Sensor

The invention relates to a sensor comprising a microwave resonator .

The prior art describes sensors comprising microwave resonators which enable measurement of dielectric properties of materials .

Each sensor exhibits a certain resonating frequency and a certain quality factor which depend on the mechanical and electrical properties of the sensor itself. By associating the a material to be examined to the sensor a change in the resonating frequency value is detected as well as a change in the sensor quality factor; these changes depend on the dielectric properties of the material to be examined, and in particular on the permittivity and the loss factor.

It follows that by detecting the change in the resonator frequency and the change in the quality factor, it is possible to obtain information on the permittivity and the loss factor of the material to be examined. In addition, the dielectric properties of the material to be examined can be correlated to the percentage of moistness and the density of the material itself.

The prior art includes an open end coaxial resonator (resonator λ/4) . Also known are sensors comprising a cylindrical resonator.

The known-type sensors comprise a radiofrequency circuit and a two-port resonator, the two ports being connected to the radiofrequency circuit by means of coaxial connections . The known-type sensors enable measurements to be made both in reflection and in transmission.

In order to obtain the maximum power transfer between the radiofrequency circuit and the resonator, the impedance at the input port must be 50 ω. Since the impedance at the sensor input port also depends on the type of material to be examined, it is necessary to

perform an impedance matching of the sensor at the input port .

A drawback of known-type sensors consists in the fact that they have very low versatility levels. In fact, the known-type sensors can be impedance-matched only by changing the mechanical properties thereof, in particular the size thereof.

It has been found that for each type of product to be examined, it is necessary to set up a sensor that, once associated to the product to be examined, exhibits an input port impedance as close to 50 ω as possible.

In other words, for each material to be examined, a dedicated sensor has to be set up.

In practice, a sensor realised for examining a certain material and which is used for examining a different material can be affected by an unacceptable loss of precision.

The sensors are equipped with a control unit which is able to calculate the values of the dielectric characteristics of a material and associate to them the values of a property of a material, for example the percentage of moisture, or the density.

The known-type sensors can be supplied by current or tension. In the first case, the sensors are provided with a loop which projects internally of the cavity of the resonator.

The loop comprises a portion of metal file which is ring- formed .

A drawback in known sensors is that the loops are very weak and exhibit a low mechanical resistance, in particular to vibrations .

In addition, it is very difficult to produce a plurality of loops which exhibit exactly the same size and shape.

As the electrical response of the sensors depends on the loop's geometry, each sensor must be individually calibrated.

The algorithm of calibration (i.e. the correlation between electrical response of the sensitive element and the dielectric properties of the material to be examined, the density or the moisture percentage) which has been set up for a sensor provided with a loop having a certain shape and a certain size, is not applicable to a further sensor which, though having the same conformation as the first sensor, differs therefrom due to undesired variations in the geometry of the loop . A further defect of known-type sensors consists in a considerable difficult of positioning of the loop when it is to be inserted in a solid material which constitutes the sensing body, for example in the case of sensors provided with a cylindrical resonator which is not in air. An aim of the present invention is to obtain a sensor comprising a microwave resonator in which the impedance at an input port of the microwave resonator can be easily adapted. A further aim is to obtain a sensor comprising a microwave resonator having a loop for current supply which is provided with considerable resistance and which can be highly repeatably realised.

A further aim is to obtain a sensor comprising a microwave resonator in which a supply device can be inserted simply in a dielectric positioned in a cavity of the resonator.

A further aim is to obtain a sensor comprising a microwave resonator in which the material to be examined is prevented from adhering to a measuring zone of the sensor itself, and in which a portion of material to be examined that has adhered to the measuring zone can be easily removed.

In a first aspect of the invention, a sensor is provided comprising a microwave resonator, radiofrequency circuit means connected to the microwave resonator, and port means through which the radiofrequency circuit means supply the microwave resonator; the microwave resonator exhibiting an impedance that depends on a product to be examined,

characterised in that an impedance matching device is associated to the port means, which impedance matching device is conformed such that the radiofrequency circuit means detect a microwave resonator impedance which is different from the said value.

In a version, the impedance matching device comprises a concentrated parameter circuit, for example an SMD (Surface Mounted Devices) circuit. In a further version, the matching device further comprises a MEMS device.

In a second aspect of the invention, a MEMS device is used in a sensor as a means for matching, to a predetermined value, an impedance encountered by radiofrequency circuit means which supply a microwave resonator. Thanks to these aspects of the invention, a sensor impedance can be matched after the sensor has interacted with a product to be examined.

This enables the power transmitted by the radiofrequency circuit means to the microwave resonator to be maximised and, consequently, also improves the efficiency of the sensor, as well the sensitivity thereof and the accuracy of the measurement .

This also enables use of the sensor to examine, with very high accuracy, materials having dielectric properties - and therefore density and moisture - which belong to a wide range of values: low-loss (tanβ <<1) , medium-loss (tanδ comparable to 1) and high-loss (tan6 >>1) .

The sensitivity of the sensor measurement is regulated by modifying the impedance at a port of the sensor using only an matching circuit.

