Method for operating a metal detector and metal detector

0001The current invention relates to a method for operating a metal detector that uses one or more operating frequencies and to a metal detector operating according to this method.

0002An industrial metal detector as described for example in <patcit id="pcit0001" dnum="US8587301B2"><text>US8587301B2</text></patcit>, is used for detecting metal contaminants in a product. When properly installed and operated, it will help reducing metal contamination and improving food safety. Most modern metal detectors utilise a search head comprising a "balanced coil system". Detectors of this design are capable of detecting all metal contaminant types including ferrous, nonferrous and stainless steels in a large variety of products such as fresh and frozen products.

0003A metal detector that operates according to the "balanced coil"-principle typically comprises three coils, a drive coil and two identical detection coils that are wound onto a non-metallic frame, each exactly parallel with the other. Since the detection coils, which typically enclose the drive coil centred in between, are identical, theoretically identical voltages are induced in the drive coils. In order to receive an output signal that is zero when the system is in balance, the first detection coil is connected in series with the second detection coil having an inversed sense of winding. Hence, under ideal conditions, the voltages induced in the detection coils, that are of identical amplitude and inverse polarity are cancelling out one another in the event that the system is in balance and no contaminant is present in an observed product.

0004However, as soon as a particle of metal passes through the coil arrangement, the electromagnetic field is disturbed first near one detection coil and then near the other detection coil. While the particle of metal is conveyed through the detection coils, the voltage induced in each detection coil is changed (typically by nano-volts). This change in balance results in a signal at the output of the detection coils that, in a receiver unit, is processed, amplified and subsequently used to detect the presence of the metal contaminant in the observed product.

0005In the receiver unit the input signal is normally split into an in-phase component and a quadrature component. The vector composed of these components has a magnitude and a phase angle, which is typical for the products and the contaminants that are conveyed through the coil system. In order to identify a metal contaminant, "product effects" need to be removed or reduced. If the phase of the product is known then the corresponding signal vector can be reduced so that a higher sensitivity results for the detection of signals originating from metal contaminants.

0006Methods applied for eliminating unwanted signals from the signal spectrum exploit the fact that metal contaminants, products and other disturbances have different influences on the magnetic field so that the detected signals differ in phase. Materials with a high conductivity cause signals with a higher resistive signal component and a smaller reactive signal component. Materials with a high magnetic permeability cause signals with a smaller resistive signal component and a higher reactive signal component. Signals caused by ferrite are primarily reactive, while signals caused by stainless steel are primarily resistive. Products, which are conductive, typically cause signals with a strong resistive component. The phase angle of the signal vectors between the resistive signal component and reactive signal component remains normally constant, when a product or contaminant is conveyed through the metal detector.

0007Above of it has been described, that under ideal conditions the balanced coil system does not provide an output signal, if no product or contaminant is present. However, as described in <patcit id="pcit0002" dnum="EP2812734B1"><text>EP2812734B1</text></patcit>, despite the fact that identical detection coils are set in a near perfectly balanced state at the factory site, there will still be occasions where the detection coil system will fail to be in a balanced state with no test items or contaminants present, resulting for example in perfectly acceptable food products being rejected. The balance of the detector can be disturbed due to mechanical impacts on the system, due to changing ambient conditions, due to metallic objects located in the vicinity of the detector or due to relaxation or aging of components. In view of the high sensitivity of metal detection systems and minute effects of contaminants on the output voltage of the coil system, imbalances may cause saturation of the receiver channels, particularly of the input amplifiers and the phase sensitive detectors, which only operate over a limited voltage signal range. In order to remove such imbalances an adjustable balance signal derived from the drive signal is combined with the output signal of the detector and varied until the combined output signal is compensated.

0008The output signal and the adjustable balance signal are individually filtered to remove one or more harmonics prior to being combined. Further, the adjustable balance signal may be adjusted for example by means of adjustable potentiometers. This compensation circuit, which must operate with highest accuracy, typically requires expensive low noise components. Still, the compensation circuit will not only counteract imbalances but will also introduce noise into the detection system, e.g. by the potentiometers. Further, the compensation circuitry applied so far is inflexible and limited to imbalances relating to the operating frequency of the drive signal. Other disturbances, which are not directly linked to the operating frequency, cannot be corrected. Still further, the adjustment process, particularly when using potentiometers, may require considerable time and may not completely cancel out the imbalance signal.

0009In <patcit id="pcit0003" dnum="US5691640A"><text>US5691640A</text></patcit> a compensation circuitry for a metal detector is disclosed, which also uses signals derived from the drive signal. These signals are applied to the phase sensitive detectors and to multipliers to which control signals are applied. The output signals of the multipliers are applied via feedback isolation transformers to the input of the receiver stage. Disadvantages of such a complex and expensive compensation circuitry are the same as described above. This circuitry also requires expensive low noise components and isolation transformers, which must be designed accurately so that no further imbalance and noise is introduced. Further, also compensation circuitry is inflexible and limited to imbalances relating to the operating frequency of the drive signal. Other disturbances, which are not directly linked to the operating frequency, cannot be corrected. Further, this circuitry may change its properties according to the selected operating frequencies. Hence, complexity of the circuitry and an additional imbalances possibly introduced will increase with a high number of operating frequencies, particularly if they are simultaneously applied.

0010The present invention is therefore based on the object of providing an improved method for operating a metal detector and to an improved metal detector operating according to this method.

0011The inventive method shall allow accurately eliminating imbalances with little effort that appear in any stage of the metal detector, particularly in the coil system, in the receiver channels, and that are caused by influences external or internal to these stages.

0012The inventive compensation unit shall be flexible and easily adaptable to correct not only imbalances but also disturbances that adversely affect the performance of the metal detector. In particular, it shall be possible to eliminate various disturbances which appear independently of the operating frequency.

