Mario Auer

Switched capacitor DC-DC converter in 65nm CMOS technology with a peak efficiency of 97%

This paper presents a highly efficient switched capacitor (SC) DC-DC converter, which is implemented in a 65nm low power CMOS process to enable SoC integration. The converter supplies a nominal output current of 30mA with an efficiency of 95%. The buck converter operates with a constant input voltage of 1.83V and produces an output voltage of 1.2V. The charge is transported in two phases from the input supply to the output capacitor. Therefore, a network consisting of switches and two so-called flying capacitors are used. This converter is intended to be used in power management units for mobile devices and to substitute complex and expensive inductor based DC-DC converters. A test chip was designed in collaboration with Infineon Technologies Austria AG. The measurement results are presented and give for an output current of 10mA a peak efficiency of 97%.

Using thermoelectric energy harvesting to power a self-sustaining temperature sensor in body area networks

In this paper a thermoelectric energy harvester with an active area of 0.5×0.2mm2 fabricated in a 130nm process is presented. By using automatic frequency adaptation the circuit is able to operate even in the sub-threshold region at about 230mV. The harvester powers a self-sustaining active wireless sensor node for temperature measurement in body area networks. Therefore it extracts power from a difference in temperature of about 1K using 2 thermo electric generators. This sensor measures the temperature and transmits data at a rate of 250kBit/s in the 868MHz ISM-band using ASK-modulation to the receiver side. The sensor is operated in a duty-cycle mode.

Design and development of a mixed signal prototyping system to achieve very high data rates for contactless applications

SummaryContactless applications like ePassport (electronic passport) demand increased communication speed for higher volumes of exchanged data. Today, those passive applications are working at the carrier frequency of 13.56 MHz, as standardized in ISO/IEC 14443, including data rates up to 848 kbps. However, higher data rates are currently a field of research. For the investigation of very high data rates a mixed signal prototyping system is introduced, based on a new concept of a FPGA-based reader prototyping system. Moreover, an introduction to advanced demodulation methods for contactless transponder devices is given.ZusammenfassungDieser Artikel beschäftigt sich mit kontaktlosen Anwendungen, dazu zählt unter anderem der elektronische Reisepass. Dieser speichert zu den herkömmlichen personenbezogenen Daten auch biometrische Merkmale, welche über eine kontaktlose Schnittstelle zum Lesegerät übertragen werden. Um höhere Datentransfers in kürzeren Übertragungszeiten zu ermöglichen, gibt es Bestrebungen, die Datenrate für derartige passive Kontaktlos-Anwendungen zu erhöhen. Diese arbeiten mit einer Trägerfrequenz von 13,56 MHz und sind in der Norm ISO/IEC 14443 standardisiert. Dieser Standard sieht derzeit Datenraten bis max. 848 kbps vor, wobei bereits Forschungstätigkeiten hinsichtlich hoher Datenraten laufen. Zur Erforschung und Analyse von hohen Datenraten in Kontaktlos-Anwendungen wird ein Mixed-Signal-System verwendet. Dieses beruht auf einem neuartigen Reader-Konzept, welches durch ein FPGA-basierendes Prototypen-System realisiert wurde. Darüber hinaus werden drei Konzepte zur Demodulation auf dem kontaktlosen Transponder vorgestellt.

Characterization of EMI-reducing spread-spectrum techniques for class-D audio amplifiers

Audio amplifiers for portable devices need to be power efficient, therefore Class-D amplifiers are used. These amplifiers employ a switching output stage, so that they generally produce a significant amount of EMI. Filtering these frequency components usually is not an option since the amplifiers have to be small and lightweight. A reduction of EMI can be achieved by using spread-spectrum techniques that distribute the emissions over a wider frequency range. Unfortunately spread-spectrum techniques can impact the audio performance. This work presents a test setup for a Class-D prototype to evaluate different spread-spectrum methods by measurement or even by listening to the audio output. Using this setup several spread-spectrum methods are compared in terms of EMI-reduction and audio performance.

Spread spectrum techniques for Class-D audio amplifiers to reduce EMI

For battery powered portable devices it is required to have circuits that combine high efficiency with a small footprint. In the case of Class-D amplifiers the high efficiency can be achieved by applying pulse-width modulation (PWM) to generate the desired output signal. Classical PWM requires an output filter to suppress the idle current. In order to avoid this filter a different modulation scheme is used, which however introduces a larger amount of electromagnetic emissions. Therefore spread-spectrum techniques are applied to distribute the emissions over a wider frequency range. In this article methods are discussed that allow the use of spread-spectrum modulation without introducing audible artifacts due to the modulation.ZusammenfassungFür batterieversorgte mobile Geräte werden Komponenten benötigt, die einerseits eine kleine Baugröße andererseits auch einen geringen Stromverbrauch aufweisen. Im Fall von Klasse-D-Verstärkern wird die hohe Effizienz dadurch erreicht, dass man Pulsweitenmodulation (PWM) verwendet, um das gewünschte Ausgangssignal zu erhalten. Klassische PWM benötigt ein Filter am Ausgang, um Ruheströme in der Last zu vermeiden. Zur Vermeidung dieses Filters setzt man andere Modulationsverfahren ein, welche ein größeres Maß an elektromagnetischen Emissionen erzeugen. Aus diesem Grund verwendet man Spread-Spectrum-Modulation, damit diese Emissionen über einen größeren Frequenzbereich verteilt werden. In diesem Beitrag werden Spread-Spectrum-Methoden vorgestellt, die auch für Audio-Anwendungen geeignet sind.

