Search

Measuring the actual junction temperature ${T}_{J}$ of power light emitting diodes used in high-end luminaires is important for implementing measures to reduce the stress and therefore extend the lifetime of these fundamental light sources. The direct measurement of this key parameter can be performed by imposing a known and precise forward current ${I}_{{probe}}$ through the LED, and tracking the bias voltage appearing at its terminals. The technique proposed in this paper is based on the same approach, but does not require a stable current source, relying instead on the exponential fitting of random I-V measurements and on the estimation of the bias voltage at the proper ${I}_{{probe}}$ through a mathematical interpolation. The technique is implemented on a custom-designed circuit, and experimental data are obtained on commercial power LEDs, allowing to assess the impact of ${I}_{{probe}}$ and sampling current range on the measurement error.

There is an increasing interest about indoor positioning, which is an emerging technology with a wide range of applications [...]

Indoor positioning of objects and people is becoming of great importance in the Internet of Things (IoT), in-home automation, and navigation in malls, airports, or very large buildings. Positioning is determined by multiple distance measurements between reference points and sensors. Distance measurement uses the time of flight of an ultrasonic signal traveling from an emitter to receiving sensors. This requires close synchronization between the emitter and the sensors and a sharp time resolution of the time of arrival (TOA) of the ultrasonic signal. Usually, TOA is detected using cross-correlation processing requiring significant computational resources at the sensors level. In this work, the synchronization is done using the RFID standard protocol features. The TOA detection is performed firstly by training off-line a Machine Learning model using as input the peaks indexes of the ultrasonic signal received and the output of a cross-correlation-based positioning system, as ground-truth. In a second phase, the positioning is evaluated and tested on-board using the previously trained model on a microcontroller. The system architecture is presented and experimental results on the positioning accuracy are shown accordingly. Results show a mean positioning error below 25 cm in 95% of the positionings in a typical room.

In the last few years, the wide diffusion of rechargeable devices has fueled the research interest in wireless power transfer (WPT) technology that offers advantages such as safety, flexibility, and ease of use. Different standards have been developed over the years but a significant part of the global interest is focused on the inductive resonant wireless power transfer. By increasing the resonance frequency, an improvement in the transfer efficiency between transmit and receive coils is generally observed, at the expense, however, of an increase in losses in the switching devices that constitute the transmitting and receiving circuits. This study concerned the performance evaluation of a WPT transmitting circuit built using Gallium Nitride (GaN) or conventional silicon (Si) switching devices, to assess their specific contribution to the overall efficiency of the system. The overall performance of two circuits, respectively based on GaN HEMTs and Si MOSFETs, were compared at frequencies of the order of MHz under different operating conditions. The theory and design choices regarding the WPT circuit, the coils, and the resonant network are also discussed. The comparison shows that the GaN circuit typically performs better than the Si one, but a clear advantage of the GaN solution cannot be established under all operating conditions.

n the field of industry 4.0, one of the sectors in which research is particularly active is the area of Predictive Maintenance(PdM), the purpose of which is to improve the industrial production process. This type of maintenance aims to predict a possible failure event, reduce non-production times and increase the quality of the processing result. The objective of this paper is to select the best Machine Learning models for a PdM application. In particular, such a model should allow making a prediction based on a real dataset, obtained by monitoring a turning process, with the aim of making the classification of the chip shape. The criteria used to choose the best model are accuracy and prediction speed (to reduce the inference time). Indeed it is crucial to spot any potential machine fault in the shortest time possible, in order to intervene before the machine fails. Hence, our goal is to choose the ML models with a lesser inference time while still maintaining high accuracy.

Indoor positioning of objects and people has becoming of great importance in the Internet of Things (IoT), in-home automation, and navigation in mall, airports or very large buildings. Position can be geometrically calculated from multiple measurements of the distance between reference points and moving sensors. Distance measurement can be done by measuring the time of flight of an ultrasonic signal traveling from an emitter to receiving sensors. This requires close synchronization between the emitter and the sensors. In this work this synchronization is done using the RFID standard protocol features. RFID tags included in the emitter and sensor circuitry are synchronized each other via the RFID reader interrogation pulse. The system architecture is explained and experimental results on the synchronization accuracy are presented. Moreover, the synchronization uncertainty influence on the positioning performances is discussed. Calculations ensure a mean positioning error below 5 cm in the volume of a typical room.