0013The inventive metal detector shall avoid the requirement of costly components and a complex compensation circuitry.

0014Furthermore, the inventive compensation unit shall accurately remove imbalances but not introduce itself noise or disturbances into the system.

0015In a first broad aspect of the invention there is provided a method for operating a metal detector, which comprises a balanced coil system with a drive coil that is connected to a transmitter unit, which provides a transmitter signal with at least one operating frequency, and with a first and a second detection coil that provide an output signal to a receiver unit, that processes a related receiver signal, which typically comprises an imbalance signal, and that provides digital in-phase and quadrature components of the demodulated imbalance signal. The metal detector is designed for detecting contaminants, particularly metal contaminants, in products that are examined by the metal detector and for this purpose are transferred through the balanced coil system of the metal detector.

0016According to the invention the method comprises the steps of <ul id="ul0001" list-style="none" compact="compact"><li><img file="EP3726257A1_D0001.tif" /> processing the digital in-phase and digital quadrature components of the imbalance signal in a signal controller for providing control data to a compensation unit that is used for compensating the imbalance signal,</li><li><img file="EP3726257A1_D0002.tif" /> applying the digital in-phase components of the imbalance signal to a first control unit, which provides an in-phase control component for the imbalance signal and applying the digital quadrature components of the imbalance signal to a second control unit, which provides a quadrature control component for the imbalance signal;</li><li><img file="EP3726257A1_D0003.tif" /> in the compensation unit, synthesizing a digital compensation signal with the frequency of the imbalance signal, which corresponds to the at least one operating frequency, and with a phase and magnitude according to the in-phase and quadrature control components provided for the imbalance signal;</li><li><img file="EP3726257A1_D0004.tif" /> converting the digital compensation signal into an analogue compensation signal, and</li><li><img file="EP3726257A1_D0005.tif" /> applying the analogue compensation signal to the balanced coil system or to the receiver signal for compensating the imbalance signal.</li></ul>

0017According to the inventive method the in-phase components and the quadrature components of the imbalance signal are detected and related in-phase and quadrature control components are determined or calculated in control units. Based on these in-phase and quadrature control components a compensation signal is synthesized with a frequency, phase and magnitude suitable for compensating the imbalance signal. Hence, at least one control loop is formed, with negative feedback of control signals to decrease the size and finally cancel out any imbalance signal. Ideally a compensation signal would be synthesized identical to the imbalance signal and would be applied 180° out of phase with reference to the imbalance signal into the signal path. In reality, preferably a PID controller is used to fully compensate the imbalance signal with control terms that have proportional, integral and derivative influence on the provided digital in-phase control components and the digital quadrature control components supplied to the compensation unit.

0018Hence, according to the invention digital in-phase and digital quadrature components of the imbalance signal are processed to deliver digital in-phase control components and the digital quadrature control components. The imbalance signal and its components can be gained in different ways. The imbalance signal can be observed when the metal detector is operative preferably without any products being transferred through the balanced coil system. The transfer of products may be delayed until the synthesized compensation signal is optimized. However, if the product signals are known or are lying in a different frequency range, then the products can be measured and the product signals may be subtracted or separated from the receiver signal, in order to obtain the imbalance signal. Hence, in the first control mode or calibration mode the imbalance signal is measured and a corresponding compensation signal is synthesized, with which the imbalance signal is fully or at least partially eliminated.

0019After the initial calibration of the metal detector has been performed in the first control mode or calibration mode the metal detector can be switched to a normal operation mode, in which products are examined and the imbalance signal is eliminated by means of the synthesized compensation signal, which has been stored. In this operation mode changes of the imbalance signal are neglected.

0020In a preferred embodiment, the metal detector is switched to a second control mode after the first control mode has been terminated. In the second control mode products are examined and the remaining imbalance signal is further observed and further eliminated. For this purpose the imbalance signal may be separated from the receiver signal for example by a filter stage, which allows the imbalance signal to pass through.

0021In the normal operation mode or in the second control mode the imbalance signal is preferably monitored and the metal detector is returned to the first control mode if the imbalance signal exceeds a predetermined threshold. In the first control mode the transfer of products through the balanced coil system is preferably interrupted until the synthesized compensation signal is optimized again. This optimization may typically be performed within a few seconds. Afterwards the metal detector is again switched to normal operation mode or to the second control mode.

0022The receiver signal may be demodulated in the analogue domain or in the digital domain to regain the baseband signals of the products, contaminants and the imbalance signal or any disturbances.

0023Demodulation of a signal in the digital domain is described for example in <u>https://www.nutaq.com/blog/digital-iq-demodulator-high-speed-adc</u>. In a digital I/Q demodulator a receiver signal, centred at an operating frequency, is sampled directly by a single high-speed ADC device. The sampled receiver signal is mixed in a digital in-phase mixer and in a digital quadrature mixer with digital cosine and sine samples of the operating frequency generated by a direct-digital synthesizer (DDS) to translate the receiver signal to in-phase and quadrature signals in the baseband. After filtering out the high-frequency component of the outputs of the in-phase mixer and the digital quadrature mixer the baseband signal is obtained. The use of processing the receiver signal in a high-speed ADC and digital I/Q demodulator typically may reduce potential errors such as gain imbalances and DC offsets when the translation from the operating frequency to baseband is performed in analogue domain.

0024Translation of the receiver signal to baseband can also be performed in the analogue domain. In this embodiment the analogue receiver signal is compared in a first phase detector with an analogue in-phase reference signal and in a second phase detector with an analogue quadrature reference signal in order to produce analogue in-phase components and analogue quadrature components of a baseband signal that are converted into digital in-phase and quadrature components.

