Wissenschaftliche Publikationen "Elektronik"

Ziel dieses Projektes ist es, durch den Einsatz von alkalischen Polymerelektrolytmembranen (engl.: anion exchange membrane, AEM) in Alkohol-Brennstoffzellen die Crossover-Leistungsverluste von protonenleitenden Membranen (engl.: proton exchange membrane), PEM) in diesem Einsatzbereich zu unterbinden. Während in protonenleitenden direkten Alkoholbrennstoffzellen (PEM-DAFC) durch den Protonentransport von der Anode zur Kathode Alkohol mitgeschleppt werden kann, findet in AEM-DAFCs der Ionenausgleich zwischen Anode und Kathode durch Hydroxid-Ionen in Gegenrichtung von Kathode zu Anode statt. Durch die Vermeidung des Mitschleppefekts kann so der Durchtritt des Alkohols durch die Membran reduziert werden. In Alk2Micro soll zunächst eine Membran-Elektroden-Einheit (MEE) entwickelt werden, die das Herzstück der Brennstoffzelle darstellt. Die MEE besteht aus der AEM, einer Katalysatorschicht für die Alkoholoxidation und Sauerstoffreduktion, sowie porösen Kohlenstoffsubstraten. Jede Komponente an sich muss für eine optimale Leistung und Lebensdauer optimiert werden, während die Gesamtkosten im Blick behalten werden müssen. Mehrere MEEs sollen dann in einem Stack zusammengeführt werden. Um die MEEs optimal und langlebig zu betreiben, soll ein Forschungsschwerpunkt hier auf korrosionsbeständige Bipolarplatten bzw. Beschichtungen gelegt werden.

In this paper the development of a compact condition classification system for electromagnetic proportional valves is shown. It allows the generation of training data as well as a fast testing and comparison of different trained neuronal networks. By using quantization and pruning, a neuronal network with drastically reduced complexity has been created, so an FPGA implementation was possible. The developed and implemented network shows a very high classification rate and can distinguish 12 different error reasons of the valves. The system requires the measurement of the supply current only, which allows a simple integration of such a false detection circuitry into existing systems. In the future, the system can be modified easily, e.g. to use and test a hardware based AI accelerator instead of the FPGA implementation.

 

Link zum Paper

It is mandatory for manufacturers of reusable medical devices to specify the maximum of allowed sterilization cycles. This work focuses on the fabrication of an autonomous sterilization cycle counter in MEMS technology. Surface microm-achining is applied utilizing vapor HF (vHF) etching and subsequent polymer anti-stiction coating as an efficient tech-nique to remove sacrificial buried oxide layers and to release micro structure without stiction. The vHF etching process has been optimized and reaches etch rates of 900 nm/min for the buried oxide with a uniformity of more than 95 % across a 100 mm wafer. Based on this process a compatible design with perforated structures is implemented. Finally, devices of the sterilization cycle counter are successfully tested by a simulated sterilization cycle temperature test.

 

Link to publication

Piezoresistive accelerometers are a popular and widely used method of acceleration sensing due to their simplicity in fabrication, packaging and inherent ruggedness. These devices usually have simpler electronics for example compared to capacitive sensors. There are products available without ASICs as well. [1] This work takes an existing high-g piezoresistive accelerometer as a reference sensor and works on optimizing this design. The work delves into different design and optimization techniques to deliver novel high-performance piezoresistive acceleration sensor prototypes. These techniques exploit the device physics and mechanics to point towards the new designs. The sensor is designed and optimized to have higher bandwidth. The adaptation of SOI wafers instead of currently used epitaxial wafers, results in smaller chip size, enabling more chips/wafer and reducing the sensor cost. Extra emphasis has been laid to maintain high sensitivity, for specified doping of the piezoresistors, while still not compromising much on the resonance frequency. Finite element analysis with ANSYS Workbench offers insights into the device performance and helps in the determination of some key sensor parameters for the new designs. It also helps to establish trends and this provides an idea for the successive models. After several iterations and incorporating different optimization techniques, new models are achieved.

 

Link to publication

In this paper we present a hybrid microsystem for acquisition and counting of sterilisation cycles. The device includes a micromechanical counter mechanism and a thermal actuator based on a shape memory alloy (SMA). The device is designed to count 100 sterilization cycles. The basic functionality is investigated on a hotplate using a thermal temperature profile with a peak temperature of 135 deg C. In this manner, counting of thermal cycles is demonstrated.

 

Link to publication

This work provides an analytical non-linear model for the capacitive transduction in MEMS transducers with perforated counter-electrodes, especially applicable to capacitive MEMS microphones. Starting from an electrostatic description of a perforated unit cell of the transducer, analytical formulations of the variable capacitance and electrostatic forces are derived, accounting for the deflection profile of a clamped circular plate. A lumped implementation into conventional circuit simulations tools is enabled via behavioral modeling based on hardware description languages, such as Verilog-A. Therefore, the analytical model is approximated via Taylor series expansions, allowing for a stable and non-linear behavioral implementation. The resulting model finally enables both a small- and large-signal analysis of capacitive MEMS microphones, precisely accounting for non-linearities in the capacitive transduction. This allows to simulate the harmonic distortion of the microphone's output signal and to account for electrostatic spring-softening in simulations of its bias voltage dependent sensitivity.

 

Link to publication

MEMS are minute electro-mechanical systems that are able to sense small changes in the environment and are responsible for a large chunk of the sensor market. MEMS inertial sensors such as gyroscopes and accelerometers have helped in realizing new products across a multitude of fields. One of the next aspects of MEMS Inertial sensors is the Internet of Things [IOT] and Industry 4.0. Hence, the following work depicts a sensor element design for vibration detection in Out Of Plane [OOP] axis (z-axis) proposed for closed loop control considering a system bandwidth of 20 kHz. The current work highlights the concept of an "anti-phase" design type sensor element developed considering either top or bottom electrodes allowing for the possibility of closed loop control. The translational motion of the sensor element along with the larger mass elements provides higher performance and lower noise. The analysis indicates a sensor element with size 1mm*1mm showcasing a specific open loop sensor sensitivity per area of 1.12 nF/g/m 2 with resonance frequency of 10.5 kHz and can be considered as a viable alternative to see-saw design type.

 

Link to publication