PUBLICATIONS

Journal Articles

An Adaptable Interface Circuit with Multi-Stage Energy Extraction for Low Power Piezoelectric Energy Harvesting MEMS

Cite as: S. Chamanian, H. Ulusan, A. Koyuncuoglu, A. Muhtaroglu and H. Kulah, "An Adaptable Interface Circuit with Multi-Stage Energy Extraction for Low Power Piezoelectric Energy Harvesting MEMS," Early access: in IEEE Transactions on Power Electronics. doi: 10.1109/TPEL.2018.2841510

This paper presents a self-powered interface circuit to extract energy from ambient vibrations for powering up microelectronic devices. The circuit interfaces a piezoelectric energy harvesting MEMS device to scavenge acoustic energy. Synchronous electric charge extraction (SECE) technique is deployed through the implementation of a novel multi-stage energy extraction (MSEE) circuit in 180nm HV CMOS technology to harvest and store energy. The circuit is optimized to operate with minimum power losses when input power is limited, and adapts well to operating conditions with higher input power. The highly accurate peak detector was validated for a wide piezoelectric frequency range from 20 Hz to 4 kHz. A charging efficiency of about 84% has been achieved for 4.75 V open circuit piezoelectric voltage excited at 390 Hz input vibration under nominal input power range of 30-80 μW. Power optimizations enable the circuit to maintain a conversion efficiency of 47% at input power level as low as 3.12 μW. MSEE provides up to 15% efficiency improvement compared to traditional SECE, and maintains power efficiency as high as possible for a wide input power range.

.
 .
Thin film piezoelectric acoustic transducer for fully implantable cochlear implants

Cite as: B. Ilik, A. Koyuncuoglu, Ş.S. Özlem, H. Külah, “Neural Stimulation Interface with Ultra-Low Power Signal Conditioning Circuit for Fully-Implantable Cochlear Implant Applications,” in Sensors and Actuators A: Physical Volume 280, 1 September 2018, Pages 38-46

This paper reports the development of a single cantilever thin film PLD-PZT transducer prototype. The device was experimentally characterized by attaching it on an acoustically vibrating membrane resembling the behavior of the eardrum. Acceleration characteristic of the sensor attached on the membrane was obtained by using a Laser Doppler Vibrometer (LDV) as the output voltage was measured by an oscilloscope. A voltage output of 114 mV was obtained, when the device was excited at 110 dB Sound Pressure Level (SPL) at 1325 Hz. This is the highest value for a thin film piezoelectric transducer in the literature to our knowledge. Using the results of a finite element analysis for this single-channel prototype, which are within 92% agreement with the experimental results, we performed an optimization study to propose a multi-frequency acoustic sensor to be placed on the eardrum for fully-implantable cochlear implant (FICI) applications. The proposed multi-channel transducer consists of eight cantilever beams. Each of these beams resonates at a specific frequency within the daily acoustic band (250–5000 Hz), senses the eardrum vibration and generates the required voltage output for the stimulation circuitry. The total volume and mass of the transducer are 5 × 5×0.2 mm3 and 12.2 mg, respectively. High sensitivity of the transducer (391.9 mV/Pa @900 Hz) enables transmission of strong signals to be the readout circuit, which can easily be processed. Expected to satisfy all the requirements (volume, mass, and stimulation signal at the hearing band) of FICI applications for the first time in literature, the proposed concept has a groundbreaking nature and it can be referred to as the next generation of FICIs since it revolutionizes the operational principle of conventional CIs.

Conference Proceedings

Bulk PZT Cantilever Based MEMS Acoustic Transducer for Cochlear Implant Applications

Cite as: A. Koyuncuoglu, B. İlik, S. Chamanian, H. Uluşan, P. Ashrafi, D. Işık, and H. Kulah, “Bulk PZT Cantilever Based MEMS Acoustic Transducer for Cochlear Implant Applications,” Proceedings 2017, Vol. 1, Page 584, vol. 1, no. 4, p. 584, Aug. 2017.

This paper presents the first acoustic experimental results of a MEMS based bulk piezoelectric transducer for use in fully implantable cochlear implants (FICI). For this purpose, the transducer was attached onto an acoustically vibrating membrane. Sensing and energy harvesting performances were measured using neural stimulation and rectifier circuits, respectively. The chip has a 150 Hz bandwidth around 1800 Hz resonance frequency that is suitable for mechanical filtering as a sensor. As an energy harvester, bulk piezoelectric transducer generated a rectified power of 16.25 μW with 2.47 VDC with 120 dB-A sound input at 1780 Hz. Among other MEMS acoustic energy harvesters in the literature, reported transducer has the highest power density (1.5 × 10⁻³ W/cm³) to our knowledge.
 .
Thin Film PZT Acoustic Sensor for Fully Implantable Cochlear Implants

Cite as: B. İlik, A. Koyuncuoglu, H. Uluşan, S. Chamanian, D. Işık, Ö. Şardan-Sukas, and H. Kulah, “Thin Film PZT Acoustic Sensor for Fully Implantable Cochlear Implants,” Proceedings 2017, Vol. 1, Page 584, vol. 1, no. 4, 2017.