0025Hence, in a first embodiment of the inventive metal detector or the implemented inventive method the receiver signal is demodulated in the analogue domain. In a second embodiment of the inventive metal detector or the implemented method the receiver signal is demodulated in the digital domain.

0026The signal controller is preferably operated in two or more modes, in which different sets of control terms are applied to the control units. After initialization of the metal detector a first control mode is preferably set and maintained until a predetermined condition is met, e.g. for a defined period of time or as long as the imbalance signal exceeds a given threshold. As mentioned above, during this first control mode or calibration mode, preferably no products are examined. The first control mode is maintained for example for period of time, in which the imbalance signal is typically reduced to a level, at which the imbalance signal has no significant influence on the measurement. Alternatively, the level, at which the imbalance signal has no significant influence on the measurement, is predetermined and compared with the actual imbalance signal. As soon as the predetermined condition is met, the first control mode is changed to a second control mode, in which measurement of products is performed while the imbalance signal is still observed and reduced in a control loop. In the first control mode the imbalance signal is quickly reduced, while in the second control mode a remaining imbalance signal is slowly driven to a minimum value or cancelled out completely. The control terms for the second control mode are selected such that signal corrections in the control loop do not lead to disturbances during measurement of the products. In particular, control terms for the control units are selected such that overshoots are avoided. In the second control mode operating conditions are therefore continuously optimized, while measurements can be performed.

0027The inventive method allows accurately eliminating imbalances with little effort that occur in any stage of the metal detector, particularly in the coil system, in the receiver channel, and that are caused by influences external or internal to these stages. The inventive compensation unit can flexibly and easily be adapted to correct any imbalances and further disturbances that adversely affect the performance of the metal detector. In particular, disturbances can be eliminated, which do not directly correspond to the operating frequency.

0028For any disturbance signal the signal controller may provide control signals and the compensation unit may generate or synthesize corresponding correction signals. The signal controller will provide information with regards to the frequency, magnitude and phase of the disturbance signal for which a correction signal will be synthesized in the compensation unit. The compensation unit may generate or synthesize correction signals in the baseband of the measured signals or offset in frequency thereto. In the event that for example frequency sidebands are detected, such frequency sidebands may be eliminated by correction signals which are synthesized accordingly.

0029In a preferred embodiment, for eliminating any disturbance signal <ul id="ul0002" list-style="none"><li><img file="EP3726257A1_D0006.tif" /> in the signal controller, a digital in-phase component and a digital quadrature component and the frequency of a disturbance signal are determined; and</li><li><img file="EP3726257A1_D0007.tif" /> a digital in-phase control component and a digital quadrature control component are determined or calculated; and</li><li><img file="EP3726257A1_D0008.tif" /> in the compensation unit a digital correction signal is synthesized in accordance with the digital in-phase control component and the digital quadrature control component of the disturbance signal; and</li><li><img file="EP3726257A1_D0009.tif" /> finally, the digital correction signal is added to the digital compensation signal.</li></ul>

0030The inventive compensation unit shall avoid the requirement of costly components and a complex compensation circuitry. The compensation unit preferably comprises a field programmable gate array unit (FPGA), which can easily be adapted to future requirements. In addition or alternatively, the compensation unit may comprise one or more frequency synthesizers, preferably direct digital synthesizers as described in <u>https://en.wikipedia.org/wiki/Direct_digital_synthesis</u>, with which the required frequencies for one or more compensation signals can be generated. Alternatively, signal data of waveforms of the at least one operating frequency or in-phase frequency components and quadrature frequency components thereof are sequentially read out from a memory unit or a look-up table. Data of one or more periods of a frequency signal can be stored with a desired resolution in the memory unit or look-up table and can be sequentially read out with a system clock to generate the related frequency. With the system clock the address of the data, which is a read out from memory, may be increased step-by-step. Hence, with a constant system clock and numerous waveforms stored in memory any related frequency can be generated with the desired resolution. However it would those be possible to change the frequency by selecting a different system clock.

0031A particular advantage of the inventive control loop is that any imbalance signal and any disturbance signal can be eliminated while no disturbing signal or noise is introduced into the metal detection system.

0032The inventive method can be used for any type of metal detector, for example for a metal detector that uses only one operating frequency or that uses multiple frequencies which are individually selectable or simultaneously applicable. For each imbalance signal, which occurs for a related operating frequency, a compensation signal is individually synthesized in a corresponding control channel.

0033Preferably, in the receiver unit or in the signal controller, the imbalance signal is separated by at least one low pass filter from other signal components. More preferably, the digital in-phase component of the imbalance signal is applied via a first low-pass filter to the first control unit and the digital quadrature component of the imbalance signal is applied via a second low-pass filter to the second control unit. Demodulating the imbalance signal typically results in DC- components. However, signals originating from products or contaminants, which are traveling through the metal detector, cause a low frequency modulation of the operating frequency. When demodulating the receiver signal, for example with reference frequencies in phase detectors, this low frequency is regained. By the low-pass filter with a corresponding cut-off frequency (e.g. &lt; 0.5 Hz), the signals of the product and contaminants and the imbalance signal can therefore be separated from one another. The influence of signals of products and contaminants in the control channel can therefore be excluded. Preferably, separation of the signals or filtering is performed by a digital signal processor.

