Wearable ultrasound technology has the potential to enable continuous, non-invasive monitoring of physiological processes, offering new opportunities for diagnostics and patient care. To achieve this, ultrasound transducer arrays must be flexible enough to conform to the human body. However, deformation of a flexible array changes the positions of individual transducer elements, which can significantly degrade the accuracy of ultrasound localisation and imaging algorithms. This thesis investigates the feasibility of using dedicated shape sensors to compensate for these deformations and thereby improve localisation accuracy in wearable ultrasound systems. A concept demonstrator consisting of two rigid ultrasound transducer arrays connected by a flexible substrate was developed. A shape estimation pipeline was developed to reconstruct the geometry of the substrate from sensor measurements and a localisation algorithm was implemented to estimate the position of a point reflector using ultrasound measurements. Both methods were validated individually in simulation and subsequently evaluated together in both simulation and a hardware demonstrator. The results demonstrate that accurate shape estimation and reflector localisation can be achieved using compact strain-based sensing.
Furthermore, the integrated system was able to align localisation estimates from both transducer arrays with sub-millimetre agreement, demonstrating successful compensation for array deformation. It is therefore concluded that shape-sensor-based deformation correction is a promising approach for wearable ultrasound systems. Future work should focus on increasing the spatial density of the shape sensors, improving shape interpolation and reconstruction algorithms, and extending the method to two-dimensional array geometries to support practical imaging applications. ... Read more

This report discusses the design and implementation of a wireless network to demonstrate a dynamic process based on the Susceptible-Infected (SI) model. The system supports up to 10 nodes that communicate over Bluetooth Low Energy (BLE) and employ a token passing based network protocol. The Relative Signal Strength Indicator (RSSI) is used to determine whether connections are formed depending on distance. Each node displays their SI state and to which sub-network they belong through an RGB LED. Furthermore, a centralized visualization system is used to combine this information to provide an overview on a screen. The data for this visualization tool is collected by a visualization node that scans the network and copies the communicated data packets to the visualization script. The data is communicated to a webpage using a WebSocket, this webpage is hosted on a local host and displays an interactive, continuously updating network plot. The model can be initialized through this interface as well. Experimental evaluation shows that the system successfully reproduces the expected probabilistic behavior of the SI model across the implemented network topologies. The prototype operates on battery power with an estimated runtime of approximately 34 hours per node. ... Read more

This report describes the design, building and testing of a high precision data acquisition system that is able to function in high pressure environments. The system uses an FPGA as its controller sending and receiving information to and from a data conversion device with the AD4630 as its ADC and the DAC8811 as its DAC. The system is tested in a pressure chamber which was pressurised up to 500 bars of pressure and the performance difference was found to be negligible. The system had to be proven to be scalable to 16 channels and these 16 channels had to fit inside a volume of 100𝑐𝑚3. ... Read more

This report describes the design, building and testing of a high precision data acquisition system that is able to function in high pressure environments. The system uses an FPGA as its controller sending and receiving information to and from a data conversion device with the AD4630 as its ADC and the DAC8811 as its DAC. The system is tested in a pressure chamber which was pressurised up to 500 bars of pressure and the performance difference was found to be negligible. The system had to be proven to be scalable to 16 channels and these 16 channels had to fit inside a volume of 100𝑐𝑚3. ... Read more

This work investigates whether the use of a ferrite core can improve the performance of a Slayer Exciter Solid State Tesla Coil. The main objective was to design, simulate, construct, and test a ferrite-core Tesla coil capable of producing long and straight discharges. COMSOL Multiphysics was used to calcu- late key electrical parameters, including inductance, capacitance, coupling factor, and AC resistance. These parameters were then used in an optimization routine to compare different coil geometries and determine a suitable final design.
The results show that increasing the inductance of the secondary coil with a ferrite core, allows smaller coil geometries to operate at suitable resonant frequencies; while keeping the resistance rela- tively low. Since the resonant frequency is directly affected by the increased inductance, the influence of operating frequency on discharge behavior was also investigated. Lower frequencies generally pro- duced longer sparks, but these discharges were often more branched. Frequencies around 350 kHz produced slightly shorter, but more stable and mostly unbranched sparks. The number of primary turns, coupling factor, and top-load size were also found to significantly affect the output performance.
An optimized ferrite-core design was constructed and compared with both air-core and alterna- tive ferrite-core coil designs. Experimental testing showed that the optimized design produced longer sparks than the comparable air-core design and performed better than the other tested ferrite-core con- figurations. The coil was able to produce sparks exceeding the target distance while operating within the required frequency range.
Overall, this work demonstrates that ferrite cores can be used effectively to improve compact Slayer Exciter Tesla coil designs and validates the usefulness of COMSOL-based optimization for high-voltage resonant systems. ... Read more

