Design of a charge amplifier for a low‐power respiration‐monitoring system
2019; Institution of Engineering and Technology; Volume: 13; Issue: 4 Linguagem: Inglês
10.1049/iet-cds.2018.5369
ISSN1751-8598
AutoresIfana Mahbub, Samira Shamsir, Salvatore A. Pullano, Antonino S. Fiorillo, Syed K. Islam,
Tópico(s)Advanced Sensor and Energy Harvesting Materials
ResumoIET Circuits, Devices & SystemsVolume 13, Issue 4 p. 499-503 Research ArticleFree Access Design of a charge amplifier for a low-power respiration-monitoring system Ifana Mahbub, Corresponding Author Ifana Mahbub Ifana.Mahbub@unt.edu orcid.org/0000-0002-5561-0880 Department of Electrical Engineering, University of North Texas, Denton, TX, 76203 USASearch for more papers by this authorSamira Shamsir, Samira Shamsir orcid.org/0000-0003-1311-8270 Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, MO, 65211 USASearch for more papers by this authorSalvatore A. Pullano, Salvatore A. Pullano orcid.org/0000-0003-3391-1698 Department of Health Sciences, University Magna Graecia of Catanzaro, Catanzaro, 88100 ItalySearch for more papers by this authorAntonino S. Fiorillo, Antonino S. Fiorillo Department of Health Sciences, University Magna Graecia of Catanzaro, Catanzaro, 88100 ItalySearch for more papers by this authorSyed K. Islam, Syed K. Islam Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, MO, 65211 USASearch for more papers by this author Ifana Mahbub, Corresponding Author Ifana Mahbub Ifana.Mahbub@unt.edu orcid.org/0000-0002-5561-0880 Department of Electrical Engineering, University of North Texas, Denton, TX, 76203 USASearch for more papers by this authorSamira Shamsir, Samira Shamsir orcid.org/0000-0003-1311-8270 Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, MO, 65211 USASearch for more papers by this authorSalvatore A. Pullano, Salvatore A. Pullano orcid.org/0000-0003-3391-1698 Department of Health Sciences, University Magna Graecia of Catanzaro, Catanzaro, 88100 ItalySearch for more papers by this authorAntonino S. Fiorillo, Antonino S. Fiorillo Department of Health Sciences, University Magna Graecia of Catanzaro, Catanzaro, 88100 ItalySearch for more papers by this authorSyed K. Islam, Syed K. Islam Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, MO, 65211 USASearch for more papers by this author First published: 30 May 2019 https://doi.org/10.1049/iet-cds.2018.5369Citations: 2AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract Home-based health care applications are rapidly gaining popularity, enabling a renewed focus on the design of low-power and low-noise front-end circuitry. In this context, the evaluation of low-frequency biomedical signals, such as the respiration pattern, benefits from the design of a front-end amplifier with reduced power consumption and low noise. Continuous efforts on improving the performances of respiration-monitoring devices have resulted in the reduction of noise and motion artefacts by increasing complexity (e.g. complex algorithms or high precision filtering) at the expense of increased power consumption. This study is focused on the design of a fully integrated charge amplifier for respiration monitoring based on a pyroelectric sensor. Simulation and test results show a power consumption of 1.8 µW, an active die area of 0.085 mm2, a bandwidth in the range from 10 mHz to 13 kHz, and a remarkable noise efficiency factor of ∼2.79, which fits well with the development of an energy efficient wearable device. 1 Introduction Non-invasive respiration monitoring has recently become very prominent for the detection of pathological conditions such as tachypnea (high respiration rate), bradypnea (low respiration rate), and apnea (absence of breathing). Continuous respiration monitoring avoiding adhesive electrodes on the skin, especially for preterm infants, is desirable to reduce long-term collateral effects (e.g. anetoderma). Although various non-contact-based respiration-monitoring systems have been reported in the literature, these schemes usually require power-hungry and expensive mechanisms for noise cancellation and filtering to obtain acceptable accuracy in detecting the respiration rate and amplitude due to