Pieter Harpe
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Pieter Harpe has a total of 161 co-authors. Here are the top co-authors by number of shared publications:
- Eugenio Cantatore
Eindhoven University of Technology
26 shared publications - Haoming Xin
Integrated Circuits Group, Eindhoven University of Technology, 5600MB Eindhoven, The Netherlands
10 shared publications - Hanyue Li
Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands
9 shared publications - Yuting Shen
Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands
9 shared publications - Sotir Ouzounov
Philips Research, Eindhoven, The Netherlands
7 shared publications - Massimo Mischi
Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands
6 shared publications
Latest Publications
- 2 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Flexible Electronics
- Vol. 5 (4) , 95-106
A Hybrid EOG System Featuring Flexible a-IGZO TFTs and a 180 nm CMOS IC
Flexible wearable sensors are becoming increasingly popular due to their capability to monitor physiological signals while ensuring high user comfort. Wearable sensors built using only flexible electronics result in inferior system performance due to the inherent limitations of flexible technologies. Flexible hybrid electronics (FHE) combines the strengths of flexible technologies (low cost, larger area coverage, user comfort) with the high performance and energy efficiency of silicon ICs. Here, an FHE system is described that addresses the challenge of system functions partitioning between flexible and silicon (Si) subsystems. The presented design approach is applied in this work to a multichannel electro-oculography (EOG) system, but can be expanded to various FHE application domains. The proof of concept fully-integrated 8-channel FHE EOG system exploits a front-end in a-IGZO TFT technology, interfaced with a custom Si IC that supports the front-end and provides conversion of the acquired signals to the digital domain. In this way the low cost, flexible form factor and the large area character of the flexible thin-film transistor (TFT) technology are leveraged, while exploiting at the same time the accuracy, computational capability and energy efficiency of the Si IC. We propose a hybrid high-pass filter (hHPF) as interface between the electrodes and the front-end, whereby TFTs and capacitors on foil are being controlled by the Si IC. The hybrid approach exploits the capabilities of flexible electronics to bring functionality close to the electrodes and its potential to cope with a high number of sensor channels, with the better precision of CMOS, enabling a fully integrated filter without discrete components. The TFT front-end converts the input signal voltages to a pulse-width modulated (PWM) representation, which is then quantized in the Si IC, to provide a digital output. The Si IC enables conversion to the digital domain with much higher efficiency compared to state-of-the-art TFT solutions. The digital output is further used in an algorithm for feature extraction and classification to detect the desired signals. The experimental electrical characterization of the system demonstrates an input-referred noise of 6.68 μVrms in a 1-30 Hz band, with a chopping frequency of 1 kHz. The input impedance offered by the integrated hHPF is 6.88 GΩ at 0.089 Hz (with multiplexing disabled). The total power consumption per channel is 84 μW. In-vivo experiments have also been performed against a gold standard (a commercial bio-potential amplifier), to confirm a faithful representation of the eye activities. The signals acquired during in-vivo experiments are successfully classified using a support vector machine (SVM) classifier, with an average accuracy of 87.6 ± 2.9%.