Connected cars and automotive car driving are emerging topics and many researchers are looking to provide new solutions and technologies that could revolutionize the driving experience. However, human car-driving is an activity that will certainly interact, in the near future, with autonomous driven vehicles. Hence, the need to provide an information layer about driving behavior and style that could benefit a traffic management system. Both insurance and car rental providers could also benefit of the knowledge of the driving behavior to custom fit insurance and rental rate. In this work, a machine learning model was deployed on a tiny hardware for performing the classification of car driving style based on inertial measurement data, augmented with in-vehicle diagnostic. The results show that an accurate (98%) model could be successfully deployed on an ultra-low-power STM32L4 Series MCU (MicroContoller Units) based on Arm® Cortex®-M4 core micro controller, capable to infer the proper classification among 4 classes in less than 5 ms after acquiring data from sensors. The system is also able to detect start and stop occurring during normal traffic driving.

Gesture recognition is a recently emerged technology that aims to redesign the interaction between humans and machines. Existing solutions usually recognize the gestures based on camera vision, wearable sensors or specialized signals (e.g., WiFi, acoustic and visible light), but they cope with the drawbacks of high energy consumption or reduced SNR with respect to the surrounding environment, which prevents them from accurately sensing finger movements. In this work, we propose a device-free gesture recognition system that use machine learning techniques on RSSI and phase values from the backscattered signals of an array of tags placed on a plastic plate to recognize different hand gestures. The system has been tested with RFC and SVM algorithms, providing excellent results both in case of stationary hand that in case of moving hand.

A detailed study about the direct measurement of junction temperature T_{J} of off-the-shelf power light emitting diodes (LED) is presented. The linear dependence on temperature of the voltage drop across the device terminals at a constant current is in particular exploited and fully characterized, in the temperature range from T = 35,,^{circ } ext{C} to 135 °C, with tests repeated at one thousand different probe currents. The accurate experimental data, obtained on several LED samples bearing two different part numbers, are reported, showing that they exhibit a high degree of linearity in wide current ranges, a circumstance that allows for a fast and reliable calibration as sensors. The measurement error is also fully characterized in terms of repeatability and stability over time, with measurements repeated after 600, 900, 1200 and 1800 hours of applied electro-thermal stress, demonstrating that the relevant sensor parameters stabilize after a few hundred hours of operation. A full set of parameters is provided for the two device models, allowing the direct use of each LED for the self-monitoring of the junction temperature and ensure compliance with their safe operating area over time. Moreover, a procedure and a simple circuit for the real time measurement of T_{J} , while the LED is on, are presented. The procedure does not require a stable current source, and relies instead on the application of a sub-threshold current ramp for such a short time that the change in the output light is not perceived by human eyes.

Indoor positioning is an emerging technology with wide applications. Augmented and mixed reality need accurate and real-time positioning of user’s limbs or direction of sight in real time, and reference points within the real environment. Positioning is also required for mall navigation, elderly people movements monitoring, and many others contexts. Indoor positioning of mobile units (MUs) can be provided by multilateration techniques that compute the location of an MU starting from distance measurements between the MU and a set of beacons. In this paper, a nonlinear closed-form solution for the trilateration problem is employed to avoid heavy numerical iterative algorithms. The closed-form solution is allowed by a particular arrangement of the beacons, placed at the vertex of a square. A positioning system is presented where MUs, exploiting the light algorithm of the closed-form positioning, are able to autonomously and privately calculate their own positions. A prototype of the positioning system has been designed, realized, and characterized for an average $4 \times 4 \times 3\,\,\text {m}^{3}$ home or office room. It includes four beacons that emit a sequence of ultrasonic chirp signals, several MUs, and a master unit that provides time synchronization via ANT transceivers onboard each MU and the master unit. Thanks to the adopted closed form solution, each tiny battery-operated MU is able to carry out all the computations onboard, including analog and digital signal processing, beacon-MU distance estimation, and finally MU positioning at a rate of 2 Hz.

