Open Access

Rectifying memristor bridge circuit realized with human skin

Oliver Pabst
Oliver Pabst
   | Dec 31, 2018
Journal of Electrical Bioimpedance's Cover Image
About this article
Previous Article
Next Article
Cite
Published Online: Dec 31, 2018
Page range: 184 - 192
Received: Dec 22, 2018
DOI: https://doi.org/10.2478/joeb-2018-0023
Keywords
© 2018 O. Pabst published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Fig. 1

Memristor bridge circuit and its implementation on human skin (a) Schematic of the four memristor bridge circuit similar to the ones presented in [15-17] with voltage source vCC and measured voltage vM. (b) The two memristor version of the memristor bridge is realized by using two voltage sources with opposite sign as illustrated in the schematic (see also [17]). (c) Memristor bridge circuit realized with human skin. Schematic of the instrumentation (top) and the corresponding electrode placement (bottom) is shown for the left-hand side. The electrode setup on the right-hand side was equivalent. Voltages vCC1 and vCC2 were applied at the CC1 and CC2 electrodes, respectively. The CC1 electrode was attached to the earlobe (variant A, chosen for 12 out of 28 subjects) or to the forehead (variant B, chosen for 16 out of 28 subjects). The CC2 electrode was always placed at the forehead. All electrodes were put in place right after each other. The stratum corneum memristor and the sweat duct memristor (under each electrode) are electrically in parallel to each other and can be modeled as one overall memristor due to the closure theorem [1]. The greyed memristor symbol under the M electrode shall illustrate that the influence of the corresponding skin to the measurement is negligible. The direction of the voltage here is from skin surface (under CC1 and CC2) to deeper skin layers while it was from deeper skin layers to skin surface in the setup used in [22]. The same photograph that illustrates the electrode placement at the earlobe has been presented in [22] under Creative Commons Attribution 4.0 International License.
Memristor bridge circuit and its implementation on human skin (a) Schematic of the four memristor bridge circuit similar to the ones presented in [15-17] with voltage source vCC and measured voltage vM. (b) The two memristor version of the memristor bridge is realized by using two voltage sources with opposite sign as illustrated in the schematic (see also [17]). (c) Memristor bridge circuit realized with human skin. Schematic of the instrumentation (top) and the corresponding electrode placement (bottom) is shown for the left-hand side. The electrode setup on the right-hand side was equivalent. Voltages vCC1 and vCC2 were applied at the CC1 and CC2 electrodes, respectively. The CC1 electrode was attached to the earlobe (variant A, chosen for 12 out of 28 subjects) or to the forehead (variant B, chosen for 16 out of 28 subjects). The CC2 electrode was always placed at the forehead. All electrodes were put in place right after each other. The stratum corneum memristor and the sweat duct memristor (under each electrode) are electrically in parallel to each other and can be modeled as one overall memristor due to the closure theorem [1]. The greyed memristor symbol under the M electrode shall illustrate that the influence of the corresponding skin to the measurement is negligible. The direction of the voltage here is from skin surface (under CC1 and CC2) to deeper skin layers while it was from deeper skin layers to skin surface in the setup used in [22]. The same photograph that illustrates the electrode placement at the earlobe has been presented in [22] under Creative Commons Attribution 4.0 International License.