0034Based on the digital in-phase control component and the digital quadrature control component the digital compensation signal can be synthesized or generated in different ways. In a preferred embodiment <ul id="ul0003" list-style="none"><li><img file="EP3726257A1_D0010.tif" /> the digital in-phase control component and the digital quadrature control component are converted into a phase control signal and into a magnitude control signal;</li><li><img file="EP3726257A1_D0011.tif" /> the phase control signal and the magnitude control signal are forwarded to the compensation unit;</li><li><img file="EP3726257A1_D0012.tif" /> a synthesizer is controlled to generate the operating frequency with a phase in accordance with the phase control signal that is applied to the synthesizer; and</li><li><img file="EP3726257A1_D0013.tif" /> in a multiplier the magnitude of the generated operating frequency is adjusted in accordance with the provided magnitude control signal for providing the digital compensation signal.</li></ul>

0035Conversion of the digital in-phase control component and the digital quadrature control component into a phase control signal and into a magnitude control signal can be done by the CORDIC algorithm (<u>https://en.wikipedia.org/wiki/CORDIC</u>) e.g. in a microcontroller or digital signal processor or FPGA provided in the signal controller or compensation unit.

0036In a further preferred embodiment <ul id="ul0004" list-style="none"><li><img file="EP3726257A1_D0014.tif" /> the digital in-phase control component and the digital quadrature control component are provided to the compensation unit;</li><li><img file="EP3726257A1_D0015.tif" /> a synthesizer is controlled to generate an in-phase frequency component and a quadrature frequency component corresponding to the operating frequency, i.e. the frequency of the imbalance signal;</li><li><img file="EP3726257A1_D0016.tif" /> in an in-phase multiplier the magnitude of the generated in-phase frequency component is adjusted in accordance with the digital in-phase control component for providing a digital in-phase compensation component of the digital compensation signal and in a quadrature multiplier the magnitude of the quadrature frequency component is adjusted in accordance with the digital quadrature control component for providing a digital quadrature compensation component of the digital compensation signal; and</li><li><img file="EP3726257A1_D0017.tif" /> the digital in-phase compensation component of the digital compensation signal and the digital quadrature compensation component of the digital compensation signal are combined for providing the digital compensation signal.</li></ul>

0037The digital compensation signals, which have been synthesized for each operating frequency, i.e. for each imbalance signal, are combined for providing a combined digital compensation signal, which is converted to the analogue compensation signal.

0038As described, the compensation signal is fully synthesized in the compensation unit without using signals generated by the transmitter unit. However, both units, the transmitter unit and the compensation unit, preferably use a common master clock so that they are synchronized. By establishing phase coherence, in the system disturbing phase errors, phase drifts and phase noise and the like are avoided.

0039The compensation signal generated in the compensation unit can be applied at every point in the metal detector, where the imbalance signal occurs. The compensation signal can be induced into the coil system preferably by means of a compensation coil that is inductively coupled with the detection coils. Alternatively, the imbalance signal can be applied to the input of the receiver unit, e.g. to a compensation coil provided on a balanced transformer, or to the signal path of the receiver unit, preferably via a subtraction module.

0040Detailed aspects and examples of the invention are described below with reference to the drawings, wherein <dl id="dl0001" compact="compact"><dt>Fig. 1</dt><dd>shows a preferred embodiment of an inventive metal detector that comprises a transmitter unit 1, a coil system 2 with a drive coil 21 and two detection coils 22, 23, a receiver unit 3, a control unit 9, and a signal controller 4 that provides data relating to an imbalance signal to a compensation unit 5, which synthesizes a compensation signal a_COMP that is introduced into the coil system 2 or into the signal path of the receiver unit 3;</dd><dt>Fig. 2</dt><dd>shows the metal detector of <figref idref="f0001">Fig. 1</figref> with the detection coils 22, 23 of the coil system 2 connected to a balanced transformer 31, which is located at the input of the receiver unit 3, and which comprises a balanced compensation coil 30 to which the compensation signal a_COMP is applied;</dd><dt>Fig. 3</dt><dd>shows the signal controller 4 and the compensation unit 5 provided for two control channels in a preferred embodiment;</dd><dt>Fig. 4</dt><dd>shows the signal controller 4 and the compensation unit 5 provided for two control channels in another a preferred embodiment;</dd><dt>Fig. 5</dt><dd>shows a flow diagram with different operating modes, in which compensation signals a_COMP are generated.</dd></dl>

0041<figref idref="f0001">Fig. 1</figref> shows a block diagram of an inventive metal detector in a preferred embodiment, which comprises a transmitter unit 1, a balanced coil system 2 with a drive coil 21 and a first and a second detection coil 22, 23, a receiver unit 3, a signal controller 4, a compensation unit 5 and a control unit 9, e.g. a personal computer that preferably comprises signal processing circuitry, such as at least one signal processor, designed for processing signals d_I, d_Q relating to products and contaminants provided by the receiver unit 3. The signal controller 4 and the compensation unit 5, which are used for compensating imbalances in the metal detector, are preferably also implemented in in the control unit 9. The signal controller 4 is preferably implemented in a digital signal processor and the compensation unit 5 is preferably implemented in a field programmable gate array (FPGA), which receives data from the signal controller 4.

0042The transmitter unit 1 provides a transmitter signal s1 with at least one operating frequency f_TX to the drive coil 21 of the balanced coil system 2 and a reference signal S_R with the operating frequency f_TX to the receiver unit 3. The transmitter signal s1 induces signals in the identical detection coils 22, 23 that are of the same amplitude but inverse polarity as long as the system is in balance, i.e. as long as no products or contaminants are conveyed through the metal detector.

0043The operating frequency f_TX may be fixed or individually selectable from a plurality of frequencies f_TX-A, f_TX-B. Alternatively, a plurality of frequencies f_TX-A, f_TX-B may simultaneously be generated in the transmitter unit 1, which is provided with at least one frequency synthesizer or frequency generator 10.

0044As soon as a product or contaminant passes through the balanced coil system 2, the signals induced in the detection coils 22, 23 will change. As a result the operating frequency f_TX induced in the detection coils 22, 23 gets modulated with a base band signal, whose amplitude and frequency are dependent on the electric and magnetic property, dimension and travelling speed of the object. Depending on the travelling speed the frequency of the baseband signal will change.