A low-voltage charge pump starting at 135mV for thermal energy harvesting in Body Area Networks

The design of a fully integrated low-voltage charge pump for thermal electric energy harvesting is presented in this paper. The proposed circuit has a startup voltage of 135mV without the need of expensive low voltage process options. This behavior is achieved by automatic frequency adaptation in correlation to the resistance of the switches. After startup the input voltage may even drop below 100mV, therefore it is well suited for small temperature gradients appearing in Body Area Networks. A test chip was fabricated in a low cost 130nm CMOS technology with an active area of 0.8 × 0.21mm2.

A Class-D output bridge with dynamic dead-time, small delay and reduced EMI

Power efficient Class-D audio amplifiers are very popular in mobile applications, despite the potential electromagnetic interference (EMI) caused by their output stage. Going to higher power levels, more effort has to be put into the reduction of EMI without sacrificing audio performance and efficiency. This work presents a Class-D output stage with an adaptive driver concept. The proposed driver monitors each output transition and dynamically adjusts dead-time and driver strength to reduce EMI while maintaining high audio fidelity. The presented concept is implemented in a 180nm CMOS technology and verified by measurement.

A process-variation compensation scheme to operate CMOS digital logic cells in deep sub-threshold region at 80mV

In this paper a simple technique to use standard digital CMOS logic cells from 80mV to 1.2V is presented. By applying a compensation network process-related variations of the logics switching threshold can be reduced by 89% for a given supply voltage. Therefore post-fabrication process steps can be avoided. The principle is introduced by a simple inverter gate and expanded to more complex NAND, NOR and Flip-Flop cells. A test chip has been fabricated in a 130nm process proving the functionality of the proposed digital cells.

Verhaltensmodellierung von Delta-Sigma-ADCs unter Verwendung von Verilog-A

ZusammenfassungIm Zuge der vorgestellten Arbeit werden universell einsetzbare Modelle entworfen, die sowohl die Simulation von zeitdiskreten als auch zeitkontinuierlichen Delta-Sigma-Umsetzern erlauben. Die Abstrahierung auf Ebene der analogen Verhaltensmodellierung ermöglicht es, diese Modelle als Bestandteil des Designs zu verwenden und schrittweise gegen Implementierungen auf Transistorebene auszutauschen. Die Flexibilität dieser Lösung gestattet es, mit geringem Aufwand Effekte nachzubilden, deren Modellierung auf einer höheren Ebene einen großen Aufwand sowohl bei der Modellbildung als auch bei der Verifikation bedeuten würde.SummaryThis article presents generic models for the simulation of discrete time and continuous time delta-sigma digital to analog converters. Abstraction at the behavioral level offers the possibility to use the developed models along with circuit level schematics and replace them by transistor circuits during the course of the design. This approach has the big advantage that higher order effects can be modeled with moderate to low effort which otherwise would result in a laborious modeling and verification task.

Design of a MEMS speaker driver

Microelectromechanical systems (MEMS) have already found widespread use in highly integrated audio applications. Currently, they are used for microphones where their small size has enabled the use of multiple microphones in a single device to improve noise cancellation. Nowaday, even loudspeakers realized as MEMS structures are possible, which brings up new challenges in circuit design that are the focus of this paper.Starting with an overview of current research activities of MEMS speakers, their advantages and disadvantages, as well as the peculiarities of the driver design are discussed. Compared to electrodynamic speakers, MEMS speakers have a different physical structure which needs to be accounted for when designing the driving circuitry. Possible driver concepts are presented and compared with special consideration of their efficiency and audio quality. Existing concepts are shown and new concepts are proposed.ZusammenfassungMikroelektromechanische Systeme (MEMS) haben bereits eine weite Verbreitung in hoch integrierten Audioanwendungen gefunden. Aktuell werden sie für Mikrofone verwendet, wo durch ihre geringe Bauform die Möglichkeit geschaffen wurde, mehrere Mikrofone in einem einzelnen Gerät zu verbauen und dadurch die Geräuschunterdrückung zu verbessern. Heutzutage sind jedoch auch schon Lautsprecher als MEMS-Strukturen realisierbar. Dies hat neue Herausforderungen im Bereich der Schaltungstechnik zur Folge, welche den Kern dieses Artikels bilden.Beginnend mit einem Überblick über aktuelle Forschungsaktivitäten von MEMS-Lautsprechern werden deren Vor- und Nachteile sowie im Speziellen deren Eigenschaften bezüglich des benötigten Treibers diskutiert. Verglichen mit elektrodynamischen Lautsprechern haben MEMS-Lautsprecher eine unterschiedliche physikalische Struktur, auf welche beim Schaltungsdesign geachtet werden muss. In dieser Arbeit werden mögliche Treiberkonzepte präsentiert und speziell in Bezug auf deren Effizienz und Audioqualität verglichen. Bereits vorhanden Konzepte werden erläutert sowie ein neues Konzept vorgestellt.