Positioning can be calculated from measurement of the distances between multiple reference points and moving sensors. A new technique to measure the distance between emitter and receiver by using ultrasonic signals in air is proposed, without the use of the time of flight measurement based on expensive time synchronization between emitter and receiver. The attenuation level that the ultrasonic signal suffers at different frequencies provides information on the emitter-receiver distance. A mathematical relationship between emitter and receiver distance and the attenuation level experienced along the traveled path by two ultrasonic signals is provided. The actual level of attenuation in the air for the employed frequencies is measured online during the ranging operations. The simulations show the effectiveness of the technique under feasible operational conditions. In particular, an error of less than ±2.5 cm is reached along two significant paths of the vertical section of a typical office room.

A circuit is presented for the on-line measurement of the junction temperature (T J ) of power LEDs. The circuit is integrated with, and makes use of, an off-the-shelf LED driver, so that the proposed architecture can be easily made compatible with existing high-end luminaire platforms. The measurement technique, based on the linear junction voltage vs. temperature relationship, shows good precision and repeatability among devices with the same part number, with an absolute error below 3 °C.

Catastrophic and accidental natural events (e.g., earthquakes, sudden floods, fires, etc.) and deficiencies in appropriate management activities require proper plans for disaster management and real-time definition of safe paths and escape routes. Emergencies can affect the structural health status of structures and infrastructures and, consequently, the safety in highly populated areas (e.g., urban contexts, occasional assembly points for crowds, theme parks, etc.) is compromised. Consequently, the use of a decision platform for the management of structures and infrastructures, which is able to gather real-time data from sensors and provide alerts and augmented information about safe paths, paves the way to the adoption of a proactive form of risk management. In this work, objectives were confined into the validation of a Near-Field Communication (NFC) system. Data coming from user smartphones using the NFC technology and sensor data deriving from the decision platform were used as input in a Machine Learning (ML) algorithm. Results show that the ML algorithm mentioned above is able to refine the safe path recognition strategy defined from the platform. In addition, the bidirectional and touch less NFC technology allows the delivery of alerts and disaster plan dissemination, also in the case of connectivity shutdown.

Expected and unexpected events affect road infrastructure leading to several issues that require maintenance interventions or prompt actions. In order to face the issues mentioned above, specially designed platforms can be used. Despite the high number of solutions presented in the last decades, it is difficult to find real applications of systems deputed to the emergency management of the abovementioned infrastructures. This paper aims at describing an innovative platform and the results of its first application. The platform consists of a system and algorithms that allow environmental and structural monitoring and emergency management. The proposed system consists of a network of wireless sensor nodes that are self-powered by mean of photovoltaic energy harvesters. Two different types of algorithms were designed to monitor the environmental and structural condition (i.e., the vibro-acoustic signature) of different types of structures and infrastructures, and to display, thanks to a dashboard, several predefined alarms, escape routes, safe place, and gathering points. The results presented in the paper refers to the first application of the platform, here tailored for an Italian theme park in the framework of the project "SICURVIA" funded by Regione Calabria, Italy. These results, presented for the first time, show that the platform is able to extract crucial information about the conditions of the thematic park’s assets, and, at the same time, this information are used to recognize an emergency, trigger the related alarm, and provide a plan to manage the extraordinary event happened.

Light Emitting Diodes (LEDs) are the longest lasting source of artificial illumination whose duration can exceed 50.000 continuous working hours. Nevertheless, they show a gradual reduction of the luminous flux due to the increase of the device temperature. In this work, a Machine Learning algorithm will be introduced and discussed, able to predict the junction temperature value of a LED in real-time while connected in the end-user circuit, taking into account current and voltage flowing in the device and, further, the actual model and aging of the LED. The algorithm was implemented on a microcontroller, showing the feasibility of performing edge machine learning on tiny yet powerful devices.

Several characteristics of the human body turn into postural behavior, recognizable also during sport activities. The presence of differences between body types could lead to different behavior of wearable and fitness-devote products. A new wearable based on machine learning techniques for the exercise detection and repetitions count is described in this work. A proper dataset has been obtained in order to offline train the network. Eventually, the machine learning algorithm has been implemented inside an edge device for real-time test e verification.

The aim of this work is to provide a comprehensive view on sensors technology for intraoral forces monitoring. State-of-the-art electronic sensors, exploitable for the application of measuring intraoral human forces are compared in this work. Furthermore, a possible configuration for data acquisition from multiple sensors, using Wheatstone bridges to detect the resistance variation and the force output of miniature strain gauges, is also provided.