Fig. 2

Results from several subjects. Applied sinusoidal voltage, vCC1, (on CC1) and measured current, i, over time. The sign of the amplitude of vCC1 was either 1 V (subject A, for example) or -1 V (subject B, for example) due to randomization and the sign of vCC2 was always opposite. In the top and middle line, the results of the recordings with a voltage frequency of 0.005 Hz are shown. Subject labelling is in accordance with the results presented in [22]. The applied voltages themselves affect the memductance of the skin and consequently the resulting current (non-linear electrical measurement). The memductance changes during the recording and can be different at the beginning of the second period which explains also the differences in the recorded current from period to period. (a) Subject A, and V at 0.005 Hz and Subject L at 0.5 Hz. The frequency of the measure is double the frequency of the applied voltages. The measured currents of a total of five subjects (2 out of 12 with the CC1 electrode at the earlobe, 3 out of 16 with the CC1 electrode at the forehead) are comparable when the applied voltage frequency is 0.005 Hz. Two of these subjects show similar results at 0.05 Hz, and one of these two even shows a similar result at 0.5 Hz (subject L). (b) Subject D and B at 0.005 Hz and subject B at 0.05 Hz. The measured current has a large magnitude in one half of the period and is more or less cut off during the other half of the period. Six subjects in total (3 out of 12 with the CC1 electrode at earlobe, 3 out of 16 with the CC1 electrode at the forehead) show a similar behavior at a frequency of 0.005 Hz and five out of them show also (more or less) similar behavior at 0.05 Hz. (c) Subject E and F. The measured currents are non-linear, but neither half-wave rectification nor frequency doubling is observed. Seventeen subjects in total show similar behavior at 0.005 Hz.
Results from several subjects. Applied sinusoidal voltage, vCC1, (on CC1) and measured current, i, over time. The sign of the amplitude of vCC1 was either 1 V (subject A, for example) or -1 V (subject B, for example) due to randomization and the sign of vCC2 was always opposite. In the top and middle line, the results of the recordings with a voltage frequency of 0.005 Hz are shown. Subject labelling is in accordance with the results presented in [22]. The applied voltages themselves affect the memductance of the skin and consequently the resulting current (non-linear electrical measurement). The memductance changes during the recording and can be different at the beginning of the second period which explains also the differences in the recorded current from period to period. (a) Subject A, and V at 0.005 Hz and Subject L at 0.5 Hz. The frequency of the measure is double the frequency of the applied voltages. The measured currents of a total of five subjects (2 out of 12 with the CC1 electrode at the earlobe, 3 out of 16 with the CC1 electrode at the forehead) are comparable when the applied voltage frequency is 0.005 Hz. Two of these subjects show similar results at 0.05 Hz, and one of these two even shows a similar result at 0.5 Hz (subject L). (b) Subject D and B at 0.005 Hz and subject B at 0.05 Hz. The measured current has a large magnitude in one half of the period and is more or less cut off during the other half of the period. Six subjects in total (3 out of 12 with the CC1 electrode at earlobe, 3 out of 16 with the CC1 electrode at the forehead) show a similar behavior at a frequency of 0.005 Hz and five out of them show also (more or less) similar behavior at 0.05 Hz. (c) Subject E and F. The measured currents are non-linear, but neither half-wave rectification nor frequency doubling is observed. Seventeen subjects in total show similar behavior at 0.005 Hz.

Fig. 3

Boxplots reflecting the percentage of rectification of the current, i, normed with the average rectified value of a sinusoidal signal (rrec, equation (7)). The boxplots are based on the evaluation of all subjects (N=28). These results are obtained from the recordings with different voltage frequencies (0.005 Hz, 0.05 Hz, and 0.5 Hz) and always shown for the first period (out of two). The horizontal line in the middle of each boxplot denotes the median; the circle indicates the mean value; and the whiskers indicate the 5% and 95% percentiles. The upper percentile is below 20% at 0.5 Hz which is in accordance with the observation that most subjects exhibited non-rectified currents at this frequency.
Boxplots reflecting the percentage of rectification of the current, i, normed with the average rectified value of a sinusoidal signal (rrec, equation (7)). The boxplots are based on the evaluation of all subjects (N=28). These results are obtained from the recordings with different voltage frequencies (0.005 Hz, 0.05 Hz, and 0.5 Hz) and always shown for the first period (out of two). The horizontal line in the middle of each boxplot denotes the median; the circle indicates the mean value; and the whiskers indicate the 5% and 95% percentiles. The upper percentile is below 20% at 0.5 Hz which is in accordance with the observation that most subjects exhibited non-rectified currents at this frequency.