0045Imbalance signals cause a constant imbalance or constant signal offset which may disturb and saturate the receiver channels. Such imbalance signals may be caused by mechanical impacts on the system, by changing ambient conditions, by metallic objects located in the vicinity of the detected or due to relaxation or aging of components.

0046In the preferred embodiment shown in <figref idref="f0001">Fig. 1</figref> the output signals of the detection coils 22, 23 are applied to an input stage 32 of the receiver unit 3. The input stage 32 typically comprises at least one amplifier and at least one filter unit. The input stage 32 delivers a receiver signal s3 via an optional subtraction unit 33 to an in-phase-channel I-CH and to a quadrature channel Q-CH. The in-phase-channel I-CH and quadrature channel Q-CH, which are identically designed, comprise each a phase detector 34I; 34Q connected to a filter stage 35I; 35Q, followed by a gain stage 36I; 36Q and an analogue to digital converter 37I, 37Q.

0047The phase detectors 34I; 34Q, which receive at their inputs the receiver signal s3 and the in-phase reference signal s_RI or the quadrature reference signal s_RQ provided by the transmitter unit 1, act as demodulators and provide at their outputs the in-phase components a_I or quadrature components a_Q of the baseband signals, which originate from the conveyed products and contaminants or imbalances. The in-phase reference signal s_R is routed from the transmitter unit 1 directly to the first phase detectors 34I and via a phase shifter 38 to the second phase detectors 34Q.

0048The analogue in-phase and quadrature signals a_I ; a_Q provided at the outputs of the phase detectors 34I; 34Q are forwarded via the filter units 35I; 35Q to the gain units 36I; 36Q, which preferably allow setting the amplitudes of the processed signals to a desired value. Subsequently the filtered and calibrated signals are converted in an analogue to digital converter 37I; 37Q from analogue in-phase and quadrature signals to digital in-phase and quadrature signals d_I, d_Q. The digital in-phase and quadrature signals d_I, d_Q provided at the output of the analogue to digital converter 37I; 37Q are forwarded to the control unit 9, where the digital in-phase and quadrature signals d_I, d_Q are analysed and an alarm is raised if a signal is detected that relates to a contaminant.

0049The control unit 9 typically comprises a processing unit and an operating program 90 designed for fulfilling the functions of an inventive single frequency or multi-simultaneous frequency metal detector. The operating program 90 is enhanced for implementing the inventive method. The signal controller 4 and the compensation unit 5 are preferably also implemented in the signal processing or signal generating circuitry contained in the control unit 9.

0050The digital in-phase and quadrature signals d_I, d_Q provided at the output of the analogue to digital converter 37I; 37Q are forwarded further to the signal controller 4. In the signal controller 4 the digital in-phase signal d_I, which comprises the in-phase component of the imbalance signal, is applied to a first control unit 41I, which provides an in-phase control component d_CI, for the imbalance signal. The quadrature signal d_Q, which comprises the quadrature component of the imbalance signal, is applied to a second control unit 41Q, which provides a quadrature control component dc_Q for the imbalance signal.

0051In the first control unit 41I and in the second control unit 41Q preferably identical control terms Kp, Ki, Kd are applied, which have proportional and/or integral and/or derivative influence on the provided digital in-phase control components d_CI, and the digital quadrature control components d_CQ. Hence, preferably a PID-controller is implemented, which is well known in the art. A PID-controller continuously calculates an error value as the difference between a desired set-point and a measured process variable and applies a correction based on proportional, integral, and derivative terms (denoted P, I, and D respectively). These values can be interpreted in terms of time: P depends on the present error, I on the accumulation of past errors, and D is a prediction of future errors, based on the current rate of change. The weighted sum of these three actions is used to adjust the process via a control element. In the inventive metal detector the measured process variables are the digital in-phase components d_I and the digital quadrature components d_Q of the imbalance signal. The set-point is zero targeting the complete elimination of the imbalance signal and the correction values are the calculated in-phase control component d_CI, and the quadrature control component d_CQ and the analogue compensation signal a_COMP derived therefrom. The control element can be regarded as the point of injection where the compensation signal a_COMP is applied.

0052The in-phase control component d_CI, and the quadrature control component d_CQ are forwarded to a converter 42, which preferably uses a CORDIC-algorithm for converting the digital in-phase control component d_CI, and the digital quadrature control component d_CQ into a phase control signal C_PH and into a magnitude control signal C_M. <figref idref="f0001">Fig. 1</figref> shows the control vector dc with the magnitude and the phase angle α, corresponding to the digital in-phase control component d_CI, and the digital quadrature control component d_CQ. The magnitude and the phase angle α of the control vector d_C corresponding to the imbalance signal are typically constant. Signals relating to products and contaminants would have a constant phase angle, but a magnitude, which changes with a frequency that corresponds to the velocity with which the products and contaminants are conveyed through the metal detector.

0053The phase control signal C_PH-A and the magnitude control signal C_M-A, which are forwarded to the compensation unit 5, together with a frequency control signal f_C-A, comprise the essential information for synthesizing a digital compensation signal d_COMP, which corresponds to the control vector d_C. In the present example the compensation signal is generated or synthesized for a first operating frequency f_TX-A. A corresponding frequency control signal f_C-A is applied together with the phase control signal C_PH-A to a frequency generator 51 in the compensation unit 5. The frequency generator 51 therefore generates a frequency signal f_G-Aα with a frequency corresponding to the frequency control signal f_C-A and a phase angle corresponding to the phase control signal C_PH-A. The generated frequency signal f_G-Aα is supplied to a multiplier 52 in which the magnitude of the generated frequency signal f_G is adjusted in accordance with the provided magnitude control signal c_M for providing the digital compensation signal d_COMP.