This paper presents design and fabrication of a MEMS-based thin film piezoelectric transducer to be placed on an eardrum for fully-implantable cochlear implant (FICI) applications. Resonating at a specific frequency within the hearing band, the transducer senses eardrum vibration and generates the required voltage output for the stimulating circuitry. Moreover, high sensitivity of the sensor, 391.9 mV/Pa @900 Hz, decreases the required power for neural stimulation. The transducer provides highest voltage output in the literature (200 mVpp @100 dB SPL) to our knowledge. A multi-frequency piezoelectric sensor, covering the daily acoustic band, is designed based on the test results and validated through FEA. The implemented system provides mechanical filtering, and mimics the natural operation of the cochlea. Herewith, the proposed sensor overcomes the challenges in FICI operations and demonstrates proof-of-concept for next generation FICIs.
 .
An Adaptable Interface Circuit for Low Power MEMS Piezoelectric Energy Harvesters with Multi- Stage Energy Extraction

Cite as: S. Chamanian, H. Ulusan, Ö. Zorlu, A. Muhtaroğlu, and H. Külah, “An Adaptable Interface Circuit for Low Power MEMS Piezoelectric Energy Harvesters with Multi- Stage Energy Extraction,” in BioCAS 2017, Oct. 2017, Turin, Italy.

This paper presents a self-powered interface circuit to extract energy from ambient vibrations for powering up microelectronic devices. The system uses a MEMS piezoelectric energy harvester to scavenge power in 5 μW to 400 μW range. Synchronous electric charge extraction (SECE) technique is utilized to transfer harvested energy to output storage with the help of a novel multi-stage energy extraction (MSEE) circuit. The circuit is optimized in 180nm HV CMOS technology to operate with minimum power losses at the lowest allowable input power, and adjusts well to higher input power due to the MSEE circuit. The circuit operation was validated for a wide piezoelectric frequency range from 20 Hz to 4 kHz. Power efficiency between 62% and 81% has been achieved for the input power range of 5 μW to 173 μW at 198 Hz input vibration. MSEE provides up to 15% efficiency improvement compared to traditional SECE to keep power efficiency as high as possible for the full input power range.
 .
Neural Stimulation Interface with Ultra-Low Power Signal Conditioning Circuit for Fully-Implantable Cochlear Implant Applications

Cite as: H. Uluşan, S. Chamanian, Ö. Zorlu, A. Muhtaroğlu and H. Külah, “Neural Stimulation Interface with Ultra-Low Power Signal Conditioning Circuit for Fully-Implantable Cochlear Implant Applications,” in BioCAS 2017, Oct. 2017, Turin, Italy.
This paper presents an ultra-low power interface circuit to stimulate auditory nerves through fully-implantable cochlear implants (FICIs). The interface circuit senses signals generated from a multi-frequency piezoelectric sensor array, and generates neural stimulation current according to input sound level. Firstly, piezoelectric sensor output is amplified, and compressed with an ultra-low power logarithmic amplifier (LA). This significantly reduces power by eliminating the compression in the next stages. Then, amplified signal is envelope-detected, and utilized as a reference for stimulation current generation using a voltage controlled current source. Finally, biphasic stimulation current is delivered to the nerves through a switch matrix. The circuit has been designed and fabricated in 180nm high-voltage CMOS technology. 8-channel stimulator dissipates about 667 μW as it generates 110 μA biphasic stimulation current, while the front-end signal conditioning unit dissipates only 51.2 μW.
 
Our Nominated Poster:

MEMRO 2018 Poster Title: A SYNTHETIC TYMPANIC MEMBRANE FOR MIDDLE EAR ACOUSTIC SENSOR TESTS OF A FULLY IMPLANTABLE COCHLEAR PROSTHESIS

Click to read the abstract: MEMRO 2018 Program Abstracts

Authors: Parinaz Ashrafi, Dilek Işık Akcakaya, Haluk Külah

Supervised Thesis

Ultra-Low Power Interface Electronics Design for Fully Implantable Cochlear Implants

Cite as: H. Uluşan, “Ultra-Low Power Interface Electronics Design for Fully Implantable Cochlear Implants, Electrical and Electronics Engineering, METU, 2019.

