Overview of The System Model, Theory, Design and Approaches
Figure 2.11 WFG architectures based on (a) three CMOS OTAs [77], (b) using three
2.9 Active Electrode
The next part of the system is the electrode. This electrode is required for the communication network between the electro-medicine device and the body, by using the conductivity of the body. Typically, these electrodes, the common transducers, are simple surface electrodes and use a passive metallic disc with wire connected to the bio- signal amplifier. The transducers may be made of stainless steel, gold or Ag/AgCl; the latter is the most used for DC derivations with low frequency signals [167]. In fact, conventional passive electrodes have a significant effect on the purity of the signal transmission. This is because the electrode wire is the main source of noise, including power line interference and the movement of the wire. Another drawback of using a passive electrode is that short circuits may happen between adjacent electrodes [168] increasing the noise generated at the metal-skin interface, which degrades the source signal quality. Passive electrodes also require unusually extensive skin preparation to improve the skin contact and connectivity, which is not practical for daily use [169]. A good quality signal can be transmitted by using an active electrode utilized buffer amplifier, as close to the transducers as possible, for shielding, and it eases mains
68
interference [167, 168, 170]. The active electrode has system connectors to supply the electronic components with power. Figure 2.18 shows different types of electrodes. In the 1960’s and 1970’s, the concept of active electrode was developed for the first time in [171, 172] respectively.
Figure 2.18: Types of electrodes: (a) passive electrodes, and (b), active electrode using commercial Op-Amps with Ag/AgCl transducer for biomedical applications [173].
In the 1960’s such an electrode had not been used, due to the large size of the IC transistor which could not be mounted on the electrode, and also due to their high cost [168]. An early active electrode was demonstrated in 1971 when [175, 176] developed an active electrode for electrocardiogram (ECG) recording, realized from
buffer amplifiers, constructed of discrete components within a metal case, connected to ground to provide shielding of the circuitry.Multielectrode probes that consisted of gold electrodes incorporated onto a silicon chip for the membrane potential recording were developed in [177-179]. In 2004 [180] employed the active electrode in a medical
acupuncture system to study the skin response to laser stimulation effect. Presently, several types of active electrode circuit have been developed for bio-signal sensing, measurement and stimulation. Examples are electro-oculography (EOG), electro- encephalography (EEG), electro-cardiogram (ECG) [167, 181, 182] electrical impedance tomography (EIT), for monitoring patient heart activities and respiration system [169, 183] and for bio-impedance measurement [184]. The main feature realized in active electrodes is signal buffering. By utilizing a buffer amplifier circuit, one can develop active electrodes for bioelectrical stimulation. A simple buffer can be employed
69
by utilizing an Op-Amp in a voltage follower configuration. A variety of active electrodes has been reported in the literature, usually, using commercial Op-Amps, for example, NJMO62M Op-Amp [168], LMC7111 Op-Amp [181], TLC272 Op-Amps [182] and LMP7701 Op-Amp [185]. Some authors [186] employed buffer circuits with class-AB output stage in a neuronal stimulation and recording application. The buffer circuit was fabricated using 0.6μm CMOS technology which can generate output current of about 10mA, using 5V supply voltage with power consumption of 150μm when no input signal is applied. Some other authors [187] developed a compact (= 0.02mm2) buffer for both voltage and current mode electro-bio-stimulation. This is a high-current class-AB voltage follower, with the OTA component based on a local common mode feedback (LCMFB) amplifier. It was fabricated with a 0.6μm CMOS
process using 5V supply voltage and power consumption of 245μW for the voltage mode, while in current mode the circuit provided output current of 300μA or 30μA with input resistance of 10kΩ or 100kΩ respectively. The design in [174] used two-stage OTAs employing PMOS input stages with large gate areas, to improve noise performance, and configured a unity-gain-buffer for simultaneous acquisition of EEG and EIT signals. It was fabricated using AMS 0.35μm technology, and power consumption in the readout mode was about 1mW. In [188, 189] a super buffer realizing a class-AB amplifier with large transistor count (22 and 20 respectively) was designed to provide large output current. It is implemented using 45nm CMOS technology with +3V supply voltage. The work reported in [190] used two-stage OTAs (transistor count 19) to design a wideband current driver circuit for cancer tissue impedance analysis. It was fabricated using a 0.35μm CMOS technology with ±2.5V supply voltage.
The aim of the above approaches was to design compact high performance circuit for the signal buffering task at low power drain, and to transmit a superior quality signal by shielding mains interference. In addition to the buffering task, a simple buffer architecture with the desired attributes that has a small transistor count while achieving large driving capability at low power consumption is attractive for a biomedical device. The proposed active electrode is based on the [186] architecture that utilized a class-AB amplifier to deliver a large output current of 10mA, which is too high for the applications of this research project. In addition, the circuit is based on 0.6μm CMOS technology using high supply voltage of 5V. Therefore, the buffer circuit of this work is developed using new 130-nm CMOS technology with low supply voltage of only 1.2V
70
and occupies a small number of MOSFETs. Since the output current of the typical OTA circuit is limited by the bias current, using a large bias current increases the power requirement. Thus, a trade-off between power consumption and slew rate exists [66]. To overcome this deficiency, an adaptive basic current scheme realizes three MOSFETs employed to release boosting of the bias current, by providing a scaled copy of the input stage differential current back to the bias current, instead of using a class-AB amplifier with four MOSFETs. Therefore, the buffer circuit can achieve an appropriate driving capacity with low power consumption, which is compatible with the requirements of the biomedical applications of this work.