No abstract

Wireless power transfer systems have been widely applied in the field of portable and implantable devices, featuring contact-free and reliable energy supply. Novel implant systems, such as brain-computer interfaces, impose the challenges of strong miniaturization and operation under loosely coupled conditions. Therefore, maximizing power transfer efficiency while decreasing the size of transmitter and receiver structures becomes a central research question. This paper presents a unified design strategy of modeling, analyzing and optimizing planar spiral coils with integrated capacitive elements, so-called capacitively segmented coils, for operation in wireless power transfer interfaces. It mathematically analyzes and experimentally verifies that the combination of capacitive coil segmentation, increased operational frequencies and geometrical coil optimization can be used to establish wireless power transfer links with comparatively high efficiency, small size and limited detuning effects in lossy dielectric environments. The paper embraces the formulation and verification of a broadband analytical link model based on partial element equivalent circuits, which is subsequently used to determine dominant coupling and loss mechanisms and to optimize the coils' geometries for high efficiency. Moreover, an extended analysis shows how the capacitive coil segmentation can effectively suppress dielectric losses and non-uniform current distributions by canceling the inductive contribution of every coil segment at the frequency of operation. Utilizing these methods, an exemplary 40.68 MHz wireless power link with a 30 mm primary and a 10 mm secondary coil is designed and evaluated: With a maximum efficiency of up to 31 % in biological tissue at 20 mm separation distance, it features efficiency levels which are up to ten times higher and a specific absorption rate which is up to five times lower compared to non-segmented systems. When operated at 150 MHz in air, efficiency levels are up to 1.5 times higher than in state-of-the-art systems of the same size.

Sensor nodes often have to work for a long time with a single battery. A change of the power source is sometimes not possible or involves effort and costs. If the node needs to be accessible for communication at any point of time it must have a radio in permanent receive state. This depletes the battery in a few days. The contradiction between long lifetime and permanent accessibility can be solved by using a separate wake up receiver on the node. In this work we present a sensor node with included wake-up receiver working in the 433 MHz ISM band. Our solution consumes 2.81lA of current in sleep state while still maintaining a realtime behaviour. The low current consumption is achieved by modulating a 125 kHz wake-up signal on the 433 MHz carrier in the sender. In the receiving node a passive demodulation circuit extracts the wake-up signal and feeds it to an 125 kHz low frequency receiver IC. An additional 16 bit address coding is used for an selective wake-up of nodes.

No abstract

The evolution of microelectronics increased the information acquired by today’s biomedical sensor systems to an extent where the capacity of low-power communication interfaces becomes one of the central bottlenecks. Hence, this paper mathematically analyzes and experimentally verifies novel coil and transceiver topologies for near-field communication interfaces, which simultaneously allow for high data transfer rates, low power consumption, and reduced interference to nearby wireless power transfer interfaces. Data coil design is focused on presenting two particular topologies which provide sufficient coupling between a reader and a wireless sensor system, but do not couple to an energy coil situated on the same substrate, severely reducing interference between wireless data and energy transfer interfaces. A novel transceiver design combines the approaches of a minimalistic analog front-end with a fully digital single-bit sampling demodulator, in which rectangular binary signals are processed by simple digital circuits instead of sinusoidal signals being conditioned by complex analog mixers and subsequent multi-bit analog-to-digital converters. The concepts are implemented using an analog interface in discrete circuit technology and a commercial low-power field-programmable gate array, yielding a transceiver which supports data rates of up to 6.78 MBit/s with an energy consumption of just 646 pJ/bit in transmitting mode and of 364 pJ/bit in receiving mode at a bit error rate of 2×10−7, being 10 times more energy efficient than any commercial NFC interface and fully implementable without any custom CMOS technology.

This paper presents a wireless passive strain sensing concept that functions by detuning a dielectric resonator. It is shown how a high Q resonator functions as a wireless passive sensor when correctly matched with an antenna. Finite element and analytical models are compared with experimental data and the sensor cross sensitivity with respect to temperature and humidity are also explored. The sensitivity of the resonance frequency to the strain, temperature and humidity is measured to be 51.6 ppm/µm, 10.09 ppm/K and -0.65 ppm/% respectively.

Near-field interfaces with miniaturized coil systems and low output power levels, such as applied in biomedical sensor systems, can suffer from severe efficiency degradation due to dynamic impedance mismatches, reducing battery life of the power transmitter unit and requiring to increase the level of electromagnetic emission. Moreover, the stability of weakly-coupled power transfer systems is generally limited by transient changes in coil alignment and load power consumption. Hence, a central research question in the domain of wireless power transfer is how to realize an adaptive impedance matching system under the constraints of a simultaneous power feedback to increase the system’s efficiency and stability, while maintaining circuit characteristics such as small size, low power consumption and fast reaction times. This paper presents a novel approach based on a two-stage control loop implemented in the primary-side reader unit, which uses a digital PI controller to maintain the rectifier output voltage for power feedback and an on-top perturb-and-observe controller configuring the setpoint of the voltage controller to maximize efficiency. The paper mathematically analyzes the AC and DC transfer characteristics of a resonant inductive link to design the reactive AC matching network, the digital voltage controller and ultimately the DC-domain impedance matching algorithm. It was found that static reactive L networks result in suitable efficiency levels for coils with sufficiently high quality factor even without adaptive tuning of operational frequency or reactive components. Furthermore, the regulated output voltage of the rectifier is a direct measure of the DC load impedance when using a regular DC/DC converter to supply the load circuits, so that this quantity can be tuned to maximize efficiency. A prototype implementation demonstrates the algorithms in a 40.68 MHz inductive link with load power levels from 10 to 100 mW and tuning time constants of 300 ms, while allowing for a simplified receiver with a footprint smaller than 200 mm2 and a self-consumption below 1 mW. Hence, the presented concepts enable adaptive impedance matching with favorable characteristics for low-energy sensor systems, i.e., minimized footprint, power level and reaction time.

No abstract