Indoor localization and tracking of persons and assets with centimeter-level accuracy for inventory, security, medical monitoring, and training, as well as gesture interfaces for domotics, is highly desirable in the framework of the emerging IoT paradigm. Low cost, tiny, battery, or batteryless operated position sensors are required. 3-D localization can be computed by combining three or more distance measurements between sensor and reference points. Our aim is to give the capability of measuring the distance from a reference point to radio frequency identification (RFID) tags. The main challenge is in the estimation of the distances with millimeter accuracy in presence of both size and power supply strict constraints, and thus with very limited computational power. An accurate ranging technique using cross-correlation and small RFID-based sensors is proposed. Its originality resides in moving the main computational efforts from the sensor to an external processing unit with sufficient computational and supply power, thus overcoming the sensor limits. The system is composed of a beacon that emits ultrasound chirps and RF sync signals, a RFID-based distance sensor, a commercial RFID reader, and a processing unit. Main advantages are the high miniaturization and low power consumption of the remote sensor, and its compatibility with existing RFID standards.

Indoor localization and tracking of persons and assets for gaming, augmented reality, medical monitoring, training, security, and inventory is an emerging technology. Localization can be computed from multiple distance measurements between reference beacons and sensors. Low cost, small, light, tiny-battery, or even batteryless operated sensors are required. Ranging can be carried out by measuring the time of flight of an ultrasonic chirp traveling from an emitting beacon to receiving sensors. That requires a tight synchronization between the beacon and the sensors. In this article, we present a ranging system based on miniature commercial modules featuring ANT wireless personal area network communication plus onboard microprocessor. A protocol is developed for microseconds synchronization and accurate onboard ranging computations are carried out through ultrasonic signal cross correlation. The sensor privately computes its own range. The ranging system is composed of a synchronizing unit, a beacon emitting ultrasonic chirps, and a battery-operated distance sensor. The sensor shows low average power consumption (less than 22 mW) and a ranging accuracy of about 1.8 mm for distances up to 2.5 m, under stable environmental conditions.

A device for wireless intraoral forces monitoring is presented. Miniaturized strain gauge sensors are used for the measurements of forces applied by tongue and lips. A sensor interface IC is able to multiplex among four sensors and a low energy transmission module, equipped with an ARM Cortex–M0 core, is used for signal elaboration and remote wireless data transmission using Bluetooth® Low Energy standard protocol. The main novelty rely in the dynamic correction of the output corrupted by the prestrain issue. Moreover, the device shows a reduced dimension and the ability to transmit data wirelessly, without the use of external cables normally used in state-of-the-art intraoral monitoring devices.

The use of 4H-SiC junction devices as effective temperature sensors is demonstrated by measuring, elaborating and comparing the experimental I- V characteristics of three different devices, namely pin and Schottky diodes and the body diode of a commercial power MOSFET. Due to the strong dependence of the I-V characteristics on temperature, the sensor performs a high sensitivity. It is shown that the correct choice of the probe current during sensing is fundamental to obtain an excellent linearity with T in the sensor response.

Energy harvesting at radio frequencies (RF) is increasingly attracting interests because of the simplicity of its concept: collecting the RF energy through an antenna and a rectification circuit. This enables power-over-distance transmission, recharging of devices and communications and many other applications. However, the transferred power levels are inadequate due to uncertainties on the scenario and a non efficient confinement of the power transfer. In this work, the feasibility of a radically new approach based on RF field intensity focusing in unknown scenarios is discussed. The proposed approach is, in principle, able to maximize the energy transfer as it aims at confining the RF power on the considered RFID device. Moreover, being the approach based on a real-time procedure, it is able to deal with moving RFID tag embedded into an unknown scenario. Finally, the achieved spatial confinement will deliver private Peer to Peer communication.

The conversion efficiency of a multi-stage RF energy harvester on a printed circuit board charge-pump, based on off-the-shelf diodes and capacitors is studied, in the UHF band, as a function of temperature. The considered temperature range varies from 25°C to 85°C, highlighting that the effects of the temperature may cause a severe degradation on the harvested power, in particular at the lowers incident power regimes. Experimental measurements are presented to show that the rectifiers quality has the biggest impact of the harvester performances. Precisely, the temperature dependent rectification ratio of silicon Schottky diodes commonly used for this application can be considered as a quality factor of the converter.