- 2 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Flexible Electronics
- Vol. 5 (4) , 95-106
A Hybrid EOG System Featuring Flexible a-IGZO TFTs and a 180 nm CMOS IC
Flexible wearable sensors are becoming increasingly popular due to their capability to monitor physiological signals while ensuring high user comfort. Wearable sensors built using only flexible electronics result in inferior system performance due to the inherent limitations of flexible technologies. Flexible hybrid electronics (FHE) combines the strengths of flexible technologies (low cost, larger area coverage, user comfort) with the high performance and energy efficiency of silicon ICs. Here, an FHE system is described that addresses the challenge of system functions partitioning between flexible and silicon (Si) subsystems. The presented design approach is applied in this work to a multichannel electro-oculography (EOG) system, but can be expanded to various FHE application domains. The proof of concept fully-integrated 8-channel FHE EOG system exploits a front-end in a-IGZO TFT technology, interfaced with a custom Si IC that supports the front-end and provides conversion of the acquired signals to the digital domain. In this way the low cost, flexible form factor and the large area character of the flexible thin-film transistor (TFT) technology are leveraged, while exploiting at the same time the accuracy, computational capability and energy efficiency of the Si IC. We propose a hybrid high-pass filter (hHPF) as interface between the electrodes and the front-end, whereby TFTs and capacitors on foil are being controlled by the Si IC. The hybrid approach exploits the capabilities of flexible electronics to bring functionality close to the electrodes and its potential to cope with a high number of sensor channels, with the better precision of CMOS, enabling a fully integrated filter without discrete components. The TFT front-end converts the input signal voltages to a pulse-width modulated (PWM) representation, which is then quantized in the Si IC, to provide a digital output. The Si IC enables conversion to the digital domain with much higher efficiency compared to state-of-the-art TFT solutions. The digital output is further used in an algorithm for feature extraction and classification to detect the desired signals. The experimental electrical characterization of the system demonstrates an input-referred noise of 6.68 μVrms in a 1-30 Hz band, with a chopping frequency of 1 kHz. The input impedance offered by the integrated hHPF is 6.88 GΩ at 0.089 Hz (with multiplexing disabled). The total power consumption per channel is 84 μW. In-vivo experiments have also been performed against a gold standard (a commercial bio-potential amplifier), to confirm a faithful representation of the eye activities. The signals acquired during in-vivo experiments are successfully classified using a support vector machine (SVM) classifier, with an average accuracy of 87.6 ± 2.9%.
- 28 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26 (6) , 8945-8955
Feasibility of Capacitive ECG R-Peak Detection for Cardiac Gating in 3-T MRI
Non-contact capacitive electrodes provide a higher comfort level and can avoid unpleasant skin preparation compared to conventional wet electrodes. This work studies the feasibility of using capacitive electrodes to measure the ECG (Electrocardiogram) R-peaks for cardiac gating inside an MRI (Magnetic Resonance Imaging) environment. A phantom test setup with ECG generation, 16 capacitive electrodes, ECG acquisition hardware, and an R-peak detection algorithm, is developed. Experiments are first conducted in a 3-Tesla static field (B0) and then combined with a switching gradient field. Various gradient conditions (e.g., directions, strengths, frequencies) and respiration-mimicking motions are included. The noise levels and artifacts of the raw data are first observed. An exemplary algorithm is then adopted to extract R-peaks from multi-channel data. The results show that under a 3T static field, the R-peaks are visible both in raw data only without motion and after processing even with motion. Under gradient field, some channels or epochs are strongly affected by artifacts, while other channels or epochs still contain relatively clean ECG data. Thanks to the multi-channel approach, this can still lead to successful R-peak detection without motion. The addition of motion deteriorates the signal quality and can lead to detection failure with the adopted signal processing method. The experimental results obtained in this work reveal the feasibility of capacitive ECG R-peak detection inside 3T MRI for the first time and provide a first step toward its future application for synchronizing the MRI of the heart.
- 28 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26 (6) , 8945-8955
Feasibility of Capacitive ECG R-Peak Detection for Cardiac Gating in 3-T MRI
Non-contact capacitive electrodes provide a higher comfort level and can avoid unpleasant skin preparation compared to conventional wet electrodes. This work studies the feasibility of using capacitive electrodes to measure the ECG (Electrocardiogram) R-peaks for cardiac gating inside an MRI (Magnetic Resonance Imaging) environment. A phantom test setup with ECG generation, 16 capacitive electrodes, ECG acquisition hardware, and an R-peak detection algorithm, is developed. Experiments are first conducted in a 3-Tesla static field (B0) and then combined with a switching gradient field. Various gradient conditions (e.g., directions, strengths, frequencies) and respiration-mimicking motions are included. The noise levels and artifacts of the raw data are first observed. An exemplary algorithm is then adopted to extract R-peaks from multi-channel data. The results show that under a 3T static field, the R-peaks are visible both in raw data only without motion and after processing even with motion. Under gradient field, some channels or epochs are strongly affected by artifacts, while other channels or epochs still contain relatively clean ECG data. Thanks to the multi-channel approach, this can still lead to successful R-peak detection without motion. The addition of motion deteriorates the signal quality and can lead to detection failure with the adopted signal processing method. The experimental results obtained in this work reveal the feasibility of capacitive ECG R-peak detection inside 3T MRI for the first time and provide a first step toward its future application for synchronizing the MRI of the heart.