This thesis details the design, implementation, and optimization of a single-switch circuit topology to drive a solid-state Tesla coil (SSTC). The goal of this project is to maximize the lengths of the discharges from the secondary coil given this switching topology. To achieve this, the thesis will look at the impact of a qausi-continuous wave (QCW) input on the discharges of the Tesla coil and compare these to the discharges of a staccato-ramped Tesla coil. A bus modulator approach was used to generate the QCW input, using a buck converter driven by a PWM signal with linearly increasing duty cycle to modulate a bus capacitor to create the QCW input signal. Additionally, this thesis analyzes and characterizes different high-power switch protection methodologies. These design optimizations combine to successfully generate discharges in excess of 40 cm with a 30 cm tall secondary coil. Experimental results revealed that there were no major differences in discharge length when comparing the QCW input to the staccato input when corrected to the same voltage. However, the results revealed that the duration of the QCW input did have a significant effect on the thickness and shape of the resulting discharges and the likeliness of internal discharges to take place. ... Read more

Wearable ultrasound technology is being developed to provide continuous, non-invasive monitoring of the human body. An important challenge in the development of such systems is determining the shape of a flexible ultrasound array, while it conforms to the body's surface. The objective of this thesis was to design, implement, and evaluate a shape sensing system and determine the suitability for it's integration into a wearable flexible ultrasound device. We began with a literature study of existing shape-sensing techniques. Multiple methods were analysed and compared. Based on these analyses we chose one technique to be used as the foundation for the system design. A complete measurement system was then developed only using commercially available components. The design was integrated into a prototype and evaluated through an experiment designed to show its ability to measure deformation. The results showed a positive correlation between the measured output and the applied deformation, demonstrating that the technique is capable of detecting its shape. However, the measurements were not stable, preventing the system from fully meeting the performance requirements established at the start of the project. Because of this we were unable to prove this technique has the capability of measuring an an accuracy needed for flexible ultrasound. Further work should focus on improving measurement stability and performing experiments focusing at accuracy and reliability. ... Read more

This work investigates whether a power amplifier (PA) designed for a lower operating frequency (10GHz) can be effectively implemented using a process that has a 𝑓𝑚𝑎𝑥 of well above 100 GHz. Operating a process far below its maximum frequency offers potential benefits in gain and efficiency, but also presents challenges. Particularly for the stability of the amplifier, which becomes increasingly critical at high gain levels.

To investegate the potential for such an RF power amplifier, Gallium Nitride (GaN) High Electron Mobility Transistors (HEMTs) are used. This is a technology that forms the current State-of-Art for microwave high-power high-efficiency MMICs. The circuit is fabricated using WIN Semiconductors NP12-01 GaN-on-SiC technology with 0.12 𝜇m gates. They offer a technology with one of the shortest gatelengths that are around in the industry.

First a Programme of Requirements was established, before starting the design and four Key Per-formance Indicators (power, efficiency, stability and size) were defined. A design process was followed that started with transistor measurements, after which an architectural design was made, followed by schematic implementation, electromagnetic (EM) modelling and final layout creation. Continuous refinement between schematic, EM and layout stages where necessary to avoid excessive simulation times. After design rule verification and stability checks, the circuit was submitted for fabrication, with MMIC samples received back approximately four months later, after which, an extensive characterization of MMICs was performed, through both small-signal as well as non-linear measurements.

All this work resulted in the following KPI’s:
• For the power, a peak power of over 38 dBm was realized under dedicated measurement conditions and over 35 dBm for nominal measurement conditions (20 V 𝑉𝐷𝑆) and a bandwidth of more than
2 GHz (-1 dB bandwidth) or 3 GHz (3-dB bandwidth) was obtained.
• For the PAE, 35% was nominally achieved over various bias points and frequencies.
• Unconditional stability was achieved.
• The size of the amplifier was 2x1.75 mm2, resulting in slightly below 2 W/mm.