In this paper, a fully passive RF Energy Harvester is presented, to be used in the UHF band. The system includes an antenna, a Dynamic Impedance Matching Network (DyIMN), a diode-based voltage rectifier and multiplier, a micropower DC/DC converter circuit, a storage capacitor and the load electronic, comprising the MCU. The DyMN is comprised of varactors that dynamically change their junction capacitance, controlled by an MCU, allowing the impedance tuning of the matching stage between antenna and rectifier.

The use of non-destructive techniques (NDT) for structural health monitoring (SHM) of civil infrastructures is crucial especially for those structures, like buildings and roads, that need efficient and reliable emergency management systems. For this reason, this study aims at designing and validating a new sensor-based NDT platform for SHM, emergency management, and decision support. Different structures of a thematic park have been monitored. The effects of different events (e.g., seismic) and failures (i.e., cracks) have been investigated using a wireless monitoring system consisting in several sensor units (properly placed), wirelessly linked to one controlling unit. The set-up of the software platform has been tuned using the data gathered through in-lab simulations and on-site experiments. Results show that the hardware and software platform paves the way to automatic detection of failures, to identify which infrastructures are still available around the monitored areas, and to generate a proper emergency plan.

Radiative (or far-field) energy replenishment for devices such as smartphones, laptops, robots, and small electric appliances paves the way to autonomous and continuous devices functioning, thus bypassing the need of operation interruptions, human maintenance activities, and replenishment by wired transformers. In this work, we investigate the feasibility of using a properly engineered antenna array able to deliver radiative power to devices in need of energy replenishment during their normal and unsupervised activity, whose locations are unknown. Both the case of single and multiple devices needing energy replenishment are addressed. A quantitative proof-of-concept study is carried out to validate the proposed approach. A 3D scenario is simulated to study the case of devices in need of energy replenishment within a standard office environment. Different antenna array configurations are investigated and the corresponding performances benchmarked against a standard installation of recharging antennas. Results confirm the outstanding capability of the proposed approach in terms of confinement and maximization of power transfer. Finally, in this framework, we also propose an efficient communication protocol that is able to manage multiple recharge demand given different operational rules.

The recent growing interest in indoor positioning applications has paved the way for the development of new and more accurate positioning techniques. The envisioned applications, include people and asset tracking, indoor navigation, as well as other emerging market applications, require fast and precise positioning. To this end, the effectiveness and high accuracy and refresh rate of positioning systems based on ultrasonic signals have been already demonstrated. Typically, positioning is obtained by combining multiple ranging. In this work, it is shown that the performance of a given ultrasonic airborne ranging technique can be thoroughly analyzed using renowned academic acoustic simulation software, originally conceived for the simulation of echographic transducers and systems. Here, in order to show that the acoustic simulation software can be profitably applied to ranging systems in air, an example is provided. Simulations are performed for a typical ultrasonic chirp, from an ultrasound emitter, in a typical office room. The ranging performances are evaluated, including the effects of acoustic diffraction and air frequency dependent absorption, when the signal-to-noise ratio (SNR) decreases from 30 to −20 dB. The ranging error, computed over a point grid in the space, and the ranging cumulative error distribution is shown for different SNR levels. The proposed approach allowed us to estimate a ranging error of about 0.34 mm when the SNR is greater than 0 dB. For SNR levels down to −10 dB, the cumulative error distribution shows an error below 5 mm, while for lower SNR, the error can be unlimited.

A numerical simulation study accounting for trap and defect effects on the current-voltage characteristics of a 4H-SiC-based power metal-oxide-semiconductor field effect transistor (MOSFET) is performed in a wide range of temperatures and bias conditions. In particular, the most penalizing native defects in the starting substrate (i.e., EH6/7 and Z1/2) as well as the fixed oxide trap concentration and the density of states (DoS) at the 4H-SiC/SiO2 interface are carefully taken into account. The temperature-dependent physics of the interface traps are considered in detail. Scattering phenomena related to the joint contribution of defects and traps shift the MOSFET threshold voltage, reduce the channel mobility, and penalize the device current capabilities. However, while the MOSFET on-state resistance (RON) tends to increase with scattering centers, the sensitivity of the drain current to the temperature decreases especially when the device is operating at a high gate voltage (VGS). Assuming the temperature ranges from 300 K to 573 K, RON is about 2.5 MΩ·µm2 for VGS &gt, 16 V with a percentage variation ΔRON lower than 20%. The device is rated to perform a blocking voltage of 650 V.

