Inductive Sensor Based on Micromachined Coil for Conductive Target Detection

This letter presents an inductive sensor based on a micromachined coil for the detection of conductive targets. The proposed sensor consists of a planar micromachined sensing coil, a signal conditioning instrumentation, and a nearby nonferromagnetic conductive target. The working principle relies on the magnetic coupling between the sensing coil, driven by AC current, and the target. Due to the eddy currents induced in the target, the magnetic flux and, thus, the equivalent inductance of the sensing coil is altered as a function of the target position. Two configurations have been experimentally adopted to explore the proposed inductive sensor. A proximity sensor has been achieved by measuring the inductance variation while changing the distance between the coil and a brass tip with a diameter of 3 mm. A nondestructive defect detector has been achieved by sliding a copper target over the sensing coil. For the proximity sensor, a sensitivity of 6500 nH/m within the range 0–50 µm has been achieved. The resolution of the proximity sensor at one standard deviation is 470 nm. For the defect detector, a 1-mm wide and 35-µm deep cut induced in the target has been detected. Experimental results have confirmed the possibility to exploit the proposed sensor in both configurations, thus validating its working principle in good agreement with the theoretical analysis.


I. INTRODUCTION
The capability to detect conductive targets or to evaluate the presence of cracks in conductive materials by means of nondestructive tests (NDTs) is of great importance.In industrial scenarios, NDTs are adopted to assess in-line material properties and trace sheet metal processing [1].Electromagnetic (EM) [2], ultrasonic [3], optical [4], and liquid penetrant [5] approaches are typically adopted to detect cracks in conductive targets.However, EM approaches have the advantage, among the others, of being time and cost-effective and employable in harsh operating conditions [2].A conventional EM method is the eddy current, which is based on the magnetic coupling between a magnetic field source and the conductive target [6].These sensors are suitable for position measurement of conductive targets through intervening nonmetallic materials, such as, plastic or glass.To this extent, they are adopted in many applications, as displacement sensors [7], as defect detectors [8], as proximity sensors [9], and as evaluators of material purity and alloy compositions [10].To further extend the application possibilities of eddy current sensors, spatial resolution and dimensions of the sensing units are critical issues.To this extent, by leveraging the advantages of MEMS fabrication processes, such as, a high degree of miniaturization and integration, micromachined sensing units can Corresponding author: A. Nastro (e-mail: alessandro.nastro@unibs.it).Associate Editor: Jeong Bong Lee.Digital Object Identifier 10.1109/LSENS.2024.3426102be fabricated [11].In this context, an inductive sensor based on a micromachined coil has been fabricated and experimentally validated.Tailored experimental setups have been developed to test the sensor as a proximity sensor and as a defect detector for conductive nonferromagnetic targets in a static regime.Both configurations have been experimentally validated, thus confirming the possibility to exploit a reliable micromachining fabrication process for the development of sensing coils for conductive target detection.
The rest of this letter is organized as follows: Inductive sensor description is presented in Section II.Experimental results are presented in Section III.Finally, Section IV concludes this letter.