0054The generated or synthesized digital compensation signal d_COMP is forwarded to a digital to analogue converter 54, which provides a corresponding analogue compensation signal a_COMP that can be applied at various points within the metal detector. The analogue compensation signal a_COMP is applied for example to the balanced coil system or to a receiver signal s2, s3 at the input of the receiver unit 3 or within the signal path of the receiver unit 3. Typically in all embodiments, the analogue compensation signal a_COMP is amplified in an amplifier 6, such as a buffer amplifier, before it is applied.

0055In a first preferred embodiment, the analogue compensation signal a_COMP is applied to a compensation coil 24, which is inductively coupled with at least one of the drive coils. In a second embodiment, the analogue compensation signal a_COMP is applied to a subtraction unit 33 in which the imbalance signal present in the system is reduced by the analogue compensation signal a_COMP preferably to zero.

0056For any other disturbance signals occurring in the metal detector the digital in-phase components and a digital quadrature components and the frequency can be determined and digital in-phase control components and a digital quadrature control components can be calculated. In the compensation unit 5, digital correction signals d_CORR in accordance with the digital in-phase component and the digital quadrature component and the frequency of the disturbance signal can be synthesized and added to the digital compensation signal d_COMP (see <figref idref="f0003">Fig. 3</figref> and <figref idref="f0004">Fig. 4</figref>).

0057Hence, with the inventive method any detected disturbances can be eliminated by analysing the disturbance signals and synthesizing a corresponding correction signal. The operating conditions and operating parameters of the metal detector are selected by means of a control program 90. The control unit 9 preferably communicates via a first control bus 91 with the transmitter unit 1 and via a second control bus 945 with the signal controller 4 and with the compensation unit 5. In preferred embodiments all processes executed in the metal detector, the conventional measurement processes and all processes relating to the inventive method are preferably controlled by the control unit 9.

0058In the event that for example two operating frequencies f_TX-A, F_TX-B are selected in the control unit 9, then the transmitter unit 1 is instructed to generate these two operating frequencies f_TX-A, f_TX-B. Further, two control channels 4A, 4B; 5A, 5B are opened for compensating imbalance signals related to these two operating frequencies f_TX-A, f_TX-B. The first control channel, which comprises a first receiver module as shown in <figref idref="f0001">Fig. 1</figref>, a first signal controller 4A and a first compensation unit 5A, is part of a first control loop in which imbalance signal for the first operating frequency F_TX-A are eliminated. The second control channel, which comprises a second receiver module as shown in <figref idref="f0001">Fig. 1</figref>, a second signal controller 4B and a second compensation unit 5B, is part of a second control loop in which imbalance signal for the second operating frequency f_TX-B are eliminated. In the receiver modules, in phase and quadrature channels I-CH, Q-CH are provided as shown for example in <figref idref="f0001">Fig. 1</figref>. In this way, every signal controller 4A, 4B is supplied with demodulated digital in-phase components d_I and digital quadrature components d_Q of the imbalance signal for the related operating frequency f_TX-A, f_TX-B.

0059In order to ensure phase coherence for all processes executed in the metal detector the at least one generated frequency signal f_G-A, f_TXG-B or in-phase frequency components f_GI and a quadrature frequency components f_GQ thereof generated in the compensation unit 5 (see <figref idref="f0004">Fig. 4</figref>) are synchronized with regard to phase and frequency with the corresponding operating frequencies f_TX-A, f_TX-B generated in the transmitter unit 1. <figref idref="f0001">Fig. 1</figref> shows that a master clock mcl is applied to a frequency generator 10 located in the transmitter unit 1, to the analogue-to-digital converters 37I and 37Q in the receiver unit 3 and to the frequency generator 51, the multiplier 52 and the digital-to-analogue converter 54 in the compensation unit 5.

0060<figref idref="f0002">FIG. 2</figref> shows the metal detector of <figref idref="f0001">Fig. 1</figref> with the detection coils 22, 23 of the coil system 2 connected to a balanced transformer 31, which is located at the input of the receiver unit 3. The output signals of the detection coils 22, 23 are applied to center-tapped primary windings of the balanced transformer 31 that mirror the detection coils 22, 23. Further, the balanced transformer 31 comprises two identical center-tapped secondary windings whose opposite tails are connected to the input stage 32 of the receiver unit 3. The balanced transformer 31 further comprises a balanced compensation coil 30 to which the compensation signal a_COMP is applied via amplifier 6 and a single-ended-to-differential-converter 60 (below: SETD-converter). The SETD converter 60 provides a pair of differential signals at its output.

0061<figref idref="f0003">Fig. 3</figref> shows the signal controller 4 and the compensation unit 5 with two control channels dedicated to the selected operating frequencies f_TX-A, f_TX-B. Each control channel comprises a signal controller 4A; 4B and a compensation unit 5A; 5B respectively. The signal controllers 4A; 4B comprise low pass filters 40I; 40Q for separating the imbalance signal from other signals relating for example to products or contaminants. Signals relating to products and contaminants are therefore barred and will not influence the processes in the control loops. The digital in-phase component d_I of the imbalance signal is applied via a first low-pass filter 40I to the first control unit 41I and the digital quadrature component d_CQ of the imbalance signal is applied via a second low-pass filter 40Q to the second control unit 41Q.

0062As described with reference to <figref idref="f0001">Fig. 1</figref>, digital in-phase and quadrature signals d_I-A, d_Q-A; d_I-B, d_Q-B provided at the outputs of the related low pass filters 40I; 40Q are forwarded to the control units 41I, 41Q. In each control channel the digital in-phase signals d_I-A; d_I-B are applied to a first control unit 41I, which provides an in-phase control component d_CI-A; d_CI-B for the imbalance signal. The quadrature signal d_QA, d_QB is applied to a second control unit 41Q, which provides a quadrature control component d_CQ-A; d_CQ-B for the imbalance signal.