breathing [1]. Pyroelectric sensor-based systems have been proposed as a potential solution that is suitable for preterm infants [2-4]. The sensor is made of polyvinylidene fluoride (PVDF) and is integrated inside a 3D-printed nasal cannula rather than having any electrode coming in contact with the body for both diagnosis and therapy. The electric charge generated by the PVDF transducer needs to be conditioned and converted into a proportional voltage signal by a charge amplifier. As the frequency of the breathing signal usually ranges from 0.1 to 1 Hz, the amplifier design has to be optimised for low-frequency signals [5, 6]. It also needs to achieve a low input-referred noise to be able to detect an even smaller breathing signal. Different designs of low-power, low-noise front-end amplifiers for biomedical sensors have been investigated by various research groups. One such topology is proposed by Denison , et al. that implements a chopper-stabilised amplifier [7]. In order to achieve a low input-referred noise, this type of amplifier requires more area, thus increasing the design complexity. A single-ended topology for the design of a low-power low-noise amplifier was proposed by Harrison , et al. [8]. Kmon , et al. presented a topology for minimising the input-referred noise along with the offset voltage correction [9]. However, not many front-end amplifiers reported in the literature can perform efficient charge conversion while achieving a very low cut-off frequency, which is particularly important for respiration-monitoring applications. Thus, the design of a low-power, low-noise front-end amplifier for the respiration signal with a frequency as low as 0.1 Hz is essential for the successful realisation of portable and wearable respiration-monitoring systems. Previously, a front-end charge amplifier designed in 0.5 µm CMOS process has been presented [10]. However, better power consumption and noise efficiency can be achieved by designing the amplifier in 130 nm complementary metal-oxide-semiconductor (CMOS) process [11]. This paper presents an in-depth design of the front-end amplifier circuit in 130 nm CMOS process optimising it specifically for the ultra-low-frequency respiration-monitoring application. The paper is organised as follows. Section 2 describes the electrical characterisation of the PVDF transducer to evaluate the amplifier design requirements and Section 3 describes the design of the charge amplifier with simulation and measurement results. This is then followed by a conclusion in Section 4. 2 Electrical characterisation of the PVDF transducer The total amount of charge on the surface of the PVDF transducer, placed inside the nasal cannula or right under the nostrils with a patch, primarily depends on the net heat flow experienced by the PVDF sensor resulting from the variation in external room temperature, the skin temperature of the patient, and the temperature of the exhaled air. If the breathing stops for a long period of time, the transducer approaches the thermal equilibrium depending on the external room temperature as well as the skin temperature. During breathing, the temperature of the sensor cyclically varies between a minimum and a maximum which mainly depends on the environmental (room temperature) and the exhaled air temperatures. The temperature of the air coming out of the nostrils roughly varies from 22 to 32°C [12]. The pyroelectric charge at the surface of the PVDF sensor can be described as (1) where Qp (t) is the charge generated by the PVDF, pQ is its pyroelectric charge coefficient, S is the surface area of the transducer, and Tp(t) is the difference between the temperatures of the upper and the lower surfaces of the transducer. The electric charge is stored across the intrinsic input capacitance, Cin, of the transducer and generates the input voltage, Vin, as (2) Using (1) and (2) for a 10 mm × 2 mm PVDF sensor with Cin of 2 pF, the minimum and the maximum generated voltage signals for a temperature variation of as low as 0.1°C to as high as 5°C can be calculated to be 24.36 µV and 860 mV, respectively. Moreover, since the voltage output for 5°C temperature variation is quite high and