In this way, the matching, and therefore the regulation of the sensitivity of the measurement, can be performed without removing the sensor from a production line. In order to match the impedance of the sensor it is not necessary to perform any complicated modifications thereon,

especially mechanical modifications, which are required in the prior-art sensors.

In a third aspect of the invention, a sensor is provided which comprises a microwave resonator, radiofrequency circuit means connected to the microwave resonator, supplying means through which the radiofrequency circuit means supply the microwave resonator, characterised in that the supplying means comprise a first element and a second element which are mutually distinct and which cooperate to define loop means for supplying current to the microwave resonator.

In a version, the first element and the second element are arranged transversally of one another in order to define loop means in a cylindrical resonator. In particular, the first element extends from a base wall of a cylindrical casing of the cylindrical resonator, projects through a cavity identified by the casing and penetrates into a seating afforded in a further base wall of the casing, opposite the above-mentioned base wall. The second element extends from a lateral wall of the cylindrical casing through a further seating afforded in the further base wall, the further seating opening into the seating, in such a way that the second element interacts with the first element . In a further version, the first element and the second element are arranged substantially parallel to one another and are mutually interconnected by a third element, and thus define loop means in an open ended coaxial resonator. In particular, the first element and the second element extend from a wall of a casing of the open-ended coaxial resonator which wall is opposite a zone predisposed to interact with a material to be examined, while the third element is housed internally of the casing and connects the first element to the second element .

Thanks to these aspects of the invention, loop means can be obtained which are made up of easily-assembled modular elements . This enables a plurality of loops to be obtained, each having the same shape and size, each to be associated to a respective resonator.

In addition, a plurality of sensors of a same type can be obtained having supply means all exhibiting the same geometry, and thus the same electrical responses. In this way, it is possible to set up a single calibration procedure which connects the measured electrical values to the properties of a material to be examined (for example dielectric function, density, granulometry, moisture percentage, salt, fat/lean mixture), the calibration procedure being applicable to all the sensors.

Further, the loop means are mechanically very robust and especially resistant to vibrations.

In addition, the loop means can be realised much more easily than the loops of the prior art, inasmuch as it is not necessary to shape a tract of metal wire, but it is sufficient to assemble components of predetermined sizes, previously realised.

In a fourth aspect of the invention a sensor is provided which comprises a microwave resonator provided with casing means and dielectric means housed in cavity means of the casing means,- radiofrequency circuit means connected to the microwave resonator, supply means through which the radiofrequency circuit means supply the microwave resonator, characterised in that the supply means comprise a supply element conformed in such a way as to extend from a first zone of the casing means to a second zone of the casing means through the dielectric means .

In a version, the supply element comprises an SMA-type launcher, or an N type launcher. Thanks to this aspect of the invention supply means for supplying a microwave resonator with current can be simply obtained.

In a fifth aspect of the invention, a sensor is provided which comprises a microwave resonator, radiofreguency circuit means connected to the microwave resonator, the microwave resonator comprising casing means internally of which dielectric means are housed, characterised in that the dielectric means comprise a first dielectric element made of a first material and a second dielectric element made of a second material . The dielectric means normally used in microwave resonators exhibit excellent dielectric properties, but are not suitable to be subjected to mechanical working, for example piercing, as they are very fragile.

Thanks to this aspect of the invention, a microwave resonator can be equipped with a first element made with a usual dielectric material having excellent dielectric properties, for example alumina (Al ₂ O ₃ ), and with a second dielectric element made of a further material, such as polytetrafluoroethylene (PTFE) , which apart from having satisfactory dielectric properties - slightly less than the properties of the first dielectric material - can also be easily subjected to mechanical working.

In this way dielectric means are obtained having excellent dielectric properties overall, and which are provided with a portion - the second dielectric element - in which mechanical working can easily be carried out .

In particular, dielectric means can be obtained in which holes can easily be made for passing a portion of supply means - for example a launcher - through which the radiofrequency circuit means supply the microwave resonator. In a sixth aspect of the invention, a sensor is provided which comprises a microwave resonator, radiofrequency circuit means connected to the microwave resonator and provided with transmitting means and receiving means, the receiving means comprising a receiver device, a terminal predisposed to read a signal coming from a port of the microwave resonator indicating a power reflected from the

microwave resonator and a further terminal predisposed to read a further signal coming from a further port of the microwave resonator and indicating a power transmitted through the microwave resonator, a circulator connected to the transmitting means, as well as to the terminal and the port, and a switch predisposed for alternatively connecting the receiver device with the terminal or with the further terminal . In a version, the switch comprises a MEMS coaxial switch (RF MEMS coaxial switch) .

Thanks to this aspect of the invention a reading can be obtained of a parameter indicating the reflected power and a further parameter indicating the transmitted power, leading to a more accurate reading of the properties of a product to be examined.

In particular the above-mentioned parameter and the further parameter can be combined, for example via a linear combination.

In a seventh aspect of the invention, a heating device is used in a sensor as an anti-adherent means for preventing a product to be examined from adhering to a measuring zone of a microwave resonator.