Smart cities need technologies that can be really applied to raise the quality of life and environment. Among all the possible solutions, Internet of Things (IoT)-based Wireless Sensor Networks (WSNs) have the potentialities to satisfy multiple needs, such as offering real-time plans for emergency management (due to accidental events or inadequate asset maintenance) and managing crowds and their spatiotemporal distribution in highly populated areas (e.g., cities or parks) to face biological risks (e.g., from a virus) by using strategies such as social distancing and movement restrictions. Consequently, the objective of this study is to present an IoT system, based on an IoT-WSN and on algorithms (Neural Network, NN, and Shortest Path Finding) that are able to recognize alarms, available exits, assembly points, safest and shortest paths, and overcrowding from real-time data gathered by sensors and cameras exploiting computer vision. Subsequently, this information is sent to mobile devices using a web platform and the Near Field Communication (NFC) technology. The results refer to two different case studies (i.e., emergency and monitoring) and show that the system is able to provide customized strategies and to face different situations, and that this is also applies in the case of a connectivity shutdown.

In a few years, the world will be populated by billions of connected devices that will be placed in our homes, cities, vehicles, and industries. Devices with limited resources will interact with the surrounding environment and users. Many of these devices will be based on machine learning models to decode meaning and behavior behind sensors’ data, to implement accurate predictions and make decisions. The bottleneck will be the high level of connected things that could congest the network. Hence, the need to incorporate intelligence on end devices using machine learning algorithms. Deploying machine learning on such edge devices improves the network congestion by allowing computations to be performed close to the data sources. The aim of this work is to provide a review of the main techniques that guarantee the execution of machine learning models on hardware with low performances in the Internet of Things paradigm, paving the way to the Internet of Conscious Things. In this work, a detailed review on models, architecture, and requirements on solutions that implement edge machine learning on Internet of Things devices is presented, with the main goal to define the state of the art and envisioning development requirements. Furthermore, an example of edge machine learning implementation on a microcontroller will be provided, commonly regarded as the machine learning “Hello World”.

A study on power light-emitting diode (LED) junction temperature sensing based on the measurements of its forward voltage is presented. The linear dependence on temperature of the voltage drop across a power LED in the wide temperature range from T = 35 up to 175°C, for different bias currents, is investigated. The experimental measurements., repeated on three devices with the same parts number, exhibit a good degree of linearity. Moreover, the proposed sensors have excellent performances at the bias current of 1.4 RNA with good repeatability and maintaining a stable output over more cycles of measurements.

A system made of a custom designed heat flow sensor, a thermo-electric heat-pump, and a microcontroller based control board is presented for the measure of the power losses in power switches, e.g. MOSFETs or BJTs, operating in dynamic repetitive regime. The heat-flux sensor consists of two highly linear proportional to absolute temperature (PTAT) CMOS sensors. The thermoelectric module holds the case of the device under test at ambient temperature, so that heat entirely flows through the calorimeter. As the dissipated power measurement relies on the measurement of a heat flow, the apparatus is particularly suitable, and provides reliable results, in fast switching devices, regardless of the switching frequency, thus overcoming the problems associated with direct electrical measurements.

An RF powered HF-RFID passive sensors platform has been realized using discrete components and printed antennas designed to resonate at 13.56 MHz, used both for energy harvesting and data transmission. The tests demonstrate the possibility of the system to operate autonomously within the reading range of a standard RFID reader, that acts both as RF power source and receiver of the data stored in the tag user memory. The microcontroller on the platform can be interfaced with an analogic sensor made with polymeric material sensible to physical parameters or chemical agents.