the maximum allowable supply voltage for the 130 nm CMOS process is 1.3 V, the closed-loop gain of the amplifier is limited to ∼6 dB. 3 Design of the front-end charge amplifier The lower cut-off frequency of the closed-loop amplifier needs to be in the sub-Hz range in order to be suitable for the respiration-monitoring application. Among different circuit topologies, a folded-cascode operational transconductance amplifier (OTA) is chosen for this work due to its simplified architecture. Furthermore, the noise and power trade-off for this system have been obtained by current splitting with the output branch and the current scaling technique [13]. To obtain a very low cut-off frequency (sub-Hz), pseudo-resistor-based feedback configuration is implemented that can reduce the active die area significantly. 3.1 Front-end amplifier design considerations The OTA, as shown in Fig. 1, consists of a differential input pair, M1–M2. The width/length (W/L) ratio and the operating condition of the transistors are summarised in Table 1. The inversion coefficient (IC) can be calculated using the ID/IS ratio, where ID is the drain current and IS is the inversion current of the transistor. If IC >0.1 but <10, the transistor is operating in the moderate inversion region. If the IC <0.1, then the transistor is operating at the weak inversion region. In addition, VEFF is the minimum drain-to-source voltage that is needed to make the transistor operate in the saturation region. The relatively wide transistors allow the devices to operate in the weak inversion region to maximise the transconductance (gm) and hence the gain. In the weak inversion region, the transistor has an exponential characteristic and acts like a bipolar junction transistor. The overall transconductance (GM) of the folded-cascode amplifier can be expressed as (3) where ID is the drain current of the transistor, VT = kT/q is the thermal voltage, and n is the slope factor (n ∼ 1.2 in the weak inversion region). The transconductance efficiency, gm/ID, is ∼34.31 V−1. All the current mirror transistors (M13−M14−M0, M15−M16, and M12−M5−M6) are designed to operate in the moderate inversion region to reduce the input-referred noise and increase the output resistance. The transconductance, gmCM of the current mirror transistors is expressed as (4) where K = 1/2µCOXW/L. The transistors in the folded branch (M3–M4) are biased in the strong inversion region to reduce their noise contribution. The bias current, Ib, is 240 nA, which makes the total power consumption of the amplifier ∼1.8 µW. The unity gain frequency, f0, or the gain bandwidth (GBW) is expressed as (5) The dominating pole frequency is (6) where RL and CL are the load resistance and capacitance, respectively. The maximum gain of the amplifier (without any load resistance) depends on the equivalent output resistance, R0, which is equal to the parallel combination of the two resistances at the output stage and can be expressed as (7) Then the mid-band open-loop gain becomes (8) where µ is the mobility and rds is the drain-to-source resistor of the transistor. The schematic of the front-end charge amplifier is shown in Fig. 2. The OTA block amplifies the breathing signal acquired by the PVDF sensor. A 2 pF capacitor, C1, and a 1 pF capacitor, C2, are used to set the gain of the closed-loop amplifier. A mid-rail voltage, Vmid, of ∼600 mV is used to bias the output signal to the mid-rail voltage for the supply voltage of 1.2 V. The load capacitance, CL, is chosen to be 1 pF. The design equations for the mid-band gain, A0, the lower and upper cut-off frequencies, fL and fH, respectively, can be expressed as [10] (9) (10) (11) Fig. 1Open in figure viewerPowerPoint Schematic of NMOS input folded-cascode amplifier Table 1. Characteristics of the NMOS input folded-cascode transistors Transistor W/L, µm/µm ID, µA IC gm/ID V−1 |VEFF| = |VGS−Vth|, mV M1, M2 10/0.5 0.24 0.008 34.31 105.2 M3, M4, M11 4/1 0.5 0.825 21.13 8.76 M5, M6 10/0.5 0.27 0.2 25.9 102.6 M9, M10 120/0.6 0.27 0.06 31.71 27.5 M7, M8 2/1 0.27 1.32 24.9 56.5 M0 2/1 0.48 0.88 28.02 25.3 M13, M14 2/1 0.5 0.52 27.88 25.3 M15, M16 5/1 0.5 0.44 30.38 8.2 M12 10/0.5 0.5 0.3 24.57 