0063In each control channel, the in-phase control component d_CI-A; d_CI-B and the quadrature control component d_CQ-A; d_CQ-B are forwarded to a converter 42, which preferably uses a CORDIC-algorithm for converting the digital in-phase control component d_CI-A; d_CI-B and the digital quadrature control component d_CQ-A; d_CQ-B into a phase control signal C_PHα; C_PHβ and into a magnitude control signal C_M-A; C_M-B.

0064In each control channel, the phase control signal C_PHα; C_PHβ and the magnitude control signal C_M-A; C_M-B, which are forwarded to the compensation unit 5, together with a frequency control signal f_C-A; f_C-B, comprise the essential information for synthesizing a digital compensation signal d_COMP-A, d_COMP-B. The corresponding frequency control signal f_C-A; f_C-B is applied together with the phase control signal C_M-A; C_M-B to a related frequency generator 51. The frequency generator 51 generates a corresponding frequency signal f_G-Aα; f_G-Bβ with a frequency corresponding to the frequency control signal f_C-A; f_C-B and a phase angle α; β corresponding to the phase control signal C_M-A; C_M-B. The generated frequency signal f_G-Aα; f_G-Bβ is supplied to a multiplier 52 in which the magnitude of the generated frequency signal f_G-Aα; f_G-Bβ is adjusted in accordance with the provided magnitude control signal C_M-A; C_M-B for providing the digital compensation signal d_COMP-A, d_COMP-B.

0065The digital compensation signals d_COMP-A, d_COMP-B. are combined in an adder 53, which preferably also combines correction signals d_CORR as described above. At the output of the adder 53 the combined digital compensation signal d_COMP is provided, which is converted in the subsequent digital-to-analogue converter 54 to the analogue compensation signal a_COMP. The analogue compensation signal a_COMP is applied to the balanced coil system 2 or to a receiver signal s2, s3 at the input of the receiver unit 3 or within the signal path of the receiver unit 3 as described with reference to <figref idref="f0001">Fig. 1</figref> or <figref idref="f0002">Fig. 2</figref>.

0066<figref idref="f0004">Fig. 4</figref> shows the signal controller 4 and the compensation unit 5 in another a preferred embodiment. Each control channel comprises a signal controller 4A; 4B and a compensation unit 5A, 5B respectively. In each control channel, the in-phase control component d_CI-A; d_CI-B and the quadrature control component d_CQ-A; d_CQ-B are forwarded directly to the related compensation unit 5A, 5B. A frequency control signal f_C-A; f_C-B is applied to the related frequency synthesizer 51/51L in the compensation unit 5A, 5B, which generates an in-phase frequency component f_GI-A; f_GI-B and a quadrature frequency component F_GQ-A; F_GQ-B corresponding to the frequency control signal f_C-A or f_C-B.

0067In each control channel, the magnitude of the generated in-phase frequency component f_GI-A; f_GI-B is adjusted in an in-phase multiplier 52I in accordance with the digital in-phase control component d_CI-A; d_CI-B for providing a digital in-phase compensation component d_COMP-I-A; d_COMP-I-B. In a quadrature multiplier 52Q the magnitude of the quadrature frequency component f_GQ is adjusted in accordance with the digital quadrature control component d_CQ-A; d_CQ-B for providing a digital quadrature compensation component d_COMP-Q-A; d_COMP-Q-B. The digital in-phase compensation component d_COMP-I-A; d_COMP-I-B and the digital quadrature compensation components d_COMP-Q-A; d_COMP-Q-B are then combined in adder 53, which provides the combined digital compensation signal d_COMP, which is converted in the subsequent digital-to-analogue converter 54 to the analogue compensation signal a_COMP.

0068The at least one frequency signal f_G-A, f_G-B or in-phase frequency components f_G, and quadrature frequency components f_GQ thereof are generated by a direct digital synthesizing or by sequentially reading out data of the waveforms of the at least one frequency signal f_G-A, f_G-B or in-phase frequency components f_GI, and quadrature frequency components f_GQ thereof from a memory unit or look-up table 51L. <patcit id="pcit0004" dnum="US5892692A"><text>US5892692A</text></patcit> describes a digital waveform oscillator that generates digitized waveforms without distortion using a lookup table.

0069<patcit id="pcit0005" dnum="US2007067123A1"><text>US2007067123A1</text></patcit> discloses an arbitrary waveform generator, which can be used for generating any desired waveform in the compensation unit. The control channels may individually use a waveform generator or may share a waveform generator. It is describes that an arbitrary waveform generator can produce arbitrary signal waveforms that may be specific to their applications. These arbitrary signal waveforms may be utilized to simulate "real world" signals that include glitches, drift, noise and other anomalies on an arbitrary signal waveform that a component, such as a device under test, will encounter when it leaves the laboratory or manufacturing floor.

0070In preferred embodiments, the inventive method uses such a waveform generator to produce compensation signals d_COMP or correction signals d_CORR as described above. Typically a waveform generator or an arbitrary waveform generator in a comparably simple embodiment will be sufficient to compensate imbalance signals or disturbance signals, which recursively occur in a metal detector. Since signal waveforms are defined point-by-point, the resolution of the output signal can be increased by using more update points, which define the generated signal waveform.

0071In <figref idref="f0003">Fig. 3</figref> and <figref idref="f0004">Fig. 4</figref> the invention has been described for two control channels or for two operating frequencies f_TX-A, f_TX-B respectively. However, in inventive metal detectors operating according to the inventive method any number of control channels can be implemented.

0072The signal controller4, 4A; 4B is preferably operated in two or more modes, in which different sets of control terms are applied to the control units .<figref idref="f0005">Fig. 5</figref> shows a flow diagram with different operating modes, in which compensation signals a_COMP are generated.