Thanks to this aspect of the invention, a microwave resonator sensor can be obtained in which material deposits do not form in proximity of a measuring zone,- also, there is no formation of air bubbles, which can hamper the effectiveness of a reading.

In particular, condensation can be prevented from forming on a portion of the container means to which the measuring zone of the sensor is associated, which condensation might mix with a part of the material, forming accumulations and encrustations which would prevent a remaining part of the material from interacting sufficiently with the microwave resonator. In an eighth aspect of the invention, a measuring apparatus is provided comprising a microwave resonator apparatus

associated to a portion of the container means for receiving a product destined to interact with the microwave resonator sensor, characterised in that it further comprises cleaning means predisposed to remove accumulations of the material from the portion.

Thanks to this aspect of the invention, a measuring apparatus is obtained in which it is not necessary to remove the microwave resonator sensor in order to clean it.

The invention can be better understood and actuated with reference to the figures of the drawings, which illustrate an example of a non-limiting embodiment thereof, in which:

Figure 1 is a block diagram of a measuring apparatus comprising a microwave resonator sensor;

Figure 2 is a block diagram of a variant of a measuring apparatus comprising a microwave resonator sensor;

Figure 3 is a perspective section view of a microwave sensor, of a resonating coaxial and open-ended type;

Figure 4 is a detail of figure 3;

Figure 5 is the view of figure 3 showing the microwave sensor associated to a wall of a container,-

Figure 6 is the view of figure 3, showing a variant of the microwave sensor;

Figure 7 is the view of figure 3, showing a further variant of the microwave sensor; Figure 8 is a perspective view in section of a microwave sensor, of a cylindrical resonator type;

Figure 9 is the view of figure 8, showing a variant of the microwave sensor;

Figure 10 is a perspective view of a microwave sensor fixed to a plate to which heating means are associated;

Figure 11 is a perspective view of a microwave sensor to which cleaning means are associated;

Figure 12 is a block diagram of a measuring apparatus comprising a microwave sensor and an impedance matching device;

Figure 13 is a diagram of an impedance matching device;

Figure 14 is a diagram of an equivalent circuit formed by a radiofrequency circuit, an impedance matching network and a microwave sensor;

Figure 15 is a graph showing experimental results of measurements of an input phase shift;

Figure 16 is a graph showing experimental results of an output phase shift;

Figures from 17 to 20 are Smith charts relating to an application example of a procedure for impedance matching; Figure 21 contains four graphs relating to an experimental application of an impedance matching procedure;

Figure 22 shows four Smith charts which correspond to the four graphs of figure 21;

Figures 23 and 24 show two equivalent circuits relating to due of the cases illustrated in figures 21 and 22;

Figure 25 is a block diagram of a further variant of a measuring apparatus comprising a microwave resonator sensor,-

Figure 26 is a table obtained experimentally through a calibration procedure of a microwave sensor of the type illustrated in figure 6;

Figure 27 shows a table obtained experimentally through a calibration procedure of a microwave sensor of the type illustrated in figure 7;

Figure 28 shows a table obtained experimentally through a calibration procedure of a microwave sensor of the type illustrated in figures from 3 to 5;

Figures from 29 to 33 show some graphs constructed on the basis of the table of figure 28.

With reference to figure 1, a measuring apparatus 1 is shown, comprising a sensor 2 - of a type provided with a microwave resonator - a radiofrequency circuit 3 connected to the sensor 1 by a coaxial connection and an electronic control unit 4 , for example an industrial PC and/or an electronic card with DSP and/or processor, connected to the radiofrequency circuit by an RS-232 and/or RS-485 serial

interface, a parallel port, an Ethernet port, analog outputs .

The measuring apparatus 1 is predisposed to perform measurements in transmission.

Figure 2 illustrates a measuring apparatus 1 in which a circulator 5 is interposed between the radiofrequency circuit 3 and the sensor 3.

The measuring apparatus 1 is predisposed to make reflection measurements.

The sensor 2 is a fringe-field lambda/4 resonating sensor having a typical response as shown in the table herein below.

where :

S ₁₁ = reflected power/incident power

S ₂₁ = transmitted power/ incident power

MUT = Material Under Test

Fr = resonance frequency

Bw = bandwidth at -3 dB

Q-factor = resonance frequency - Bw ratio

Amp = width of S ₂ i or S ₁₁ at resonance peak

SWR = stationary wave ratio

Real = real part of impedance at Fr expressed in Ohm

Imag = imaginary part of impedance at Fr expressed in Ohm.

Alternatively, the sensor 2 can be a cylindrical resonating sensor.

The sensor 1 body is made of stainless steel 304, or steel

306, or Anticorodal aluminium, the contact plugs are made of virgin polytetrafluoroethylene PTFE) , the connectors/ports are SMA and/or N type.

The sensor 1 can operate in a temperature range of between 0

- 80 ⁰ C .