While amplifiers with higher absolute power at X-band exist, the NP12 process is unlikely to surpass the state-of-the-art in this regard due to its limited maximum drain voltage, which constrains the achievable output power. However, the measured efficiency and gain are promising for a first design iteration. Notably, the implementation of the stability analysis for this power amplifier took a significant effort, but was proved highly effective as no instability was observed at all during the measurements. Two variations of the amplifier were implemented to investigate the possibility of reusing the source via-hole for two adjacent output stage transistors. This variation, with re-used via-holes, worked similar to the baseline design, but exhibited a lower PAE.

In conclusion, the expectation is strengthened that the advantages of a mm-wave process used at X-band frequencies can pay off in a modest increase of the efficiency and a significantly increased gain. Finally, the work resulted in one conference publication, and two tutorial documents for future students (an AWR student guide and MMIC mounting tutorial). ... Read more

This thesis discusses the integration of the Illuminator software with hardware components, for the purpose of creating a table-top, Plug-and-Play energy system demonstrator. The thesis starts by introducing the motivation for an energy system demonstrator in Chapter 1. An interactive, visual demonstrator can help educate people on energy systems in a way that
is more intuitive. The main product should show the power flow on the table, implement Plug-and-Play dynamics, and be scalable. LED strips are used to visualize power flows in the table-top network, in combination with the Digispark ATtiny85. For determining the topology, static ID pairs were used. A double simulation is used to implement Plug-and-Play dynamics.
The first simulation configures the topology used by the second simulation, based on the hardware connections. The second simulation runs the Illuminator simulation. During this simulation, checks are run to see whether a physical connection has changed, such as a cable being unplugged. The simulation and then starts the reconfiguration process again.
Testing the reliability and run-time of the implementation is documented in Chapter 6, with a focus on how well the implementation scales with the size of the simulation. It was concluded that the Digispark’s communication with the Raspberry Pi would often stall, requiring error correction to be implemented. Even then, the data transfer to the Digispark from the Raspberry Pi fails on the first try an average of 48% of the time. Determining how long a setup takes to reconfigure was estimated using a computer, since the Raspberry Pi’s aren’t powerful enough to simulate dozens of models. It was determined that a reconfiguration of 20 models takes about 100 seconds. ... Read more

Tailless flapping-wing drones mimic the flight mechanics of insects and offer unique advantages in agility and maneuverability compared to rotor-based drones. Yet, their limited onboard computational resources and non-linear flight dynamics complicate active attitude control. Neural network-based controllers have shown promising control performance for this task but they exceed a flapping-wing drone's onboard computational budget. To this end, the Sparse Identification of Nonlinear Dynamics (SINDy) algorithm offers a promising solution by distilling a neural network-based controller into a simplified mathematical expression. This expression is better suited for mapping onto an FPGA which provides the power efficiency essential for an energy-constrained drone. The current work presents an automated workflow to translate the simplified controller into an HDL description, maximizing DSP block usage for power and resource efficiency. Demonstration of this workflow on the pendulum simulation as a proof-of-concept has shown its efficacy. The included optimization techniques have resulted, on average, in a 40% reduction of DSP block usage, without compromising controller performance. This scalable, platform-agnostic workflow streamlines the design of a controller's hardware implementation and allows its future application to a flapping-wing drone. ... Read more

The Energy System Integration Demonstrator, developed by the Illuminator team at TU Delft, is a modular, tabletop tool designed to educate and engage citizens in the benefits and challenges of the energy transition. It simulates important aspects of national electricity grids through a combination of Raspberry Pi control systems, 3D models, and LED visualizations. The B.Sc. graduation project focused on enhancing the demonstrator’s visualization and interaction capabilities. The project subgroup developed dynamic 3D models such as houses, windmills, cars, solar panels, carbon emissions, and batteries. On-screen elements are also created to represent key simulation agents like the battery state-of-charge and green energy percentage. Multiple dashboard versions were designed to tailor the experience to different stakeholder audiences. The report details the design, implementation, and validation of the visualized components. ... Read more