- Open Access
- 21 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Solid-State Circuits Society
- Vol. 5, 454-455
Special Section on Temperature Resilient Systems and Circuits
- 21 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Solid-State Circuits Society
- Vol. 5, 454-455
Open AccessSpecial Section on Temperature Resilient Systems and Circuits
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 73 (6) , 3986-3995
Shallow Neural Network-Based Error Compensation of a Non-Linear SAR ADC Sampling Stage
Non-idealities of the track-and-hold (T&H) circuit in analog-to-digital converters (ADCs) can severely limit the ADC linearity. In practical applications, traditional calibration techniques are ineffective in suppressing these effects due to the multi-dimensional dependency of the non-idealities on e.g., input voltage, frequency and temperature. To address this, we propose a calibration solution that combines circuit knowledge of the error mechanisms with a data-driven approach into a shallow neural network (NN) containing just 5 neurons. Due to the small size of the NN, the output of the network can be interpreted and referred back to the underlying circuit non-idealities improving the predictability and reliability of the model. The effectiveness of the NN-based calibration is verified on silicon measurements of 10 different 12-bit SAR ADC samples manufactured in a 65nm CMOS technology, demonstrating an average improvement up to 6.8dB and 13.3dB on the measured SNDR and SFDR, respectively. To the best of the author’s knowledge, this is the first demonstration of a compact, interpretable NN-based calibration solution tested over amplitude, frequency, and temperature for multiple samples, while maintaining physics-based insight into the correction process.
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 73 (6) , 3986-3995
Shallow Neural Network-Based Error Compensation of a Non-Linear SAR ADC Sampling Stage
Non-idealities of the track-and-hold (T&H) circuit in analog-to-digital converters (ADCs) can severely limit the ADC linearity. In practical applications, traditional calibration techniques are ineffective in suppressing these effects due to the multi-dimensional dependency of the non-idealities on e.g., input voltage, frequency and temperature. To address this, we propose a calibration solution that combines circuit knowledge of the error mechanisms with a data-driven approach into a shallow neural network (NN) containing just 5 neurons. Due to the small size of the NN, the output of the network can be interpreted and referred back to the underlying circuit non-idealities improving the predictability and reliability of the model. The effectiveness of the NN-based calibration is verified on silicon measurements of 10 different 12-bit SAR ADC samples manufactured in a 65nm CMOS technology, demonstrating an average improvement up to 6.8dB and 13.3dB on the measured SNDR and SFDR, respectively. To the best of the author’s knowledge, this is the first demonstration of a compact, interpretable NN-based calibration solution tested over amplitude, frequency, and temperature for multiple samples, while maintaining physics-based insight into the correction process.
- 19 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Magazine
- Vol. 17 (2) , 82-83
Learn to Establish and Achieve Professionalism at Asian Solid-State Circuits 2024 Was a Success (LEAP) [Women in Circuits Corner]
Provides society information that may include news, reviews or technical notes that should be of interest to practitioners and researchers.
- 19 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Solid-State Circuits Magazine
- Vol. 17 (2) , 82-83
Learn to Establish and Achieve Professionalism at Asian Solid-State Circuits 2024 Was a Success (LEAP) [Women in Circuits Corner]
Provides society information that may include news, reviews or technical notes that should be of interest to practitioners and researchers.
- Open Access
- 19 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 53594-53607
An Ultra-Low-Leakage Microcontroller With Configurable Power Management for Energy Harvesting IoT Devices
This paper presents a power management unit (PMU) architecture designed for energy-harvesting IoT devices, integrating a dual-capacitor system, an ultra-low-leakage balloon-based microcontroller (MCU), and a fully digital controller. The PMU dynamically switches between an active capacitor, an idle capacitor, and a backup battery, optimizing energy use in different operational modes. This configuration optimizes capacitor voltages for various operational modes, applying a higher voltage to the active capacitor for enhanced performance and a lower voltage to the idle capacitor for ultra-low-leakage data retention. Additionally, a fully digital PMU controller replaces conventional analog components, leveraging a digital voltage monitor and converter-free voltage regulation. The test chip is designed and fabricated using a 28-nm FD-SOI technology. The balloon-based MCU measurement shows that it can retain data during idle mode with a power consumption of only 3nW at 0.6V, achieving a 4400× reduction in leakage power compared to conventional MCUs without balloon memory. Furthermore, the results of the system measurement demonstrate a substantial reduction in battery usage from 88% in traditional setups to 0%, while ensuring continuous operation and data retention.
