Improvement of AC-based Electrical Capacitance Tomography hardware/Kintamu signalu pagristos elektrines talpines tomografijos elektronikos tobulinimas.
Electrical Capacitance Tomography (ECT) is a method to obtain permittivity distribution of mixture by measuring capacitances between electrodes placed around the sensing area. ECT systems (Fig.1) are used for multiphase flow detection in the oil industry , chemistry , agriculture  and in research.
[FIGURE 1 OMITTED]
In the last two decades AC-based ECT techniques have made a big progress in terms of software for image reconstruction [4, 5] and hardware, especially digital electronics parts [6-8]. On the other hand analog electronics has known only few improvements and the main principle remains the same, i.e. the use of charge amplifier  or so called auto-balancing bridge .
[FIGURE 2 OMITTED]
Use of charge amplifier (Fig.2) makes the tomograph stray capacitance immune by allowing voltage excitation and current measurement principle for capacitance measurement . One of the most significant changes in analog part of electronics was the use of three switches to connect excitation voltage to excitation electrode . This change has the main benefit of increasing capacitance measurement range to lower values by eliminating the influence from capacitance of the switches in OFF mode ([C.sub.sw]) which acts in parallel to the measurement capacitances. Meanwhile, the noise created by non-linear capacitance of the switch is also removed.
With the design described above, the input measurement circuit is working properly, even if it encounters some limitations. The first problem concerns the measurement of capacitances which are much smaller than [C.sub.sw]. It has been found that the input circuit has a too large input signal and the output voltage goes into saturation when the electrodes are switched into excitation mode because of the [C.sub.sw] coupling. The capacitance [C.sub.sw] depends on the chosen switch and varies from one to hundreds of pF while capacitances between electrodes in ECT sensors can be less than 10 fF. Non linear distortions appear because of signal saturation, temperature increases  and hardware becomes instable.
Another problem concerns the sensitivity of the input circuit which usually lacks of flexibility. Indeed, the sensitivity of the input circuit depends on the feedback of the charge amplifier. Impedance of tenths of kilo Ohms is used as a feedback to make the input circuit sensitive enough. This generates an impedance of feedback (parallel junction of [R.sub.f] and [C.sub.f] in Fig.2) which is larger than the impedance of switch in OFF mode because of a too large [C.sub.sw]. Therefore, the switch cannot be used normally. Hardware is not flexible and cannot be switched to resistance or conductance mode where the sensitivity of the input circuit should be decreased up to a thousands times.
[FIGURE 3 OMITTED]
The last limitation concerns the standard protocol to calibrate ECT hardware in order to normalize the capacitance measurement. The first calibration is made for the sensing area being empty in order to eliminate the standing capacitances by setting offset values in DAC (Fig.3). The second calibration point is made for the sensing area filled with high permittivity mixture and DC gains (in Fig. 3) are calibrated by setting them as high as possible. Non flexible algorithm of AC gains is used .
Based on above limitations, an improved AC-based ECT hardware is described in this paper. The hardware developed by Yongbo He et al.  was modified and corresponding improvement are described in the three following sections. The first section describes the method to improve switch combination for electrode connection to charge amplifier in order to obtain a more stable hardware. The following section presents the new feedback switching circuit that allows changing sensitivity of input circuit. The last section, before the conclusion, is devoted to the methodology used for the obtention of an additional calibration step for AC gains in order to improve the flexibility of the hardware in order to accommodate any kind of 2D or 3D ECT sensor layouts.
Electronics connection to sensor electrodes
Nowadays, ECT system can handle from 6 to 64 electrodes. Each electrode operates either in excitation or in measurement mode. The mode is changed by connecting the electrode to excitation or measurement circuit using switches. Switch in OFF mode has capacitance [C.sub.sw]. Electronics with switch combination shown in Fig.2 is suitable only for ECT sensors with interelectrode capacitance much higher than [C.sub.sw].
Simulation results have shown that using of one switch [S.sub.m1] for electrode connection to measurement circuit is not enough in some cases. Problem appears when the hardware is used for ECT sensor having capacitance between electrodes much smaller than [C.sub.sw]. In that case the sensitivity of the input circuit should be as high as possible by setting feedback impedance (see parallel combination of [C.sub.f] and [R.sub.f] in Fig.2) as high as possible. However, the increase of feedback impedance is limited not only by the smallest interelectrode impedance but also by the impedance of [S.sub.m1] in OFF mode. The input current in the charge amplifier resulting from the capacitance [C.sub.sw] of [S.sub.m1] may be too large when the impedance of measurement switch in OFF mode is smaller than the feedback impedance while electrode is working in excitation mode. Consequently, the output voltage presents saturation, harmonic distortions and additional harmonics with high frequency forms. This has direct impact on the system instability and serious damage of the ECT hardware can occur because of the sudden rise of components temperature.