II. INDUCTIVE SENSOR DESCRIPTION
The proposed inductive sensor consists of a planar micromachined sensing coil, a signal conditioning instrumentation, and a nearby conductive target.By driving the sensing coil with AC current, an alternated magnetic field is generated, thus inducing eddy currents in the conductive target and in turn establishing a magnetic coupling between the coil and the target.The eddy currents generate a magnetic field with an opposite direction to that of the sensing coil thus altering the magnetic flux in the coil and in turn its equivalent inductance.The variations of the sensing-coil equivalent inductance can be induced by changing the magnetic coupling with the conductive target.Fig. 1 shows two configurations adopted to explore the proposed inductive sensor.A proximity sensor has been achieved by measuring the inductance variation of the coil while changing the distance d between the sensing coil and the conductive target, as shown in the schematic representation of Fig. 1(a).By varying d, the coupling between the target and the coil, and therefore the equivalent coil inductance, will change [12].To prevent damages of the sensing coil, a safety clearance ε has been considered.A nondestructive defect detector has been achieved by sliding a conductive target over the sensing coil at a fixed distance d, as shown in the schematic representation of Fig. 1(b).The presence of defects in the target, e.g., scratches or cracks, will vary the inductance of the sensing coil thus allowing their detection.Specifically, in correspondence of defects, such as, scratches, it is expected that the localized rupture or absence of conductive target will modify the coupling between the target and the coil and thus changes the coil inductance [13].By considering a conductive nonferromagnetic target, the coupling between the target and the sensing coil can be modeled in the electrical domain, as shown in Fig. 2. The electrical impedance Z(ω) = V exc (ω)/I C (ω) in the frequency domain is [14] where R = R C + R E and L = L C -L E are the equivalent resistance and inductance of the electrical impedance Z(ω), respectively.R C and L C are the resistance and the inductance of the sensing coil.R E and L E are the resistance and inductance variations due to the eddy current within the conductive target seen from the primary of the transformer, defined as [13] and (3) R T and L T are the equivalent resistance and inductance of the target while the mutual inductance M is and k is the coupling factor between the primary and secondary of the transformer and increases while decreasing the distance between the target and the coil.As it can be observed from (1), the equivalent resistance R and inductance L of the impedance are R C +R E and L C -L E , respectively.Since both R E and L E are positive as described in ( 2) and (3), it can be deduced that R will increase and L will decrease due to the presence of a conductive nonferromagnetic target, while in the absence of the target, R E and L E are zero because the mutual inductance nulls.

III. EXPERIMENTAL RESULTS
Fig. 3(a) shows the top view of the fabricated planar micromachined inductive sensing coil.The sensing coil has been fabricated by exploiting the same platform described in [15] for integrated galvanically isolated transformers.This process is a dedicated platform to fabricate planar inductors and transformers based on gold metal layers, whose thickness can be in the range of 3-10 µm and which are defined by means of electrochemical deposition.Thick gold metals guarantee small resistances and thus high inductor quality factor.
More generally than the scope of the present study, the proposed sensing coil embeds an additional coil buried underneath an insulation layer.Specifically, the two coils are separated by a polyimide layer to provide electrical insulation.To this extent, a polyimide with high dielectric strength (in the range of kV/mm) was chosen and a thickness of tens of µm was achieved between the two coils by means of a multicoating approach, i.e., adding a sequence of polyimide layers through lithography steps [15].In this letter, only the upper coil has been adopted.The buried coil has been left electrically unconnected, i.e., open circuit, to dismiss the effect of currents circulating within it.To experimentally validate the possibility to detect conductive targets by means of a micromachined sensing coil, both configurations described in Section II have been tested by exploiting tailored experimental setups.A brass tip with a diameter of 3 mm has been adopted as conductive target, as shown in Fig. 3(b).
Fig. 3(c) shows the experimental setup adopted to validate the inductive proximity sensor configuration.The fabricated micromachined sensing coil has been bonded on a PLCC68 package for electrical connection to the external circuit board.In this work, the inductance of the sensing coil has been measured by means of an impedance analyzer (HP4194A) by applying a sinusoidal excitation signal v exc (t) with rms amplitude A exc = 1 V and excitation frequency f exc = 10 MHz, as shown in the inset of Fig. 4. The distance d between the coil and tip center has been varied by a positioning stage with a step size of 10 µm within the range 0-1350 µm.To avoid damages to the wire bonding of the sensing coil, a safety clearance ε = 100 µm has been kept.The variation of the equivalent sensing coil inductance L -L 0 has been measured and shown as a function of d in Fig. 4. L 0 = 178.93 nH refers to the inductance measured when the tip target is at ε from the coil center, i.e., the minimum value of L. By reducing d, the coupling between the target and the coil increases, thus reducing the inductance coil as predicted by the theory described in Section II, whereas by increasing d, the inductance asymptotically reaches the value achievable in the absence of target.Fig. 5 shows the obtained inductance variation within the range d = 0-50 µm, i.e., closest to the sensing coil where the inductance variation can be considered linear with respect to d.Within the considered range the sensitivity of the proposed proximity sensor has been estimated by a linear fitting as 6500 nH/m.By performing ten repeated measurements of the inductance for each d step, maximum standard deviations σ L of 3.1 pH have been obtained.The measurement repeatability, which sets the resolution of the inductive proximity sensor, is estimated at one standard deviation σ resulting in 470 nm.Fig. 6(a) shows the experimental setup adopted to validate the inductive defect detector configuration in a static regime.The target has been fabricated by adopting a commercially available single-layer printed circuit board (PCB) with a standard copper layer thickness of 35 µm deposited on an FR4 substrate.As a preliminary test, to emulate the presence of defects three 1-mm wide and 35-µm deep cuts distanced by 19 mm have been included in the target, as shown in Fig. 6(b).The width of the cuts has been set to be compliant with the sensing coil dimensions whose outer diameter is 1.8 mm.The target has been kept at a fixed distance d = 2 mm from the sensing coil center.To detect the defect, the target position x has been slid by 9 mm over the sensing coil by a positioning stage with a step size of 0.5 mm.The variation of the equivalent sensing coil inductance L -L 0 has been measured and shown as a function of x in Fig. L 0 = 172.26nH refers to the inductance measured in the presence of a uniform target.The inductance of the sensing coil has been measured by means of the impedance analyzer by applying the same sinusoidal excitation signal employed in the proximity sensor configuration and by performing four repeated measurements for each step.The localized absence of a conductive target, i.e., the defect, reduces the coupling between the target and the coil, thus increasing the coil inductance.A variation of the inductance L can be measured 4 mm away from the defect and reaches its maximum at the defect position, in consistency with the sensing coil dimensions.