This project addresses the challenge of monitoring large, dynamic classrooms by proposing a privacyoriented multimodal data acquisition system tailored for MMLA. Traditional learning analytics rely on unimodal data and fail to capture complex classroom interactions. In contrast, MMLA leverages multiple data sources to better understand learning behaviors. Current systems lack adaptability, userfriendliness, and privacy considerations, impeding their integration into classrooms. The proposed system comprises static and dynamic nodes, with dynamic nodes worn by individuals and static nodes strategically placed in classrooms. Data features, selected on MMLA relevance, are transmitted wirelessly to the static node for storage and analysis. Privacy is prioritized by avoiding sensitive data collection and adhering to GDPR guidelines. The design ensures adaptability, supporting additional sensors and seamless integration into various educational settings. This foundational system enables future research while addressing ethical and technical challenges in large-scale classrooms.
... Read more

The goal of this project is to create a power supply for an electrical discharge machine (EDM). This machine can drill holes and create shapes in metal objects by evaporating material using discharges from the electrode to the workpiece. The electrode should be supplied with a high-voltage pulse wave in order to initate sparks. The power supply produces this waveform by turning a DC source on and off using a transistor as a switch. A current-limiting resistor that is placed between the source of the MOSFET and the electrode prevents short circuits during discharges. The circuit was first created with a DC voltage of 15V. A gate driver circuit to drive this MOSFET was designed to reduce power losses and to make the frequency and duty cycle adjustable. Then the design was improved by adding a diode to remove oscillations caused by the parasitic inductance of the power resistor. Afterwards, the circuit was built on a custom PCB to reduce parasitics. A filtering capacitor was added to remove high-frequency noise, after which a stable 15V pulse signal was obtained. A microcontroller is employed to regulate the pulse duration, detect spark occurrences, and relay this information to a speed regulator. This establishes a closed-loop system aimed at optimizing the overall system operation. To achieve this, circuitry is employed to convert the power supply output into a digital format, enabling the utilization of the microcontroller’s specialized hardware for rapid sampling. ... Read more

This paper presents the design and development of an Electrical Discharge Machining (EDM) device aimed at achieving precise hole creation in diverse metal materials. The EDM device was conceptu- alized and constructed based on specific requirements identified by the group. The design process encompassed four main stages: dielectric fluid selection, electrode design, control system develop- ment and power supply. In the dielectric fluid stage, the importance and criteria for selecting an appropriate fluid were discussed, resulting in the choice of distilled water for its superior dielectric properties. The electrode design stage followed a similar methodology, leading to the selection of a copper rod as the optimal electrode material. The control system stage detailed the development of an open-loop, manual, and closed-loop control system, emphasizing the utilization of Klipper software for precise electrode control. The power supply section outlined three primary circuits: the power source, power amplification circuit, and square wave generator circuit. Detailed schematics, component justifications, and optimization values for key parameters were provided to enhance power supply efficiency. Experimental evaluation demonstrated the capability of the EDM device to effectively create holes in various metals using an open-loop control system. Additional experiments focused on parameter variations within the power supply setup further illustrated their impact on machining performance. The discussion highlights the challenges encountered throughout the project, particularly the constraints imposed by limited time, which prevented the realization of all initially set requirements. Despite these challenges, the EDM device successfully met most of the specified objectives, showcasing promising results for future refinements and applications in precision machining. ... Read more

This thesis presents the design, implementation, and evaluation of the RF (Radio Frequency) section of an Open-Hardware Vector Network Analyzer (VNA) intended for quantum research applications. The project aims to create a cost-effective, modular VNA system that fulfills the functional requirements necessary for qubit readout and other quantum measurements.
In the initial chapters, the overall architecture of the VNA is outlined, with specific attention to the power budget and system requirements. The RF generation principles are examined, and a range of RF generators are tested to ensure they meet the signal quality standards, such as spurious emissions and harmonic content. The performance of various RF mixers is also evaluated and found to be sufficient for the RF system.
Experimental results demonstrate the system’s capability to measure the S21 parameter of a resonator cavity, comparable to commercial VNAs. This validates that the RF system meets the specified requirements and can be effectively used in quantum research.
Future work suggested includes the measurement of generator frequency/phase stability over time and exploring the feasibility of implementing power sweeps to enhance the system’s functionality. The findings of this thesis contribute to the development of accessible and flexible tools for quantum technology research, promoting further advancements in the field. ... Read more