71.4 Fig. 2Open in figure viewerPowerPoint Schematic of the folded-cascode OTA with triple-well NMOS pseudo-resistor For implementing the pseudo-resistors (M1–M2) double-gated n-type triple-well metal-oxide-semiconductor field-effect transistors with a W/L ratio of 0.36/0.24 μm are chosen. To achieve an ultra-low cut-off frequency without using an enormous resistor or capacitor in the feedback loop and thus minimising the chip area, diode-connected n-type metal-oxide-semiconductor (NMOS) transistors M1 and M2 are used to emulate large resistances of up to hundreds of GΩ. These transistors operate in the subthreshold region at a very low VDS. It is similar to the pseudo-resistor proposed by Holleman and Otis [14] except for the use of two triple-well NMOS transistors instead of single-well transistors, which is the novelty of this work. To reduce the noise contribution, the substrate of the transistor is isolated by implementing the triple-well pseudo-resistor configuration. The cross-sectional view of a triple-well NMOS transistor is illustrated in Fig. 3, which shows that the p-well is embedded within a deep n-well creating an isolated body for the NMOS from the p-substrate. Fig. 3Open in figure viewerPowerPoint Cross-section of an NMOS transistor in a triple-well 130 nm CMOS process 3.2 Simulation and experimental results As the voltage across the pseudo-resistor changes in small steps, the equivalent resistance increases up to hundreds of GΩ. Fig. 4a shows the difference in current (ΔI) due to the small incremental voltage change (ΔV). Fig. 4b shows the incremental resistance values of the pseudo-resistors for different ΔV values. As shown in Fig. 4(A), the incremental current change is in the fA range for −1.2 V < ΔV < 0.6 V range and the resistance value is ∼0.5 TΩ. The resistance value is almost constant over this range. The open-loop mid-band gain of the OTA is ∼58.18 dB, which is depicted in the simulation result as shown in Fig. 5. With the load capacitance CL of 1 pF, the simulated unity GBW and the phase margin are 5.5 MHz and 82.56°, respectively. The simulated input offset voltage of the OTA is 134 μV. Simulation and measurement results of the closed-loop gain of the OTA with feedback are shown in Fig. 6. From the plot, it can be observed that the simulated mid-band gain is 6 dB and the lower (fL) and the upper (fH) corner frequencies are 11.2 mHz and 34.7 kHz, respectively, with 10 pF load capacitance. Fig. 4Open in figure viewerPowerPoint Simulation results of the proposed pseudo-resistor(a) Simulation result of the current versus incremental voltage across the pseudo-resistor and, (b) Equivalent resistance of the pseudo-resistor for different incremental voltages Fig. 5Open in figure viewerPowerPoint Simulation results of the open-loop gain (black) and the phase (blue) plots of the OTA in Fig. 1 Fig. 6Open in figure viewerPowerPoint Simulation (black) and measurement (red) results of the closed-loop gain of the front-end amplifier in Fig. 2 The experimentally measured mid-band gain is 6.23 dB with fL being <10 mHz and fH ∼13 kHz. The reason for the measured gain to be greater than the simulation result can be attributed to the mismatch of the input capacitor C1. With the addition of the parasitic capacitance, the value of C1 might have also increased which increased the gain slightly. The actual measured load capacitance, CL, is also higher than 10 pF due to the parasitic capacitance in the printed circuit board, thus reducing the upper corner frequency fH. Even though the upper cut-off frequency is lower than the simulation result, it is still much higher than the bandwidth of the breathing signal. Although the increased bandwidth in the measurement results of the amplifier increases the input-referred noise, it does not affect the overall power consumption and other performance parameters of the amplifier. Simulation and measurement results of the input-referred voltage noise spectrum of the OTA are depicted in Fig. 7. The input-referred noise is measured using the SR785 Dynamic Signal Analyzer and the SR570 Preamplifier. By integrating the simulated noise over the 125 mHz to 12.8 kHz bandwidth, a noise voltage