[FIGURE 4 OMITTED]
The new circuit shown in Fig.4 works as follows. Switches [S.sub.m1] and [S.sub.m3] are in OFF position and switch [S.sub.s2] is in ON position while electrode is set to excitation mode. In the case when electrode is working in measurement mode, the above switches are set to the opposite mode. The decrease of the output voltage in comparison to the standard circuit shown in Fig.2 is presented in Fig.5. This comparison has been performed using PSpice simulation.
[FIGURE 5 OMITTED]
One can see from the simulation results that the additional switches [S.sub.m2] and [S.sub.m3] (T-switch) allow decreasing significantly the output voltage by minimizing the parasite current flow to input of the charge amplifier. Output voltage and parasite current were minimized 1820 times at the excitation frequency [f.sub.ex] of 500 kHz.
Flexible sensitivity of input circuit
AC-based ECT hardware has input circuit based on charge amplifier (Fig.2.) and sensitivity of input circuit is dependent on its feedback . In the case where it is desired that the tomograph works in other modes than for capacitance measurement, ie. to allow resistance or conductance measurement to be done, there is the need to significantly decrease the sensitivity of the input circuit.
One standard way to achieve such requirement is to use additional amplifier with programmable gain . Another way is to use switching feedback as it is shown in Fig. 6. However, using simple switching feedback can have some drawbacks.
[FIGURE 6 OMITTED]
The PSpice simulation of the switchable feedback circuit shown in Fig. 6 has shown that the sensitivity can be changed to high only for low frequencies. However, standard AC-based ECT hardware uses high frequencies up to 10 MHz  and simple feedback switching is therefore not suitable because of the parasite effect from the [C.sub.sw], coupling. Indeed, switch in OFF position has too low isolation and signal passes through the [C.sub.sw] of [S.sub.f1] which inevitably decreases the sensitivity.
One solution that has been already proposed is to use additional switches  in order to annihilate the coupling effect from the [C.sub.sw]. This is illustrated in Fig.7 with the use of switches [S.sub.f2] and [S.sub.f3].
[FIGURE 7 OMITTED]
The comparison between the input circuits in high sensitivity mode shown in Fig.6 and Fig.7 was again carried out using PSpice simulation presented in Fig.8. The corresponding Figure also presents the case with the input circuit having only one feedback resistance [R.sub.f1].
[FIGURE 8 OMITTED]
The simulation results given in Fig.8 has shown that the simple feedback switching circuit presented in Fig.6 allows switching only up to 100 kHz while the new circuit presented in Fig.7 can be used to switch feedback up to 14 MHz. The suggested feedback switching made hardware able to change sensitivity of input circuit in all the frequencies used in ECT.
Algorithm for flexible calibration of hardware
Standard calibration algorithm is optimized only for special sensor or group of sensors .
Hardware calibration was improved by adding the AC gains calibration step which's block diagram is shown in Fig.9,a. Calibration of AC gains was made before Offset and DC gains calibration and with the same material like for DC gains calibration inside the sensor.
Before measuring signals DC gains were set to 1 and the offsets for standing capacitances were set to 0 as it is presented in the block diagram in Fig.9, b.
[FIGURE 9 OMITTED]
AC gains should be set not larger than the division of maximum ADC value by the measured ADC value as it is shown in the equation below.
ACPGA [less than or equal to] K [ADC.sub.max]/[ADC.sub.value], (1)
here [ADC.sub.max] and [ADC.sub.value] are the--maximal and measured ADC values, respectively; ACPGA is the AC programmable gain, K is a safety coefficient.
It is suggested to use safety coefficient K [approximately equal to] 0,7 for AC gains. It is difficult to fill sensor in some cases  so this coefficient should be used to protect hardware against signal saturation.
Calibration of AC gains by adding additional calibration step allowed compensation of wider range of standing capacitances, allowed to set sensitivity more accurate, made hardware flexible and suitable for any kind of 2D or 3D ECT sensors.
The following conclusion can be drawn concerning the improvement of AC-based ECT hardware:
* Additional switches for electrodes connection to measurement circuits prevent hardware from signal saturation, which is accompanied by a significant drop of the hardware temperature and therefore minimization of power dissipation.