IV. CONCLUSION
An inductive sensor based on a planar micromachined coil for the detection of conductive targets has been presented in this letter.The proposed sensor leverages on the magnetic coupling between the sensing coil, driven by AC current, and a conductive nonferromagnetic target.By changing the position of the target, a variation in the equivalent coil inductance has been measured.Two different configurations have been adopted and experimentally verified.A proximity sensor has been accomplished by measuring the inductance variation at different distances between the coil and a brass tip.A nondestructive defect detector has been succeeded by sliding a copper target over the sensing coil.Experimental results have shown the possibility to detect conductive targets with a resolution of 470 nm within the range 0-50 µm and to detect cuts of 1 mm wide and 35 µm deep.Experimental data have confirmed the possibility to adopt the proposed micromachined coil for the reliable detection of conductive targets.This can pave the way for the development of a miniaturized sensor probe by adopting more than one micromachined coils to detect targets in 2-D or 3-D.

Fig. 1 .
Fig. 1.(a) Schematic representation of the proposed inductive sensor adopted as proximity sensor and (b) defect detector for conductive targets.

Fig. 3 .
Fig. 3. (a) Image of the fabricated micromachined sensing coil.(b) Enlarged view of the tip target.(c) Developed experimental setup for proximity sensing of conductive target.

Fig. 4 .
Fig. 4. Measured inductance variation L -L 0 (blue dots) and linear fitting (dotted red line) as a function of the distance between the sensing coil and the conductive tip target.Schematic representation of the measurement setup (inset).

Fig. 5 .
Fig. 5. Measured inductance variation L-L 0 (blue circles) and linear fitting (dotted red line) as a function of the distance between the sensing coil and the conductive tip target within the range 0-50 µm.

Fig. 6 .
Fig. 6.(a) Developed experimental setup for defect detection in conductive targets.(b) Bottom view of the fabricated target.

Fig. 7 .
Fig. 7. Measured inductance variation L -L 0 as a function of the position x.