S41, S42; S43, S44 the first and second end of the drive coil L61 are alternately connected to the drive voltage V.sub.D and ground potential, respectively.

(15) The converter 4 converts the PXM-signal s.sub.PXM into a drive current i.sub.D which comprises desired frequency components, preferably harmonics of the lowest frequency, for example according to the formula, which has been used to determine the PXM-signal s.sub.PXM:
i(ωt)=I sin(ωt)+I/3 sin(3ωt)+I/7 sin(7ωt)+I/17 sin(17ωt).

(16) As described below with reference to FIG. 4 the PXM-signal s.sub.PXM this preferably created according to such a formula so that after conversion in the converter 4 a drive current i.sub.D e.g. with the four frequency components of this formula or a close approximation thereof is generated. Other harmonics are preferably avoided or suppressed. In order to increase sensitivity the desired harmonics are expanded. Further, it would be desirable to generate high coil currents i.sub.L61 with comparably small drive currents i.sub.D. These objects are reached by guiding the drive current i.sub.D via an admittance unit 5 to the drive coil L61.

(17) In the embodiment shown the admittance unit 5 comprises several branches, each provided with a capacitor C51; C52; C5n and an inductor L51; L52; L5n. The number n of branches corresponds to the number of frequency components present in the drive current i.sub.D. Each of the branches C51, L51; C52, L52; C5n, L5n forms together with the drive coil L61 a resonant circuit tuned to the corresponding frequency components ω, 3ω, 7ω, 17ω of the drive current i.sub.D. The coil currents i.sub.L61 in the drive coil L61 at resonance are significantly larger than the drive current i.sub.D. Hence, on the one hand, the drive current i.sub.D flowing in the drive switches S41, S42; S43, S44 can be reduced, while high coil currents i.sub.L61 are reached. The converter 4 can therefore be dimensioned for lower currents and can be built at reduced costs.

(18) The metal detector of FIG. 1a is tuned to a specific set of frequencies of the PXM-signal s.sub.PXM stored in the drive controller 2. The admittance unit 5 with its branches C51, L51; C52, L52; C5n, L5n is fixed to resonate together with the drive coil L61 at this set of frequency components ω, 3ω, 7ω, 17ω.

(19) FIG. 1b shows the metal detector 1 of FIG. 1a in an embodiment with a converter 4 with two drive switches S41, S42 that form a half-bridge and that are controlled by means of a selectable PXM-signal s.sub.PXM, such as a pulse-width or pulse-density modulated signal, provided by the drive controller 2. The drive switches S41, S42 provide a drive current i.sub.D via the admittance unit 5 to the drive coil L61. The drive switches S41, S42 form a half-bridge circuit that is connected on one side to a first voltage potential V.sub.D, e.g. the first drive voltage, and on the other side to a second voltage potential V.sub.S, e.g. a second drive voltage. The drive switches S41, S42 are connected at a centre tap of the half-bridge circuit and are controlled such that the first end of the drive coil L61, which is connected to the centre tap, is alternately connected to the first voltage potential V.sub.D and to the second voltage potential V.sub.S.

(20) Further, as described above, in preferred embodiments, the PXM-signal s.sub.PXM can be generated online and forwarded to the converter 4. In FIG. 1b a selector switch S2 is provided, which is controlled by the control unit 9 by means of a control signal ctrl. The selector switch S2 can be set to receive a PXM-signal s.sub.PXM-STORED provided by the memory module 23 or to receive a PXM-signal s.sub.PXM-ONLINE provided online by a processor unit 25, e.g. a digital signal processer DSP, which is controlled by the control unit 9 by means of a control signal ctrl. In the processor unit 25 a program is implemented, with which suitable pulse-width modulated signals and/or pulse-density modulated signals can be generated. The processor unit 25, which, together with the other circuitry, is preferably integrated into the control unit 9, may also generate PXM-signals that are stored in the memory unit 23 for later use.

(21) In all embodiments discussed, the PXM-signal s.sub.PXM may be selected from the memory unit 23 and/or from the processor unit 25 with any configuration of drive switches S41, . . . , S44 present and with any configuration of the admittance unit 5, if present. Hence, the features of the individual embodiments can freely be combined. In particular, processor unit 25 can most advantageously be used to generate a PXM-signal s.sub.PXM online, with any set of operating frequencies. At the same time the admittance unit 5 may automatically be tuned to the same set of operating frequencies. The processor unit 25 may replace the memory unit 23 in all disclosed circuits or may be used as an alternative source for the PXM-signal s.sub.PXM.