* Suggested feedback switching for charge amplifier makes ECT hardware more flexible and suitable for ERT measurements.
* Additional calibration point allows calibration of AC gains to be made which results in a more flexible system that can be more suitable for any kind of 2D or 3D sensors, especially when the distance between electrodes located in different planes need to be compensated.
The authors would like to thank prof. Erling A. Hammer from University of Bergen for valuable comments.
The work is funded by the European Community's Sixth Framework Programme--Marie Curie Transfer of Knowledge Action (DENIDIA, contract No.: MTKD-CT-2006-039546).
Received 2010 04 06
[1.] Ismail I., Gamio J. C., Bukhari S. F. A., Yang W. Q. Tomography for multi-phase flow measurement in the oil industry // Flow Meas. Instrum.--2005.--16.--P. 145-155.
[2.] Tapp H. S., Peyton A. J., Kemsley E. K., Wilson R. H Chemical engineering applications of electrical process tomography // Sensors and Actuators, B: Chemical.--2003. 92.--P. 17-24.
[3.] Grudzien K., Romanowski A., Chaniecki Z., Niedostatkiewicz M., Sankowski D. Description of the silo flow and bulk solid pulsation detection using ECT // Flow Meas. Instrum.
[4.] Yang W. Q., Peng L. Image reconstruction algorithms for electrical capacitance tomography // Measurement Science and Technology.--2003.--14 p.
[5.] Banasiak R., Soleimani M. Shape based reconstruction of experimental data in 3D electrical capacitance tomography // NDT and E International.
[6.] Huang S., Green R. G., Plaskowski A., Beck M. S. A high frequency stray-immune capacitance transducer based on the charge transfer principle // Instrumentation and Measurement, IEEE Transactions on.--1988.--No. 37.--P. 368-373.
[7.] Yang W. Q., York T. A. New AC-based capacitance tomography system // Science, Measurement and Technology, IEEE Proceedings.--1999.--P.47-53.
[8.] YongBo H., Radoslaw W., Robert B., Meng X., Dominik S. Design of TCP/IP Based Flexible Three Dimensional Electrical Capacitance Tomography, 2009.
[9.] Erling A. Hammer Multi modality tomography systems--State of the art and possible future applications, 2008.
[10.] Dumbrava V., Svilainis L. The Automated Complex Impedance Measurement System // Electronics and Electrical Engineering.--Kaunas, Technologija, 2007.--No. 4(76).--P.59-62.
[11.] Kuroda S. Simple stray-free capacitance meter by using an operational amplifier // IEEE Transactions on Instrumentation and Measurement, 1983.--IM-32.--P.512-513.
[12.] Vamos A. Design of High Temperature Operational Amplifiers // Meres es automatika.--1981.--No. 29.--P. 169-176.
[13.] Sheng-Yu Peng, Qureshi M. S., Hasler P. E., Hall N. A., Degertekin F. L. High SNR capacitive sensing transducer // Proceedings of Circuits and Systems (ISCAS 2006).--2006.--P. 1178-1181.
[14.] Marashdeh Q., Fan L.--., Du B., Warsito W. Electrical Capacitance Tomography--A Perspective // Ind Eng Chem Res, 2008.--No. 47.--P. 3708-3719.
[15.] Im D., Kim H.--., Lee K. A CMOS resistive feedback differential low-noise amplifier with enhanced loop gain for digital TV tuner applications // IEEE Trans. Microwave Theory Tech.--2009.--No. 57.--P. 2633-2642.
[16.] Hadi B., Berruti F., Briens C. New calibration methods for accurate electrical capacitance tomography measurements in particulate-fluid systems // Industrial and Engineering Chemistry Research. No. 2009.--48.--P. 274-280.
D. Styra, L. Babout
Department of Computer Engineering, Technical University of Lodz, Stefanowskiego str. 18/22, 90-924 Lodz, Poland, phone: +48 42 6312750, e-mail: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||ELECTRICAL ENGINEERING/ELEKTROS INZINERIJA|
|Author:||Styra, D.; Babout, L.|
|Publication:||Elektronika ir Elektrotechnika|
|Date:||Sep 1, 2010|
|Previous Article:||Optimization of two-mass electromechanical system/Dvimases elektromechanines sistemos optimizavimas.|
|Next Article:||Influence of colour difference over visibility in multimedia projection/Spalvu skirtumo itaka daugiaterpes projekcijos matomumui.|