The radiofrequency circuit 3 enables scalar measurement of parameters S ₂ i and Sn in a frequency range of between 500 MHz and 3000 MHz . The radiofrequency circuit 3 has the following characteristics :

Operating temperature 0 í 50 ⁰ C

Supply 12 í 24 VDC

Characteristic impedance 50 ω Max. consumption 5 W

Typical output power 10 dBm

Frequency range 500-700 MHz

S _2x dynamic 50 dB (-10 í -60 dB)

Revealed signal dynamic range (Sn) 20-23 dB Frequency synthesis accuracy 500 Hz

Response width accuracy +/- 0. IdB (between 0 í 50 ⁰ C, with calibrated system)

Minimum step frequency 50 KHz

Sweep time < 10ms The connectors/interfaces have the following characteristics:

PC remote interface DB9

RF interface SMA F

Communication protocol RS232

Input channel A/D converter 16 bit The circuit control interface software enables management of the following parameters :

Frequency range on which to perform the measurement;

Start frequency, frequency step width and number of points to synthesise; Serial port baud rate selection;

Start command for performing a measurement : the system responds by transmitting the number of bytes requested (number of synthesised points*2 bytes) ;

A control of transmitted bytes is performed (1 checksum byte generated by XOR of bytes identifying the measurement) .

Figures from 3 to 5 show the microwave sensor 2 of open- ended coaxial resonator type, comprising a substantially- cylindrical casing internally of which an elongate body 7 is arranged, coaxial to the casing 6 and having a smaller transversal dimension than a transversal dimension of the casing 6, such that between the casing 6 and the elongate body 7 there is an annular hollow space 8.

The elongate body 7 has a substantially constant transversal section. The sensor 2 further comprises a flange 9 to which the elongate body 7 is connected, which flange 9 can be fixed to the casing 6 in order to close an open end 10 of the casing 6. The casing 6 further comprises a further open end 11, opposite the open end 10, to which a plug 12 is associated, the plug 12 being made of a material which is transparent to microwaves .

The further open end 11 is predisposed to interact with a material to be examined. In particular, the further open end 11 can be associated to a wall 65 of a hopper, not illustrated, and be predisposed to receive a certain quantity of the above-mentioned material . The sensor 2 further comprises a launcher 13, of known type, for example an SMA. launcher, connected to a port of the sensor 2 and predisposed to detect an output signal indicating the transmitted power.

The launcher 13 is supplied with tension (probe) . The sensor 2 further comprises a supply device 14 predisposed to supply the sensor 2 with current.

The supply device 14 defines a loop 15 short-circuited on the flange 9.

The loop 15 is constituted by a first portion 16 and a second portion 17 arranged substantially parallel to the elongate body 7 and by a third portion 18 arranged

transversally - and substantially perpendicular - to the first portion 16 and the second portion 17.

The first portion 16, the second portion 17 and the third portion 18 are contained in a same plane. The first portion 16 is constituted by a further launcher 19, of known type, for example an SMA launcher, or an N launcher.

The second portion 17 is constituted by a bar 20 extending from a head 21 associable to the flange 9. The head 21 comprises a threaded zone 22 engaging in a corresponding threaded hole 23 sunk in the flange 9 and opening internally of the casing 6 into the hollow space 8. The third portion 18 comprises a crossbar 24 having a first end 25 associated to a tip zone 26 of the further launcher 19 and a second end 27 associated to a further tip zone 28 of the bar 20.

The crossbar 24 comprises - at the first end 25 - a first hole, not illustrated, predisposed to receive the tip zone 26 and - at the second end 27 - a second hole, not illustrated, arranged to receive the further tip zone 28.

The crossbar 24 can be solidly fixed to the further launcher 19 and to the bar 20 by welding, for example soft-soldering connecting the tip zone 26, housed in the first hole, to the first end 25, and a further soft-soldering connecting the further tip zone 28, housed in the second hole, to the second end 27.

The further launcher 19, the bar 20 and the crossbar 24 make it possible to obtain a loop 15 in a much simpler way than with the prior art, i.e. a loop obtained by bending a metal wire to obtain a ring.

Further, the loop 15 is geometrically much more easily reproducible than the loops of the known apparatus. This means that during manufacturing sensors can be obtained which have geometrically identical loops and therefore equal electrical reactions .

It follows that a same calibration procedure can be used for a plurality of sensors having the same geometrical and electrical characteristics and it is not necessary to set up a calibration procedure for each sensor, as happens in the case of sensors having loops, and therefore electrical responses, which differ from one another.

In addition, the loop 15 is fixed extremely effectively to the flange 9 and is particularly resistant to vibrations. Further, the sensor 2 can be supplied with current through the loop 15, which facilitates an impedance matching of the port connected to the loop 15, in a way which will be more fully discussed herein below.

Figure 6 illustrates a sensor 2 comprising an elongate element 7 provided with a body region 80 and an end region 81, closer to the further open end 11, which has a decreasing transversal section in a distancing direction from the body region 80.

In other words, the end region 81 has a substantially truncoconical shape, a larger base 82 of the cone being adjacent to the body region 80 and a smaller base 83 of the cone being distant from the body region 80.

Figure 7 illustrates a sensor 2 comprising an elongate element 7 provided with a further body region 84 and a further end region 85, closer to the further open end 11, which has a growing transversal section as it distances from the further body region 84.