- 19 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 53594-53607
Open AccessAn Ultra-Low-Leakage Microcontroller With Configurable Power Management for Energy Harvesting IoT Devices
This paper presents a power management unit (PMU) architecture designed for energy-harvesting IoT devices, integrating a dual-capacitor system, an ultra-low-leakage balloon-based microcontroller (MCU), and a fully digital controller. The PMU dynamically switches between an active capacitor, an idle capacitor, and a backup battery, optimizing energy use in different operational modes. This configuration optimizes capacitor voltages for various operational modes, applying a higher voltage to the active capacitor for enhanced performance and a lower voltage to the idle capacitor for ultra-low-leakage data retention. Additionally, a fully digital PMU controller replaces conventional analog components, leveraging a digital voltage monitor and converter-free voltage regulation. The test chip is designed and fabricated using a 28-nm FD-SOI technology. The balloon-based MCU measurement shows that it can retain data during idle mode with a power consumption of only 3nW at 0.6V, achieving a 4400× reduction in leakage power compared to conventional MCUs without balloon memory. Furthermore, the results of the system measurement demonstrate a substantial reduction in battery usage from 88% in traditional setups to 0%, while ensuring continuous operation and data retention.
- 27 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (8) , 3935-3946
An Efficient, Dynamically Adjustable, Multipurpose Ultrasound Digitizer Array in 40 nm CMOS
In this work, a test chip for a 32 channel ultrasound imaging digitizer for intra-cardiac echocardiography is presented. The focus of this design is on area- and power efficiency, as well as multi-purpose usage for various catheters. It contains 32 individual analog front-ends and 12 bit SAR ADCs operating at 40MS/s to enable off-chip digital beamforming. The analog front-end is programmable in power, bandwidth, gain and can be partially bypassed. The front-end bypass and slew-rate can be dynamically adjusted to save power during a receive period. On-chip supply regulation is included, which can be duty-cycled between receive periods to power down the system and save power. This leads to a power supply consumption between 4.5 and 14.1mW when always on, or between 0.37 and 1.15mW when duty-cycling at an 8% ratio. When operating at maximum power, an SFDR of 57.5dB and SNR of 56.1dB are achieved.The ADC performance can optionally be improved by clocking the array in 4 separate clock phases, resulting in lower peak currents, less power supply disturbance and an overall linearity improvement of 6dB. The 32 channel array occupies 1mm2 including decoupling capacitance, combining low area and low power operation.
- 27 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (8) , 3935-3946
An Efficient, Dynamically Adjustable, Multipurpose Ultrasound Digitizer Array in 40 nm CMOS
In this work, a test chip for a 32 channel ultrasound imaging digitizer for intra-cardiac echocardiography is presented. The focus of this design is on area- and power efficiency, as well as multi-purpose usage for various catheters. It contains 32 individual analog front-ends and 12 bit SAR ADCs operating at 40MS/s to enable off-chip digital beamforming. The analog front-end is programmable in power, bandwidth, gain and can be partially bypassed. The front-end bypass and slew-rate can be dynamically adjusted to save power during a receive period. On-chip supply regulation is included, which can be duty-cycled between receive periods to power down the system and save power. This leads to a power supply consumption between 4.5 and 14.1mW when always on, or between 0.37 and 1.15mW when duty-cycling at an 8% ratio. When operating at maximum power, an SFDR of 57.5dB and SNR of 56.1dB are achieved.The ADC performance can optionally be improved by clocking the array in 4 separate clock phases, resulting in lower peak currents, less power supply disturbance and an overall linearity improvement of 6dB. The 32 channel array occupies 1mm2 including decoupling capacitance, combining low area and low power operation.