of 6.85 µV or 4.847 µVrms is evaluated. Due to the limitation of the instrument noise voltage 60 >100 dB 104.37 PSRR, dB — 55 50 >70 — 74.62 4 Conclusion In this paper, the design of a low-power, low-noise charge amplifier is proposed for efficient charge-to-voltage conversion. The charge amplifier is designed using 130 nm CMOS process with a power consumption of only 1.8 µW, a − 3 dB cut-off frequency of 10 mHz to 13 kHz and an NEF of 2.79. The simulation and the measurement results show significant improvement over existing amplifier designs, validating the proposed amplifier to be a good candidate for the front end of a wearable respiration-monitoring system based on the PVDF sensor. 5 References 1Pullano, S.A., Mahbub, I., Bianco, M.G., et al.: 'Medical devices for paediatric apnea monitoring and therapy: past and new trends', IEEE Rev. Biomed. Eng., 2017, 10, pp. 199– 212 2Mahbub, I., Hasan, M., Pullano, S.A., et al.: ' A low power wireless sleep apnea detection system based on pyroelectric sensor'. IEEE Topical Conf. Biomedical Wireless Technologies, Networks, and Sensing Systems, San Diego, January 25–28, 2015 3Manjunatha, G.R., Rajanna, K., Mahapatra, D.R., et al.: 'Polyvinylidene fluoride film based nasal sensor to monitor human respiration pattern: an initial clinical study', J. Clin. Monit. Comput., 2013, 27, (6), pp. 647– 657 4Mahbub, I., Islam, S.K., Shamsir, S., et al.: ' A low power wearable respiration monitoring sensor using pyroelectric transducer'. United States National Committee of URSI National Radio Science Meeting (USNC-URSI NRSM), Boulder, USA, 4–7 January 2017 5Cook, C., Sutherland, J., Segal, S., et al.: 'Studies of respiratory physiology in the newborn infant. III. Measurements of mechanics of respiration', J. Clin. Invest., 1957, 36, pp. 440– 448 6Rodríguez-Molinero, A., Narvaiza, L., Ruiz, J., et al.: 'Normal respiratory rate and peripheral blood oxygen saturation in the elderly population', J. Am. Geriatr. Soc., 2013, 61, pp. 2238– 2240 7Denison, T., Consoer, K., Santa, W., et al.: 'A 2 µW 100 nV/rtHz chopper-stabilized instrumentation amplifier for chronic measurement of 130 neural field potentials', IEEE J. Solid-State Circuits, 2017, 42, pp. 2934– 2945 8Harrison, R.R., Charles, C.: 'A low-power low-noise CMOS amplifier for neural recording applications', IEEE J. Solid-State Circuits, 2003, 38, pp. 958– 965 9Kmon, P., Gryboś, P.: 'Energy efficient low-noise multichannel neural amplifier in submicron CMOS process', IEEE Trans. Circuits Syst. I, Regul.Pap., 2013, 60, pp. 1764– 1775 10Mahbub, I., Shamsir, S., Islam, S.K., et al.: ' A low noise front-end amplifier for pyroelectric transducer based respiration monitoring system'. IEEE 60th Int. Midwest Symp. on Circuits and Systems (MWSCAS), Boston, MA, August 6–9, 2017 11Mahbub, I., Oh, T., Shamsir, S., et al.: ' Design of a pyroelectric charge amplifier and a piezoelectric energy harvester for a novel non-invasive wearable and self-powered respiratory monitoring system'. IEEE Region 10 Humanitarian Technology Conf. (R10-HTC), Dhaka, 2017, pp. 105– 108 12Höppe, P.: 'Temperatures of expired air under varying climatic conditions', Int. J. Biometeorol., 1981, 25, pp. 127– 132 13Assaad, R., Silva-Martinez, J.: 'Enhancing general performance of folded-cascode amplifier by recycling current', Electron. Lett., 2007, 43, pp. 1243– 1244 14Holleman, J., Otis, B.: ' A sub-microwatt low-noise amplifier for neural recording'. 29th Annual Int. Conf. of the IEEE Engineering in Medicine and Biology Society (EMBS), Lyon, France, 22–26 August 2007 15Yang, T., Holleman, J.: 'An ultralow-power low-noise CMOS biopotential amplifier for neural recording', IEEE Trans. Circuits Syst. II, Express Briefs, 2015, 62, (10), pp. 927– 931 16Liu, L., Zou, X., Goh, W.L., et al.: '800 nW 43 nV/√Hz neural recording amplifier with enhanced noise efficiency factor', Electron. Lett., 2012, 48, (9), pp. 479– 480 17Zheng, J., Ki, W.H., Tsui, C.Y.: 'A fully integrated analog front End for biopotential signal sensing', IEEE Trans. Circuits Syst. I, Regul.Pap., 2018, 65, (11), pp. 1– 10 Citing Literature Volume13, Issue4July 2019Pages 499-503 FiguresReferencesRelatedInformation
Referência(s)