This thesis, conducted as part of the Bachelor Graduation Project at TU Delft, focusses on designing the communication and storage components of a data acquisition system for Multimodal Learning Analytics (MMLA). The goal is to create an adaptable system that improves learning outcomes in large, dynamic classrooms such as Tellegen Hall. Current systems often fail to address issues of privacy and integration, limiting their effectiveness in real-world applications. Theproposeddesignemployswearabledynamicnodesfordatacollection, andutilisesstaticnodesand arootnodeforwirelesscommunication. WirelesscommunicationusestheESP-NOWwirelessprotocol, and data is stored in a time-series database, InfluxDB, running on a Raspberry Pi 5. A Graphical User Interface(GUI) built in Grafana, allows monitoring and visualises the data. The system was tested for reliability and performance, showing a high success rate of 99.93% in data transmission. It enables a network consisting of, theoretically, up to 200 nodes. Future improvements include real-time analytics and data security mechanisms. ... Read more

The focus of this report is the design and implementation of a universal gate driver for half-bridge circuits. This design includes a programmable dead-time controller and adjustable properties to ensure efficient and reliable switching on a variety of high-voltage applications. Furthermore, this work includes a comparative understanding of all design choices, such as the options considered for the dead-time controller, the gate driver component and the circuit of the output phase. The designed system is expected to receive as input a PWM signal and generate two output signals to drive a halfbridge configuration. The output pulses must be generated with a programmable dead-time and at the input switching frequency. The end product of this design will be used in the TU Delft Power Electronics Lab. The specific requirements are determined in order to meet the lab’s needs and this report further describes the process from acquiring the technical requirements, the circuit design, and finally, the prototype evaluation. ... Read more

The increase in non-linear loads of modern electronics raises concerns over power quality. Additionally, existing power-quality analyzers are expensive and not intended for household use. This thesis aims to develop a single-phase, user-friendly power-quality analyzer using a Raspberry Pi 4 Model B with an emphasis on low cost and class S specifications. The design was split into modules consisting of analog to digital conversion, voltage sensing, and current sensing. Sub-modules were added for circuit protection and PCB. Various approaches are discussed before circuit design, simulation, and testing occur. A functioning prototype was assembled on a dedicated PCB while not exceeding the set budget of €250.00. However, it could not be determined whether the class S specifications were achieved due to insufficient testing. A variety of improvements have been suggested. ... Read more

This thesis describes the design of a low-cost class-S power quality analyzer based on a Raspberry Pi 4. Capable of detecting power frequency, magnitude, voltage dips and swells, harmonics, and total harmonic distortion, the project consists of four main components: the communication protocol, the interface, the algorithms for power quality parameters, and the database. Each with its subdivisions. The communication module, which uses the I2C protocol, is chosen for its high sampling rates and built-in acknowledgement system, ensuring robust and fast operation. The interface module features a secure login process with two-step verification, a graphical user interface for real-time monitoring, and integration with algorithms for power quality parameter calculation. The algorithm module includes Fast Fourier Transform for harmonic detection, zero crossing method for power factor and frequency, and peak detection for voltage dips and swells. The database, powered by MariaDB on Raspberry Pi 4, securely manages the received data with restricted access for increased security, allowing remote access only from specified IP addresses. ... Read more

This thesis presents the design and development of a UV-C LED-based seed treatment machine aimed at enhancing seed quality by the extermination of pathogens. The research covers design choices, including a round irradiation pattern, consisting of two rings with three and nine LEDs for the inner and outer ring respectively, the use of a quartz plate as a holding plate for seeds for its high UV-C light permeability capabilities, the use of a vibration motor underneath the main operational stack for seed movement, and the use of Ethernet ports for power distribution and communication. The thesis discusses a comparative study between square and circular plate configurations, evaluating their performance using simulation results. Safety considerations were prioritized in the design, and appropriate precautions were implemented throughout the design process. The thesis also highlights the iterative design process for the mechanical system, discussing challenges encountered and improvements made to achieve a functional and robust prototype. Results demonstrate successful integration of components and achievement of objectives. The thesis concludes with discussions on the strengths, limitations, and future enhancements of the UV-C LED-based seed treatment machine. The research presented in this thesis provides valuable insights for further advancements in seed treatment technology, contributing to sustainable agricultural practices. Due to time constraints, conclusive results of testing on seed with this machine could not be obtained yet. ... Read more