0073According to the flow diagram the following steps are performed <ul id="ul0005" list-style="none"><li><img file="EP3726257A1_D0018.tif" /> the metal detector is initialized (SYSTEM START)</li><li><img file="EP3726257A1_D0019.tif" /> then the metal detector is set to a first control mode (CONTROL MODE 1);</li><li><img file="EP3726257A1_D0020.tif" /> in CONTROL MODE 1 a first set of control terms Kp, Ki, Kd is applied to the first control unit 41I and to the second control unit 41Q, which control terms Kp, Ki, Kd are selected in such a way that the digital in-phase control components d_CI, and the digital quadrature control components d_CQ are provided with a first correction factor;</li><li><img file="EP3726257A1_D0021.tif" /> CONTROL MODE 1 is maintained for a defined period of time or as long as the imbalance signal exceeds a given threshold;</li><li><img file="EP3726257A1_D0022.tif" /> in the event that the defined period of time has lapsed or the imbalance signal has fallen below the given threshold a second control mode (CONTROL MODE 2) is entered and the metal detector is changed to a state, in which measurement processes can be executed;</li><li><img file="EP3726257A1_D0023.tif" /> in CONTROL MODE 2 a second set of control terms Kp, Ki, Kd is applied to the first control unit 41I and to the second control unit 41Q, which control terms Kp, Ki, Kd are selected in such a way that the digital in-phase control components d_CI and the digital quadrature control components d_CQ are provided with a second correction factor that is lower than the first correction factor and that leads to reduced corrections; and</li><li><img file="EP3726257A1_D0024.tif" /> CONTROL MODE 2 is maintained as long as the imbalance signal does not exceed the given threshold.</li></ul>

0074CONTROL MODE 1 is maintained for example for period of time, in which the imbalance signal is typically reduced to a level, at which the imbalance signal has no significant influence on the measurement. Alternatively, the level, at which the imbalance signal has no significant influence on the measurement, is predetermined and compared with the actual imbalance signal. As soon as the imbalance signal is reduced below the corresponding threshold, CONTROL MODE 1 is changed to CONTROL MODE 2, in which measurement of products is performed while the imbalance signal is still observed and reduced in the control loop. In CONTROL MODE 1 the imbalance signal is quickly reduced, while in CONTROL MODE 2 a remaining imbalance signal is slowly driven to a minimum value. The control terms Kp, Ki, Kd for the second control mode are selected such that signal corrections in the control loop do not lead to disturbances during measurement of the products. In particular, control terms Kp, Ki, Kd for the control units are selected such that overshoots are avoided. In the second control mode operating conditions are therefore continuously optimized, while measurement can be performed.

Reference signs list

0075<dl id="dl0002" compact="compact"><dt>1</dt><dd>transmitter unit</dd><dt>10</dt><dd>frequency generator in the transmitter unit 1</dd><dt>2</dt><dd>coil system</dd><dt>21</dt><dd>drive coil</dd><dt>22, 23</dt><dd>detection coils</dd><dt>24</dt><dd>compensation coil</dd><dt>3</dt><dd>receiver unit</dd><dt>31</dt><dd>balanced transformer</dd><dt>32</dt><dd>input stage with amplifiers and/or filters</dd><dt>33</dt><dd>subtraction unit</dd><dt>34I, 34Q</dt><dd>phase detectors</dd><dt>35I, 35Q</dt><dd>filter units</dd><dt>36I, 36Q</dt><dd>gain units; controllable gain units</dd><dt>37I, 37Q</dt><dd>analogue to digital converters</dd><dt>38</dt><dd>phase shifter</dd><dt>4; 4A, 4B</dt><dd>signal controller; signal controllers</dd><dt>40I, 40Q</dt><dd>low pass filter</dd><dt>41I, 41Q</dt><dd>control units</dd><dt>42</dt><dd>converter (CORDIC)</dd><dt>5; 5A, 5B</dt><dd>compensation unit; compensation units</dd><dt>51, 51L</dt><dd>frequency synthesizer, look up table</dd><dt>52; 52I, 52Q</dt><dd>Multiplier</dd><dt>53</dt><dd>adder</dd><dt>54</dt><dd>A/D converter</dd><dt>6</dt><dd>Amplifier</dd><dt>60</dt><dd>single-ended-to-differential converter</dd><dt>9</dt><dd>control unit preferably comprising - the signal controller 4 - the compensation unit 5 - a digital signal processor for processing signals relating to products contaminants and/or imbalances</dd><dt>a_I, a_Q</dt><dd>analogue in-phase and quadrature signals</dd><dt>a_COMP</dt><dd>analogue compensation signal</dd><dt>C_M</dt><dd>magnitude control signal</dd><dt>C_PH</dt><dd>phase control signal</dd><dt>d_I, d_Q</dt><dd>digital in-phase and quadrature signals</dd><dt>d_CI, d_CQ</dt><dd>digital in-phase and quadrature control signals</dd><dt>d_COMP</dt><dd>digital compensation signal</dd><dt>d_CORR</dt><dd>digital correction signal</dd><dt>f_C</dt><dd>frequency control signal</dd><dt>f_G</dt><dd>generated frequency signal</dd><dt>f_TX; f_TX-A, f_TX-B</dt><dd>operating frequencies</dd><dt>Kp, Ki, Kd</dt><dd>control terms</dd><dt>mcl</dt><dd>master clock</dd><dt>s1</dt><dd>transmitter signal</dd><dt>s2</dt><dd>input signal at the receiver unit 3</dd><dt>s3</dt><dd>signal in the receiver path</dd><dt>S_RI, S_RQ</dt><dd>reference signals</dd></dl>