In other words, the further end region 85 has a substantially truncoconical shape, a further smaller base 86 of the cone being adjacent to the further body region 84 and a further larger base 87 of the cone being distant from the further body region 84.

Figure 8 illustrates a cylindrical resonating sensor 2 comprising a first half-shell 29 and a second half-shell 30 which are mutually connectable by connecting means, for example screws .

The first half-shell 29 and the second half-shell 30 define casing means 90 which delimit a cavity 31, internally of which annular dielectric means 32 are inserted. A first tubular portion 33 is associated to the first half- shell 29 and a second tubular portion 34 is associated to the second half-shell 30. The first tubular portion 33 and the second tubular portion 34 cooperate to define conduit means 35 predisposed to be crossed by a material to be examined. The sensor 2 comprises a further supply device 36 which is provided with a further loop 37 predisposed to enable a current supply to the sensor 2.

The further loop 37 comprises a portion 38 arranged substantially parallel to the conduit means 35 and a further portion 39 arranged transversally - and substantially perpendicular - to the conduit means 35.

The portion 38 and the further portion 39 are contained in a same plane.

The portion 38 is constituted by a still further launcher 40, of known type, for example an SMA launcher, or an N launcher.

The still further launcher 40 comprises a body 41 passing through a first hole 42, realised in the first half-shell 29, and through a second hole 43, realised in the dielectric means 32. The still further launcher 40 comprises a tip zone 44 received in a third hole 45 realised in the second half- shell 30.

The further portion 39 is defined by a further bar 46 extending from a further head 47 associable to the second half-shell 30.

The further bar 46 comprises a further threaded zone 48 engaging in a corresponding threaded portion 49 of hole means 50 afforded in the second half-shell 30 and communicating with the third hole 45. To obtain the third loop 37, after the body 41 of the still further launcher 40 has been positioned in the first hole 42

and in the second hole 43 and the tip zone 44 has been positioned in the third hole, the further bar 46 is introduced into the hole means 50 in such a way that an end 51 of the further bar 46 is placed in contact with the tip zone 44 of the still further launcher.

The threaded zone 48 of the further bar 46 engages with the threaded portion 49 of the hole means 50 such that the further loop 37 is provided with a good level of mechanical resistance. The still further launcher 40 and the further bar 46 enable a further loop 37 to obtained simply,- the further loop 37 crosses the dielectric means 32.

The dielectric means 32 comprise a ring 52 made of alumina (AI2O3) , which, as is known in the prior art, exhibits dielectric properties which make it particularly suitable to be used in sensors comprising microwave resonators. The dielectric means 32 further comprise a further ring 53 made of virgin polytetrafluoroethylene (PTFE) . The ring 52 and the further ring 53 are arranged coaxially, the further ring 53 having an internal diameter which substantially corresponds to an external diameter of the ring 52, in order for the further ring 53 to surround the ring 52. The ring 52 is of a height, i.e. a dimension measured parallel to the conduit means 35, which is substantially equal to a further height of the further ring 53. Polytetrafluorethylene (PTFE) has dielectric properties which make it suitable to be used for realising sensors comprising microwave resonators . Further, the polytetrafluoroethylene (PTFE) has mechanical properties which make it very much easier to work than alumina (Al ₂ O ₃ ) .

In particular, the second hole 43 is obtainable very easily in polytetrafluoroethylene (PTFE) , while complex working would be required to obtain the same second hole 43 in alumina (Al ₂ O ₃ ) .

Thanks to the invention, then, manufacturing costs of the sensor 2 can be considerably limited by using dielectric means 32 comprising, in a zone closest to the conduit means 35, an alumina (Al ₂ O ₃ ) ring and, in a more distant zone from the conduit means 35, in which a seating is to be made which will contain a portion of the further supply device 36, a further ring made of an easily-workable material such as polytetrafluoroethylene (PTFE) . Figure 9 illustrates a cylindrical resonator sensor 2 in a variant version.

The sensor 2 comprises a supply element 91 cooperating with the casing means 90 to define a still further loop 92 predisposed to enable a current supply to the sensor 2. The supply element 91 comprises a further launcher 93, of known type, for example an SMA or N type launcher.

The further launcher 93 is provided with a further body 94 - made of an electrically insulating material - passing through a further first hole 95, afforded in the first half- shell 29, and through a further second hole 96, afforded in the dielectric means 32.

The further launcher 93 comprises a further tip zone 97 - made of an electrically conducting material - received in a further third hole 98 afforded in the second half-shell 30. The further tip zone 97 is placed in contact with the second half-shell 30 - also made of an electrically conducting material - so as to constitute the still further loop 92. A countersunk 99 seating is afforded in the second half- shell 30, which leads into the further third hole 98 and is predisposed to receive a quantity of solder which electrically and mechanically connects the further tip zone 97 to the second half-shell 30.

Alternatively, the countersunk seating 99 can be absent. In this case, the further tip zone 97 and the further third hole 98 are conformed such that the further tip zone 97 interacts with the walls delimiting the further third hole 98.

Thanks to the invention, a current supply device to the sensor 2 can be realised simply.