The design of a complete photovoltaic (PV) system and the precise evaluation of its performances under different conditions requires the use of an accurate simulation model. In this work, a SPICE model of a complete PV system, including a detailed model of PV cells and a multilevel inverter with load and energy storage elements, is presented. The simulation of the system as a whole allows determining the role of components and parameters as solar irradiation and cell temperature on the system performance. The global conversion efficiency and the total harmonic distortion (THD) of the output waveform are analysed in detail. The role and sizing criteria of storage elements placed in parallel with the PV modules are also analysed for various operating conditions. A PV system, including a specifically designed multilevel inverter, was built to perform efficiency and THD measurements that were compared with the simulation results to evaluate the model effectiveness.

A wireless CMOS temperature sensor chip performing an embedded RF antenna is presented. The RF carrier is produced by a three-stage ring oscillator that also acts as sensing element exploiting a linear dependence of oscillation frequency on temperature. The antenna is a square loop fabricated using the top metal layer provided by the 0.35 μm CMOS technology. Oscillation frequency shifts due to bias voltage fluctuations are circumvented by using a mathematical technique. The microchip provides an RF power level of -45 dBm at a distance of 1 m. The average power consumption is about 50 μW.

A calorimetric apparatus for the measurement of the power losses in switching devices, e.g. MOSFETS, is presented. The technique uses an heat-flux sensor made of two, new and custom designed, proportional to absolute temperature (PTAT) CMOS sensors, and allows to achieve accurate results without the use of costly measurement equipment. The calorimetric apparatus allows the measurement of the power dissipated by devices operating in repetitive dynamic conditions, without the problems associated with electrical measurements. Coupled to a control/readout circuit and a thermoelectric module (TEM), it allows the determination of the power dissipated by a semiconductor device, regardless of its switching frequency.

A CMOS sensor fabricated in 0.35μm technology, specifically designed for the monitoring of High Concentration Photo-Voltaic (HCPV) modules, is presented. The microchip was designed to monitor temperature and illumination of each solar cell in a module. Temperature is measured by monitoring the base-emitter voltage of two coupled, diode connected, bipolar transistors, while the illumination sensor is an integrated p-n junction photodiode. A custom communication protocol is implemented in the chip to allow the sharing of a two-wire communication resource among the cells.

In this paper, an alternative approach for the measuring and transmission of vital parameters related to High Concentration PhotoVoltaics (HCPV) modules under illumination has been proposed, allowing recovering information from every individual cell. The valuable cell parameters such as temperature and illumination are measured with a dedicated VLSI integrated circuit (IC) that contains also an integrated 2.4 GHz transmitter with an On-Chip-Antenna (OCA) - made on the top metal layer of the same die - that sends the collected data wirelessly to a base station allowing a complete diagnostic of the module performances over the years. On-Off Keying (OOK) modulation is used to send the sensors data. The IC has been designed and realized in AMS CMOS 0.35µm technology.

An autonomous RFID sensor platform is designed and implemented using off-the-shelf discrete components. It includes a dipole antenna to wirelessly transmit the stored data to a reader according to the EPCglobal Class-1 Gen2 standard. The rectifier operates with received power as low as −10dBm. The microcontroller dissipates below 500nW in sleep mode, allowing an optimal energy harvesting and storage from the RF source. The input impedance is dynamically adapted to respond over a range of input power and loads. The collected power is used to write the data captured by sensors into the user memory of an RFID tag at a user selectable rate. The results demonstrate the feature of the system to perform operations up to a distance of 1.5 m from an RF energy source of 30dbm, thus it is a suitable candidate for use in passively powered sensor networks.

An RF powered UHF-RFID passive sensors platform was realized using discrete components and printed antennas designed to resonate at 868 MHz, used both for energy harvesting and data transmission. The tests demonstrate the possibility for the system to operate autonomously within the reading range of a standard RFID reader, that acts both as the RF power source and the receiver of the data stored in the tag user memory. The microcontroller can be interfaced on the same substrate with a sensor made of polymeric materials, sensible to physical parameters or chemical agents.

An high-resolution CMOS temperature sensor with high linearity in the range from 20°C to 90°C is presented. The circuit uses substrate bipolars as temperature transducers. The temperature sensor consists of two building blocks: a Proportional To Absolute Temperature (PTAT) voltage generator and a differential amplifier. The second block is necessary to obtain reasonable voltage level that can be either read with a simple setup using it as analog input for an ADC. The simulation predictions are confirmed with experimental data resulting from the IC fabricated in a standard 0.35µm CMOS technology.