Fig. S1

Alternating current (AC) voltage-current plot of subject B at the earlobe, shown for the third period of applied sinusoidal voltage with frequency of 0.05 Hz and amplitude of 1.2 V. These data (not shown before) are obtained from the experiment presented in [22]. The two pinched points are indication that the stratum corneum memristor (in parallel with the capacitive properties of the stratum corneum) is dominating the measurement and that the galvanic contact through the sweat ducts was not given (see [22]).
Alternating current (AC) voltage-current plot of subject B at the earlobe, shown for the third period of applied sinusoidal voltage with frequency of 0.05 Hz and amplitude of 1.2 V. These data (not shown before) are obtained from the experiment presented in [22]. The two pinched points are indication that the stratum corneum memristor (in parallel with the capacitive properties of the stratum corneum) is dominating the measurement and that the galvanic contact through the sweat ducts was not given (see [22]).

Fig. S2

Results from several subjects (voltage current plots). These are the results of the two-memristor bridge circuit shown for the same subjects that are presented in Fig. 2. Here the measured current, i, is plotted over the applied sinusoidal voltage vCC1 (on CC1); the sign of the applied sinusoidal voltage vCC2 under CC2 was opposite. Even though pinched hysteresis loops may cross the second and fourth quadrant in this presentation (plotted with regard to vCC1, but not vCC2) it does not mean that the corresponding memristor circuit is active. The plot of subject E, for example, just means that the skin under CC2 is dominating the measurement. The results are shown over two periods (the grey dotted plot presents the first period and the blue solid plot presents the second period). In the top and middle line, the results of the recordings with a voltage frequency of 0.005 Hz are shown. (a) Subject A, V at 0.005 Hz and Subject L at 0.5 Hz; all exhibited (more or less) full-wave rectified currents in Fig. 2. The corresponding V-I plots show pinched hysteresis loops with large lobe areas in the third and fourth quadrant but relative small areas in the first and second. (b) Subject D and B at 0.005 Hz and subject B at 0.05 Hz; all exhibited more or less half wave rectified currents in Fig. 2. The corresponding V-I plots show pinched hysteresis loops with relative large lobe area in either the third or fourth quadrant. Within the other quadrant (either fourth or third quadrant, respectively), the corresponding branch of the loop does not span a large area but is close to the x-axis instead. (c) Subject E and F. Since one skin site (under CC1 or CC2) dominates each measurement, the V-I plots reflect the pinched hysteresis loops of single skin memristors.
Results from several subjects (voltage current plots). These are the results of the two-memristor bridge circuit shown for the same subjects that are presented in Fig. 2. Here the measured current, i, is plotted over the applied sinusoidal voltage vCC1 (on CC1); the sign of the applied sinusoidal voltage vCC2 under CC2 was opposite. Even though pinched hysteresis loops may cross the second and fourth quadrant in this presentation (plotted with regard to vCC1, but not vCC2) it does not mean that the corresponding memristor circuit is active. The plot of subject E, for example, just means that the skin under CC2 is dominating the measurement. The results are shown over two periods (the grey dotted plot presents the first period and the blue solid plot presents the second period). In the top and middle line, the results of the recordings with a voltage frequency of 0.005 Hz are shown. (a) Subject A, V at 0.005 Hz and Subject L at 0.5 Hz; all exhibited (more or less) full-wave rectified currents in Fig. 2. The corresponding V-I plots show pinched hysteresis loops with large lobe areas in the third and fourth quadrant but relative small areas in the first and second. (b) Subject D and B at 0.005 Hz and subject B at 0.05 Hz; all exhibited more or less half wave rectified currents in Fig. 2. The corresponding V-I plots show pinched hysteresis loops with relative large lobe area in either the third or fourth quadrant. Within the other quadrant (either fourth or third quadrant, respectively), the corresponding branch of the loop does not span a large area but is close to the x-axis instead. (c) Subject E and F. Since one skin site (under CC1 or CC2) dominates each measurement, the V-I plots reflect the pinched hysteresis loops of single skin memristors.
eISSN:
1891-5469
Language:
English
Publication timeframe:
Volume Open
Journal Subjects:
Engineering, Bioengineering and Biomedical Engineering, Biomedical Electronics, Life Sciences, Biophysics, Medicine, Biomedical Engineering, Physics, Spectroscopy and Metrology
Journal RSS Feed