In addition, the device is realised with a modular component and is highly repeatable. Further, by realising, in the casing means, a seating predisposed to receive a point portion of a launcher and providing for the launcher to cross the dielectric means from side to side, a current supply device of a microwave resonator can be realised using a launcher of a commercially-available type.

The above-described sensors, with reference to figures from 3 to 9 (in particular the casing 6, the elongate body 7 and the flange 9 of the open-ended coaxial resonator sensor 2 and the first half-shell 29, the second half-shell 30, the first tubular portion 33 and the second tubular portion 34 of the cylindrical resonator sensor 2) are made of stainless steel, or Anticorodal aluminium, and are subjected to a surface treatment with SurTec ^® 650 which stabilises the electrical characteristics of the above-cited materials over time.

Experiments have shown that the surface layer that forms following treatment with SurTec 650 stabilises over time the electrical properties of the structure of the sensing body. SurTec ⁰ 650 treatment can be done by immersion of the sensor in the product. This makes the surface treatment operations very simple and rapid and therefore is an aid to production cost limitation. A further surface treatment used together with SurTec 650 or alternatively thereto is hard anodisation. Experimentation has shown that the surface layer of aluminium oxide which forms following anodisation treatment does not harm the electrical properties of the sensing body at the operating frequencies. With reference to figure 10, a sensor 2 is fixed to a plate 54 associated to a wall 55 of a conduit along which a

product to be examined by the sensor 2 advances, for example in order to perform dielectric property measurements, e.g. moisture measurements and/or product density measurements. The wall 55 is provided with an opening, not illustrated, which enables the product to interact with the sensor 2.

The conduit 56 can be a bunker predisposed to receive and transport a product, such as wood-chip or fibre, or can be a hopper predisposed to receive and transport inert powders (for example atomised ceramic powders) . Heating means 56 are associated to the plate 55, which heating means 56 are predisposed to heat the plate 55 and the sensor 2 in proximity of a measuring zone of the sensor 2 predisposed to interact with the material to be examined. The heating means 56 comprise a plurality of heating elements 66 fixed to the plate 55.

In particular, the heating means 56 comprise four heating elements 66 positioned in proximity of the corners of the plate 55. The heating means 56 keep the plate 55 and the sensor 2 at a predetermined temperature, for example within a range of 30- 70 ⁰ C. The heating means 56 prevent the product from forming encrustations and deposits close to the above-mentioned measuring zone of the sensor 2. In particular, the heating means 56 maintain the plate 55 and the sensor 2 at an equal or higher temperature to that of the product, thus preventing any water vapour present in the product from condensing on the plate and from becoming mixed up with the product and creating agglomerations which would considerably reduce the measuring sensitivity of the sensor 2, or which might even render it unusable.

In particular, the heating means 56 keep the plate 55 and the sensor 2 at a controlled temperature, improving accuracy of measurement . With reference to figure 11, a sensor is shown which is associated to a side 57 of a container 58 predisposed to remove a material to be examined from a tube 59.

The sensor 2 and the container 58 are supported by a slide truck 60 activated by an actuator 62 to slide on guides 61 between a sample gathering position and a rest position. When the slide truck 60 is in the rest position, a vertically mobile brush 63, as indicated by the arrow F, penetrates internally of the container 58 to remove any encrustations of product from the side 57; eventual presence of encrustations could affect the sensitivity of the sensor 2. The invention includes a procedure of impedance matching of the sensor 2 and a device therefor. The impedance matching procedure of the sensor 2 is described herein below. Description of the problem. With reference to figure 12, once calibration of the cables has been carried out, the impedance Z _s perceived by the sensor 2 is 50 ω while the impedance measured by the radiofrequency circuit 3 at the port 1 (Zi _n ) depends on the product which has been placed before the sensor 2. The maximum transfer of power between circuit and sensor occurs in the conjugate matching condition: the circuit must "see" a 50 ω impedance, while towards the circuit the sensor must "see" a complex conjugate impedance Zi _n ^* of Zi _n . Diagram of the matching network. With reference to figure 13, the matching network which mostly simply enables a match to be made at the port 1 of the resonator is a simple concentrated parameter circuit in which there is a series reactance (inductive or capacitive) and a parallel susceptance (inductive or capacitive) . The signal is carried through a microstrip (characteristic impedance Z _{c =} 50 ω) .

The microstrip exhibits a gap between the two SMAs in which the series SMD component can be inserted; the parallel SMD component is soldered between the strip and the ground. The RF circuit is connected to the SMA on the Input side, while the SMA on the Output side is connected directly to the sensor port.

The equivalent circuit formed by the RF circuit, matching circuit and sensor is shown in figure 14.

The input-output shifts at input and output of the matching network are due to the SMA connectors on the card (and the adaptors) and to the microstrip which is of a certain length.

If the line carrying the signal is sufficiently short not to lead to width attenuation, the two shifts can be modelled with a transmission line without loss of characteristic impedance Z _c =50 ω the shift of which is easy to measure with a network analyser.