Hasan Ulusan presenting his PhD work, 2018/a>

Cochlear implants are one of the most successful neural prosthesis where users could go from being profoundly deaf to enjoying high degree of speech perception. However, subsequent aesthetic concerns, damage risks and high power dissipation associated with bulky external units of cochlear implants have redirected recent studies to fully implantable cochlear implants (FICI). Although, implantable sensors occupies the largest portion of the previous researches on FICIs, design of the low powered signal conditioning interface circuit is a bottle-neck to accomplish a FICI system. In this thesis, a novel fully integrated interface circuit is designed to process signals from an implantable multi-frequency piezoelectric (PZT) cantilever set for stimulation of the auditory neurons. The 1st generation FICI interface is focused on power dissipation of the front-end signal conditioning circuit. Power of the front-end circuit is reduced through a novel logarithmic amplifier design that combines amplification and compression stages. The conditioned signal controls the biphasic rectangular current pulses for neural stimulation. Single channel performance of the circuit has been tested with a thin film pulsed-laser deposition (PLD) PZT sensor. The interface generates biphasic current in the range of 110-430 µA for acoustic input of 60-100 dB sound pressure level (SPL). Power dissipation of the front-end signal conditioning and the overall system for 8-channel operation is projected from measurement as 51.2 and 691.2 µW, respectively. After validating the 1st generation interface performance, power dissipation and input range of the sensor front was improved at the 2nd generation FICI interface through novel current mode circuits. The conditioned signal is converted into biphasic neural stimulation current with a 7-bit user-programmed DAC to enable patient fitting (calibration). The proposed circuit introduces an optimized stimulation current waveform to reduce the electrode voltage and hence supply voltage of the stimulator (most power hungry part) by about 20%. The designed system has been tested with up to 60 dB input dynamic range (40-100 dB SPL) while the minimum threshold and maximum comfort levels of the system are 0 and 1 mA, respectively. The 8-channel interface has been validated to be fully functional with the front-end and the overall circuit power dissipation of 19.7 and 471.7 μW, when excited by a mimicked speech signal. The proposed 2nd generation interface electronics is the first sub-500 μW FICI interface that provides more than 30 years of operational lifetime, and reduces the healthcare cost and risks associated with surgical battery replacements. The implantable device performance of the 2nd generation FICI interface electronics has been validated through in-vivo tests on a guinea pig. After validating healthy and partially-deafened hearing performance of the guinea pig, the electrical stimulation performance of controlled current stimulator and the FICI interface electronics were tested and compared. The neural stimulation capability of acoustically excited FICI system with intra-cochlear electrodes has been validated with 55 dB hearing threshold and 45 dB input dynamic range. The proposed system is the first FICI interface electronics with ultra-low power dissipation and wide dynamic range that is also validated with in-vivo tests. Keywords: Fully Implantable Cochlear Implant, Neural Stimulation, Bionic Ear, Ultra-Low Power, Interface Electronics
 .
Design and Implementation of an Interface Circuit for Piezoelectric Energy Harvesters

Cite as: S. Chamanian, “Design and Implementation of an Interface Circuit for Piezoelectric Energy Harvesters, Electrical and Electronics Engineering, METU, 2018.

Micro-fabricated piezoelectric transducer is a prominent harvesting method due to its small size and relatively high energy density. However, the available interface circuits (IC) in the literature for piezoelectric energy harvesters (PEH) are generally designed for macro-scaled versions having a power output in the range of hundreds μW. The efficiency of such systems is significantly diminished when input power drops to tens of μWs or less, which is the pertinent power output range for micro-fabricated devices. Therefore, it is necessary to develop efficient electronics to extract energy from low output power levels of Microelectromechanical systems (MEMS) piezoelectric energy harvesters. The main aim of this thesis is to develop ICs that can efficiently extract energy from the MEMS piezoelectric energy harvesters and charge storage element for powering up micro-electronic devices. In the first IC, a novel multi-stage energy extraction method is proposed to optimize the implementation of the synchronous electrical charge extraction (SECE) converter. This optimization allows downsizing of the external inductor without affecting the power-conversion efficiency. Then, a charge management approach is presented to speed up the charging of the large storage element. The advantage of this method is that it accelerates the transition from passive mode to active mode. Several circuit techniques are introduced to enhance practicability of the energy harvesting IC. An autonomous system is achieved through a start-up circuit with power management Circuit that initiates the circuit from no primary charge. Implementations of active negative voltage converter and new ultra-low-power peak detector expand operating frequency range of the IC from 100 Hz to 4 kHz. Finally, self-adapting multi-stage energy extraction (MSEE) enhances power conversion efficiency for a wide input-power range. Maximum charging efficiency of 84 % is achieved with a 1 mH external inductor, while MEMS PEH is excited at 390 Hz. Second IC introduces a novel nonlinear switching technique aiming to boost extracted energy from low coupling factor PEHs and provide load-independent energy extraction with a single inductor. The idea is to enhance effective damping force of the PEH by processing piezoelectric voltage through a set of switches and an inductor. A novel maximum power point (MPP) sensing approach is proposed to achieve the optimal operation point of the proposed circuit regardless of input excitation level, for the first time in literature. The IC can efficiently harvest energy from shock vibrations, as MPP circuit adjusts optimal point regardless of the variation in the available energy on PEH. In the end, an efficient hybrid energy-harvesting interface is presented to simultaneously scavenge power from electromagnetic and piezoelectric sources, while driving a single load. The total simultaneously extracted power from both harvesters is more than the power obtained from each independently. The hybrid IC reaches up to 90% conversion efficiency with output power level of 100 μW. A wearable harvesting prototype consisting of custom-made electromagnetic harvester, off-the-shelf PEH, and the proposed interface circuit is built and tested to harvest energy from body movement.
 .
Thin Film Piezoelectric Acoustic Transducer for Cochlear Implant Applications

Cite as: Bedirhan İlik, “Thin Film Piezoelectric Acoustic Transducer for Cochlear Implants Applications,” Electrical and Electronics Engineering, METU, 2018.