(22) FIG. 2 shows the metal detector 1 of FIG. 1a equipped with a drive controller 2, which provides a first PXM-signal s.sub.PXM1 used for controlling drive switches S41, S42 and a corresponding second PXM-signal s.sub.PXM2 used for controlling drive switches S43, S44. The first PXM-signal s.sub.PXM1 is stored in memory module 23A and the second PXM-signal s.sub.PXM2 stored in memory module 23B preferably at corresponding addresses. The address counter 22 can therefore synchronously address both memory modules 23A and 23B in order to simultaneously read out the first PXM-signal s.sub.PXM1 and the second PXM-signal s.sub.PXM2. The memory modules 23A and 23B may store PXM-signals s.sub.PXM1 and s.sub.PXM2 that are inverted to one another. Having two PXM-signals makes it possible for the converter 4 to have 0 volts differential on its outputs when both PXM-signals are at ground potential or both are at the potential of the drive voltage V.sub.D. This allows the generation of trapezoidal waves and better current control.

(23) FIG. 3 shows the metal detector 1 of FIG. 1a equipped with a drive controller 2 that allows selecting one of a plurality of stored PXM-signals s.sub.PXM, each provided with a specific set of frequency components, and with an admittance unit 5 that allows selectively connecting at least one of a plurality of branches, which preferably comprise each at least one capacitor C51; C52; C5n and at least one inductor L51; L52; L5n, to the drive coil L61 in order to create resonant circuits tuned to the frequency components of a selected PXM-signal s.sub.PXM. Data of each PXM-signal s.sub.PXM are stored individually in a corresponding memory module 231; 232; 23n. The branches of the admittance unit 5 can individually be activated by means of switches S51, S52, S5n which are actuated by means of a selector 50.

(24) In order to select a specific PXM-signal s.sub.PXM with a desired set of operating frequencies ω1, ω2, ω3, ω4 and to select the corresponding resonant circuits or branches C51, L51; C52, L52; C5n, L5n in the admittance unit 5, the control unit 9 provides a frequency select signal sf for example to the address counter 22, optionally to the memory unit 23, and to the selector 50. The address counter 22 will then address the selected memory module 231, 232 or 23n and the selector 50 the corresponding switches S51, S52, S5n.

(25) The metal detector 1 of FIG. 3 can therefore selectively be tuned to any set of frequencies selected for a specific product and potential contaminants. Resonance circuits can be tuned by adding capacitors and inductors e.g. by means of switches, such as electronic switches. Values of these items may also be changed electronically.

(26) FIG. 4 shows a mathematically determined waveform of the drive current i(ωt):
i(ωt)=I sin(ωt)+I/3 sin(3ωt)+I/7 sin(7ωt)+I/17 sin(17ωt)
that includes four frequency components ω, 3ω, 7ω and 17ω but no disturbances. Further shown is a related PXM-signal s.sub.PXM that has been determined by approximation of a triangular signal i.sub.DA to the determined waveform of the mathematically determined drive current i(ωt). Since the drive current i.sub.D and the coil current i.sub.L61 have only two possible gradients for FIGS. 1 and 3 and three possible gradients including zero for FIG. 2, the mathematically determined waveform of the drive current i(ωt) is resembled or approximated by using triangular or trapezoidal segments. The mathematically determined waveform of the drive current i(ωt) is shown in a dashed line. The waveform of the approximated triangular signal i.sub.DA closely follows the waveform of the mathematically determined drive current i(ωt). In the first half of the period or the positive half wave, the maxima of the triangular signal i.sub.DA are set to the maxima of the mathematically determined drive current i(ωt). In the second half of the period or the negative half wave, the minima of the triangular signal i.sub.DA are set to the minima of the mathematically determined drive current i(ωt). The approximated triangular signal i.sub.DA is not the actual drive current i.sub.D but ideally its mirror image. The approximated triangular signal i.sub.DA is converted to a PXM-signal s.sub.PXM, which is then converted in the converter 4 to the actual drive current i.sub.D, the mirror image of the approximated triangular signal i.sub.DA. In FIG. 4 with brackets it is indicated that the actual drive current i.sub.D at least approximately also corresponds to the approximated triangular signal i.sub.DA. If however, if higher frequencies are suppressed, then the virtual drive current i.sub.D will rather resemble the mathematically determined drive current i(ωt).