The graphs of figures 15 and 16 report the results of the measurements of the shift of Sn at the two ports (since the reflected power runs through the microstrip twice, the values are double the sought shift Aφ) .

With the resonating frequency to be matched known, the inlet and outlet shift values can be calculated.

From these calculations the length 1 of the 50 ω equivalent transmission line can be derived:

2Aφ=2β=^-l -> 1 =^-λ ^{ψ H} λ 2π

Matching procedure.

The matching is performed at a predetermined frequency (resonance frequency of the product to be matched) . The input impedance is measured at that frequency.

EXAMPLE: Product X ^ fr = 597.1 MHz; Zin = 8.67 + J25.04

With reference to figures from 17 to 19, the Z _ser i _es and Yp _ara ii _e i values enabling matching are those which move the impedance from the centre of the Smith chart (Point 1 = 50

ω) to Point 2 which represents the conjugate complex impedance Z _in ^* = 8.63 - J25.04.

To f ind the Z _series and Ypa _ra ii _e i values it is advisable to perform a shift de-embedding .

If the transmission line is effectively at 50 ω, the reflection coefficient is not attenuated but only phase shift. This means that the shift generated by the line is clockwise on a circumference centred in the origin of the Smith chart (points with constant | Si ₁ 1 ) . For this reason, since point 1 is exactly in the centre, the first shift (input shift) does not cause shifts on the Smith chart. The point 2 is downstream the output shift, in order to go upstream the shift it is necessary to follow the circumference at | Sn | constant in an anticlockwise direction. At 597.1 MHz the output shift is:

Aφ=-*(-0.1968*597.1+5.4163)= -56.0465

/ = 0.15568A

The point of arrival (point 3) is on the circumference centred on point 1 at -0.155 λ from point 2 (the minus sign is because of the anticlockwise movement) . At this juncture it is necessary to shift from Point 1 to

Point 3 through a series reactance and a parallel susceptance.

A series reactance leads to a shift on the circumferences with constant R in the Smith chart of impedances; the shift is clockwise if the reactance is inductive (ωL) , and anticlockwise if it is capacitive (-1/ωC) .

Parallel susceptance, on the other hand, leads to a shift on the circumferences at G constant in the Smith chart of the admittances; in this case the shift is clockwise if the susceptance is capacitive (coC) and anticlockwise if inductive (-1/ωL) .

Point 3 has an impedance of Z=204.44-j*173.48 which corresponds to an admittance of Y=O .0028+j*0.0024.

The series reactance must lead to a shift on the circumference R=50 ω up to point 4 which will have the same

R as point 1 (50 ω) and the same G as point 3 (0.0028 S) .

The parallel susceptance will be the one that leads from point 3 to point 4 by moving on the circumference at G =

0.0028 S.

The reactance going from points 1 to 4 is an inductance (movement in a clockwise direction) of 33.2 nH. Going from point 4 to point 3 is achieved with a parallel capacity

(clockwise movement) of 2.5 pF.

Matching procedure at Is ₁₁ ] = -25dB

With reference to figure 20, taking the preceding case but with the difference that no match to 50 ω is sought, but rather it is desirable that the value of | Sn | be predetermined (for example-25 dB) at the desired frequency

(597.1 MHz) . The point which corresponds to the above value in the Smith chart must be found. The input impedance of a circuit can be derived starting from the value of Is ₁₁ I and arg (S ₁₁ ) :

1 + S ₁₁ |(cos(arg(S _n )) + j sin(arg(S _n )))

1- 1- S _n 1(COsCaTg(S ₁₁ ))+ jSiB(OTg(S ₁₁ ))) If matching is to be IS ₁₁ I = -25dB = 0.056235, keeping the sensor over-coupled, arg (S ₁₁ ) = 90° can be fixed.

The impedance value on the Smith chart of point 1 (starting point) will be:

The starting point of the matching is no longer at the centre of the Smith chart, so the shift introduced by the microstrip in input will lead to a shift which was not previously taken into consideration. This shift is a clockwise rotation on the circumference at )Sn| = -25dB. The displacement is calculated from the preceding graphs: At 597.1 MHz the output shift is:

Aφ=-*(-0.034*597.1-20)=-20.1507

I = 0.5597λ

The rotation leads to point 5 which has an impedance of

Z=53,583-j*4,6 ω.

At this point the steps are identical to the preceding case and the matching is obtained with a series L and a parallel C to reach point 3. From here, with the output shift point 2 is reached (arrival point) .

The above-described matching procedure enables a circuit having concentrated parameters to be set up, for example an SMD (Surface Mounted Device) circuit, comprising a series reactance (inductive or capacitive) and a parallel susceptance (inductive or capacitive) by which the impedance at the sensor input port can be adapted to 50 ω. The sensor 2 - destined to interact with a certain number of materials - can be equipped with a group of concentrated parameter circuits, each of which enables the sensor to be adapted in impedance, when associated to a material. The above-described matching procedure enables matching devices to be realised which can be associated to microwave sensors to make them suitable for interacting at very high sensitivity with a certain number of products having properties, for example density and/or humidity, the values of which belong to a predetermined range, without there being any need to intervene mechanically on the sensor in order to change the geometry thereof and, consequently, the impedance thereof .

