Subminiature eddy current transducers for studying metal-dielectric junctions

Based on a transformer eddy current transducer (ECT), a probe has been designed to study metal-dielectric-metal structures. The structural diagram of the probe is given and the basic technical data are stated (the number of windings is 10–130 turns, and the value of the initial permeability of the core μmax = 500). The scheme that uses the computer as a generator and receiver of signals from windings is considered. The measurement procedure allowing one to detect defects in laminate composites with a high accuracy is described. The transducer was tested on the layered structure consisting of paper and aluminum layers with a thickness of 100 μm each in which the model defect was placed. The dependences of the ECT signal on the defect in this structure are given.

A subminiature eddy current transducer [1] has been designed to monitor physical parameters when studying the properties of metal-dielectric junctions. The monitored parameter is the conductance value of the material and its distribution over the surface and in the thickness of the studied object. The eddy current transducer (ECT) is connected to the sound card of the personal computer operating under control of the special software. The software controls the voltage applied to the generator winding of the transducer and reads voltage values from the measuring winding in terms of arbitrary units, which are thereafter, with allowance for the preliminary calibration, transformed into conductance values.
The exciting winding of the subminiature trans ducer consists of 10 turns, and its diameter is 0.13-0.12 mm. The measuring winding consists of 130 turns and has a diameter of 0.05-0.08 mm. To minimize the influence of the exciting winding on the received sig nal, the scheme contains the compensation winding, connected to the measuring winding so that the volt age of the exciting winding is subtracted. It consists of 20 turns. The copper wire with a diameter of 5 µm is used for winding turns. The windings are wound round a pyramid shaped core. The core is made of a 2000 НМ3 ferrite having an initial permeability of 500. The scheme of the subminiature ECT is shown in Fig. 1. The characteristics of the designed transducers allow one to efficiently localize the magnetic field within 2500 µm 2 and ensure a significant depth of its penetra tion into the studied object [2].
The ECT is connected to the sound card of the per sonal computer PC, operated under control of the spe cial software (SW) (Fig. 2). The SW was developed in the C++ language for Windows operational systems.
Using the mixer subsystem of Windows, the SW applies the voltage to the exciting winding of the trans ducer, specifying the level and frequency of the sinuso idal digital signal of the virtual generator.
The digital signal from the virtual generator arrives at the input of the digital to analog converter DAC of the sound card, from the output of which the analog signal is applied through the power amplifier A to the exciting winding (E) of the converter. Being transmit ted through the exiting winding of the ECT, the sinu soidal signal creates the electromagnetic field, which induces the emf in the measuring winding (M) of the ECT. This voltage arrives at the microphone input of the sound card and next through the preamplifier PA at the input of the analog to digital converter (ADC) of the sound card. The analog signal is converted into the digital one and transmitted to the processing and con trol unit of the SW. The processing and control unit records the digital signal level in arbitrary units corre sponding to voltage values at the measuring winding.
This level is assumed to be the zero level corre sponding to the voltage level at the measuring winding without the monitored object; i.e., in the absence of the monitored object, the indicator shows zero corre sponding to the zero conductance value.
The use of the computer sound card enables one, while scanning, to vary the frequency of the electro magnetic field, created by the exciting winding of the converter, from 20 Hz to 2 kHz.
The designed ECT allows one to efficiently study metal-dielectric junctions in miniature laminate metal-polymeric composite objects. Similar compos ites can contain several metallic layers separated by thin polymer dielectric interlayers [2]. Typical defects Abstract-Based on a transformer eddy current transducer (ECT), a probe has been designed to study metal-dielectric-metal structures. The structural diagram of the probe is given and the basic technical data are stated (the number of windings is 10-130 turns, and the value of the initial permeability of the core µ max = 500). The scheme that uses the computer as a generator and receiver of signals from windings is con sidered. The measurement procedure allowing one to detect defects in laminate composites with a high accu racy is described. The transducer was tested on the layered structure consisting of paper and aluminum layers with a thickness of 100 µm each in which the model defect was placed. of such materials include, e.g., discontinuities of layers and bridging between layers. The earlier designed ИЭНМ 5ФА device [3] (conductance meter for non ferromagnetic materials) was used to study metaldielectric-metal layer structures, and a specially mod ified Fourier analyzer was used to measure ampli tude-frequency characteristics. The structure composed of the alternating 100 µm thick aluminum foil and the paper also having the 100 µm thickness was used to demonstrate the applicability of the proposed method. As a model defect, a hollow parallelepiped with 300 µm thick walls was placed between layers. Figure 3 shows the picture observed when the probe moves above the layered medium inside which a defect is located. The signal level from the measuring winding characterizes conductance values on the studied area. For the 1000 Hz basic operating frequency, the volt age introduced into the measuring winding was 130 ± 2 mV. Domains 1 and 2 in the graph, in which the volt age level drops to 115 mV, correspond to the parallel epiped walls. This change in the signal amplitude is 11% from the signal level corresponding to the defect free area of the sample. In this case, oscillations of the signal amplitude on the defect free area do not exceed 4 mV, being 3% of the signal level corresponding to the defect free area of the sample.
The amplitude changes during the tests of the transducer at other frequencies are well noticeable on the graphs shown in Fig. 4. As the frequency increases, these changes are caused by a smaller depth of pene tration of eddy currents into the layered structure and increased influence of various small cracks on the sur face of the layered structure. As the frequency decreases, the field of eddy currents more deeply pen etrates into the studied object. In this case, the influ ence of the model defect is not observed.   When the measurement frequency is equal to 6000 Hz (Fig. 4a), the model defect is still well notice able. However, the amplitude oscillations on the defect free area of the sample already exceed 7% of the signal level corresponding to the defect free part of the sample. When the monitored object with unknown defects is studied, similar changes in the amplitude can erroneously be interpreted as defects. At the prob ing frequency of 500 Hz (Fig. 4b), the oscillations of the amplitude on the defect free area are insignificant. However, the change in the amplitude on the area of the defect itself does not exceed 3% of the signal level corresponding to the defect free part of the sample. In laboratory and production measurements, a similar amplitude oscillation can be caused by external actions, which are not associated with the presence of defects.
When the operating frequency of the device goes beyond the indicated limits, results of the measure ments will be distorted by amplitude oscillations caused by microcracks on the surface of the sample or by a decrease in the field localization inside the layered structure. Figure 5 shows the measurement results at a frequency of 7000 Hz. As can be seen from the graph, the changes in the amplitude caused, in this case, by microcracks on the sample surface are significantly higher than the changes in the amplitude caused directly by the defect.
The defect was located at a distance of 600 µm from the probe in the heart of the layered structure. Up to the defect occurrence depth equal to 1400 µm, the explicit dependence of the transducer response on the position of the transducer above the defect was observed. By fixing the change in the transducer response amplitude caused by the defect, it is possible to change the frequency of the current in the exciting winding so that the eddy currents are concentrated in composite layers located above the defect. The solu tion to the inverse problem allows one to determine the defect occurrence depth. Upon the calibration of the Fourier analyzer against typical defects, it is possi ble to use the ИЭНМ 5ФА device for diagnosing composite multilayer materials with a thickness of 1-1400 µm.