(27) Approximation by triangular or trapezoidal segments has the advantage that unwanted signals occur remote from the selected frequency components ω1, ω2, ω3, ω4 and therefore have no significant impact on measurement. Further the typical location of such disturbing signals in the Fourier spectrum is known, wherefore such disturbing signals can easily be suppressed in the input stage 7 of the receiver unit 11 by filter means selected accordingly. U.S. Pat. No. 8,473,235 mentioned above discloses a circuit with filter stages located subsequent to the phase detectors. In the present invention, filtering efforts are smaller. However any known filtering technique can also be applied to the signal delivered by the detection coils L62, L63 before or after demodulation, i.e. before and/or after the phase detector 8.

(28) By means of the determined triangular signal i.sub.DA the switching angles α1, α2, . . . of the PXM-signal s.sub.PXM can be determined, which is required to control the drive switches S41, S42, S43, S44 in the converter 4. These switching angles α1, α2, . . . are positioned at the relative maxima and minima of the determined triangular signal i.sub.DA. Thereby the falling edges of the PXM-signal s.sub.PXM are set to occur at the maxima of the determined triangular signal i.sub.DA and the rising edges of the PXM-signal s.sub.PXM are set to occur at the minima of the determined triangular signal i.sub.DA. The obtained PXM-signal s.sub.PXM or PXM-signals s.sub.PXM1, s.sub.PXM2, . . . is/are then stored in the memory unit 23, i.e. in one of the memory modules 231, 232, 23n; 23A, 23B.

(29) FIG. 4 shows the mathematically determined waveform of the drive current i(ωt), the approximated triangular signal i.sub.DA and the determined PXM-signal s.sub.PXM for the length of one period of the lowest frequency ω. Repetitively reading out the data of the PXM-signal s.sub.PXM from the related memory module 231, 232, 23n; 23A, 23B allows therefore to establish a continuous stream of the PXM-signal s.sub.PXM.

(30) FIG. 5 shows the coil current i.sub.L61 in the drive coil L61, the currents i.sub.L51, i.sub.L52 in the branches of the admittance unit 5 (see FIG. 3) and the drive current i.sub.D delivered by the drive switches S41, S42, S43, S44 in the converter unit 4 for the complete frequency spectrum of the metal detectors of FIG. 1a, FIG. 1b, FIG. 2 and FIG. 3. While the gradient of the coil current i.sub.L61 extends almost linear, the curve of the drive current i.sub.D shows a strong decay at each resonant frequency frequencies f.sub.RES1, f.sub.RES2 so that the drive current i.sub.D is well below the coil current i.sub.L61 at these positions of the spectrum. Hence, with comparably small drive currents i.sub.D with drive frequencies f.sub.D1, f.sub.D2 set to the resonant frequencies f.sub.RES1, f.sub.RES2 of the admittance unit 5 and drive coil L61, or vice versa, large coil currents i.sub.L61 can be reached. An explanation for this advantageous effect can be given with regard to the currents i.sub.L51, i.sub.L52 appearing in the branches of the admittance unit 5 which currents i.sub.L51, i.sub.L52, at the resonant frequencies f.sub.RES1, f.sub.RES2 of the admittance unit 5, are equivalent to the coil current i.sub.L61 in the drive coil L61. The admittance unit 5 contains in its branches passive components, e.g. inductors L51; L52, L5n and capacitors C51, C52, C5n, which when in resonance together with the drive coil L61 cause the current to circulate between the drive coil L61 and the branches of the admittance unit 5. Advantageously, the power circulation at the predetermined drive frequencies f.sub.D1, f.sub.D2 set to the resonant frequencies f.sub.RES1, f.sub.RES2 is restricted within the loop formed by the admittance unit 5 and the drive coil L61 making the driving-point admittance to be zero on an ideal lossless system. Consequently, the drive currents i.sub.D flowing at the drive frequencies f.sub.D1, f.sub.D2 through the drive switches S41, S42, S43, S44, typically MOSFETs, will be much lower than the coil currents i.sub.L61 flowing through the drive coil L61 and the branches of the admittance unit. Among other advantages, enables the extension of the spectrum towards lower frequencies and the possibility of driving a low impedance drive coil L61.

(31) In the drawings, preferred embodiments of the admittance unit 5 are shown. However, any other circuitry that preferably selectively allows reaching resonant circuits operating at defined frequencies f.sub.RES1, f.sub.RES2 are of course also applicable.