The above-described matching procedure, however, requires numerous calculations to be made, as well as the practical realisation of concentrated parameter circuits. The invention includes use of a MEMS device (Micro-Electro- Mechanical-Systems) for adapting the impedance of the sensor when empty or when loaded with product, in order to optimise the bulk moisture and density measurements. In this way the measurement sensitivity is regulated by changing the impedance at the sensor ports using only a matching circuit. A considerable advantage is that the matching and therefore

the regulation of the sensitivity of the measurements can be performed in-line, which means avoiding complex modifications to the sensor, as is the case with mechanical adjustment.

In other words , the micro switches of the MEMS device are extremely rapidly activated to obtain the desired impedance value .

In particular, the device can realise a reconfigurable matching network with discrete components, using commercial

Teravicta ^® RFMEMS switches. The MEMS device enables impedance matching of the sensor 2 with extreme simplicity and rapidity, and without the sensor

2 having to be demounted from the plant it is installed in.

This is particularly advantageous in the case of plants comprising a large number of sensors . Tests have been carried out on material constituted by wood chip and fibre using a sensor associated to a bunker wall along which the material is advanced.

Samples of material were examined, having various levels of moisture (cases A, B, C and D) . Figure 21 shows the progress of Sn according to the frequency when the sensor is not impedance-matched, when the sensor is impedance-matched in the absence of product (i.e. in the presence of air) , when the sensor is impedance-matched interacting with the material of case A and when the sensor is impedance-matched interacting with the material of case D.

Figure 22 shows the Smith charts corresponding to the diagrams of figure 21.

Figure 23 shows the matching network for case A, while figure

24 shows the matching network for case D. With reference to figure 25, a measuring apparatus 1 comprises a microwave sensor 2 and a radiofrequency circuit 3 connected to the microwave sensor 2 and provided with transmission means 71 and receiving means 72.

The receiving means 72 comprise a receiving device 73, a terminal 74 connected to a port 75 (port 1) of the sensor 2 and predisposed to detect parameter Si ₁ and a further

terminal 76 connected to a further port 77 (port 2) of the sensor and predisposed to detect parameter S ₂₂ • The measuring apparatus further comprises a circulator 5 connected to the transmission means 71, to the terminal 74 and a switch 70 predisposed alternatively to connect the receiving device 73 with the terminal 74 or with the further terminal 75.

The switch can comprise an RF MEMS coaxial switch. The switch 70 enables a more accurate measurement of the product .

In particular, the switch 70 enables obtaining a response which is a combination, for example linear, of Sn and S ₂ i. The invention also provides a calibrating procedure, i.e. a procedure which enables relations to be made between dielectric quantities of the material 2 measured using the sensor 2 and material properties such as density and/or the degree of moisture.

The calibration procedure of the sensor is described herein below. The sensor calibration is done in the laboratory and is separated into three distinct stages. STAGE 1: CHARACTERISATION OF THE MATERIAL.

1. Determination of the range of moisture and density to be monitored; 2. Laboratory measurements on 3-4 samples of material with moistness distributed within the selected parameter; 3. Laboratory measurements on the same sample, generating where possible variations of product density or selecting samples having different densities; 4. For each measurement the microwave values measured by the sensor can be associated to a humidity and density value ;

5. Analysis of the data using CST Microwave Studio ^® simulations in order to associate to each measurement a corresponding dielectric constant and a loss tangent; the sensor response is saved in touchstone format, then

imported into the simulator which has the task of finding the pair (ε', tanD [ie. the ratio ε''/ε']) with which the simulated response is as close as possible to the measured response . STAGE 2 : CHARACTERISATION OF THE SENSOR

6. Determination of the (ε', tanD) parameters corresponding to the maximum moisture and density variations which the production process will induce on the product;

7. Mapping using CST Microwave Studio ⁸ parametric simulations of the variations the sensor will see when measuring the product ;

8. Each (ε', tanD) pair can be associated to corresponding microwave parameter values.

STAGE 3 : DATA ANALYSIS

9. Study of the connection between (ε' _; tanD), Moisture (M%) and Density (p) for the product considered;

10. Determination of the relations

11. Study of the connection between (ε',ε") and the sensor microwave parameters; 12.Determination of the relations:

This procedure enables obtaining a calibration which can be transportable from sensor to sensor (even if they do not belong to the same family) . in fact, STAGE 1 and points 9 and 10 of STAGE 3 are common to all the sensors and depend only on the product .

Figures from 26 to 28 contain some tables obtained by experimentally applying STAGE 1 of the above-described

calibration procedure.

Figures from 29 to 33 contain some graphs, constructed using the data from the table of figure 28.

Once the calibration procedure has been set up and trimmed, the sensor 2 can be used as an instrument for performing nondestructive examinations of materials in order to measure the values of some properties thereof.

In particular, the examinations can be performed on significant quantities of material, which enables errors to be kept at a minimum.

In particular, by using considerable quantities of material it is possible to minimise changes and lack of homogeneity due to air present in the product .