- 18 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 18 (5) , 1100-1111
An Event-Based Neural Compressive Telemetry With >11× Loss-Less Data Reduction for High-Bandwidth Intracortical Brain Computer Interfaces
Intracortical brain-computer interfaces offer superior spatial and temporal resolutions, but face challenges as the increasing number of recording channels introduces high amounts of data to be transferred. This requires power-hungry data serialization and telemetry, leading to potential tissue damage risks. To address this challenge, this paper introduces an event-based neural compressive telemetry (NCT) consisting of 8 channel-rotating Δ-ADCs, an event-driven serializer supporting a proposed ternary address event representation protocol, and an event-based LVDS driver. Leveraging a high sparsity of extracellular spikes and high spatial correlation of the high-density recordings, the proposed NCT achieves a compression ratio of >11.4×, while consumes only 1 µW per channel, which is 127× more efficient than state of the art. The NCT well preserves the spike waveform fidelity, and has a low normalized RMS error <23% even with a spike amplitude down to only 31 µV.
- 18 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Circuits and Systems
- Vol. 18 (5) , 1100-1111
An Event-Based Neural Compressive Telemetry With >11× Loss-Less Data Reduction for High-Bandwidth Intracortical Brain Computer Interfaces
Intracortical brain-computer interfaces offer superior spatial and temporal resolutions, but face challenges as the increasing number of recording channels introduces high amounts of data to be transferred. This requires power-hungry data serialization and telemetry, leading to potential tissue damage risks. To address this challenge, this paper introduces an event-based neural compressive telemetry (NCT) consisting of 8 channel-rotating Δ-ADCs, an event-driven serializer supporting a proposed ternary address event representation protocol, and an event-based LVDS driver. Leveraging a high sparsity of extracellular spikes and high spatial correlation of the high-density recordings, the proposed NCT achieves a compression ratio of >11.4×, while consumes only 1 µW per channel, which is 127× more efficient than state of the art. The NCT well preserves the spike waveform fidelity, and has a low normalized RMS error <23% even with a spike amplitude down to only 31 µV.
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 59 (4) , 1194-1203
A 1.8–65 fJ/Conv.-Step 64-dB SNDR Continuous- Time Level Crossing ADC Exploiting Dynamic Self-Biasing Comparators
This work presents a power-efficient level crossing (LC) ADC designed to digitize sparse signals. It uses dynamically self-biased comparators, which require minimal current when the input voltage is far from a decision threshold. It also uses a DAC architecture which avoids the signal attenuation commonly present in prior LC ADC works, improving the achievable SNDR. The prototype is designed and implemented in a 65-nm CMOS technology, and occupies an area of 0.0045 mm $^{2}$ . In a 20 kHz bandwidth, the LC-ADC achieves a 64 dB SNDR. Thanks to the proposed techniques a power efficiency of up to 1.8 fJ/conv.-step is achieved for sinusoidal inputs. For sparse biopotential signals, a FoM $_{\text{W}}$ as low as 0.9 fJ/conv.-step was measured. This makes the prototype interesting for e.g., biomedical applications that make use of spike-based processing.
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 59 (4) , 1194-1203
A 1.8–65 fJ/Conv.-Step 64-dB SNDR Continuous- Time Level Crossing ADC Exploiting Dynamic Self-Biasing Comparators
This work presents a power-efficient level crossing (LC) ADC designed to digitize sparse signals. It uses dynamically self-biased comparators, which require minimal current when the input voltage is far from a decision threshold. It also uses a DAC architecture which avoids the signal attenuation commonly present in prior LC ADC works, improving the achievable SNDR. The prototype is designed and implemented in a 65-nm CMOS technology, and occupies an area of 0.0045 mm $^{2}$ . In a 20 kHz bandwidth, the LC-ADC achieves a 64 dB SNDR. Thanks to the proposed techniques a power efficiency of up to 1.8 fJ/conv.-step is achieved for sinusoidal inputs. For sparse biopotential signals, a FoM $_{\text{W}}$ as low as 0.9 fJ/conv.-step was measured. This makes the prototype interesting for e.g., biomedical applications that make use of spike-based processing.