|
SMART SENSOR SYSTEMS
APRIL 20-24, 2020
TUDelft, Delft, The Netherlands
|
|
Designing Smart Sensor Systems
Kofi Makinwa, TU Delft, the Netherlands
Smart Sensor Systems are systems in which sensors and dedicated interface electronics are integrated on the same chip, or at least in the same package. The design of such sensors requires a multidisciplinary approach that takes the characteristics and requirements of the whole system into account. The interface electronics needs to be designed in such a way that it does not limit sensor performance. Since sensors are often relatively slow, the necessary precision can often be achieved by the use of dynamic techniques such as chopping, auto-zeroing and dynamic element matching. As an example, the design of a state-of-the-art wind sensor will be described.
|
|
Measurement and Calibration Techniques
Michiel A.P. Pertijs, TU Delft, the Netherlands
This lecture discusses the basics of measurement and calibration. Calibration procedures are essential for establishing the accuracy of a sensor in relation to standards. The lecture discusses how smart sensors differ from conventional sensors in how they are calibrated and how they are used after calibration. Various calibration techniques are introduced, as well as various trimming and correction techniques that can be used to adjust smart sensors after calibration. The lecture also explores the possibility of realizing self-calibrating smart sensors. Various forms of self-calibration are discussed, including the co-integration of additional sensors to compensate for cross-sensitivity, and the co-integration of an actuator to generate a calibration signal locally. Three case studies, of a smart temperature sensor, a smart wind sensor and a self-calibrating Hall sensor, are included to illustrate the various concepts.
|
|
Analog-to-Digital Converters
Marcel Pelgrom, Pelgrom Consult, the Netherlands
The basic principles of analog-to-digital conversion will be discussed as well as some characteristic parameters. Different conversion architectures allow optimum conversion for various signal types. Often the trade-off is between accuracy, bandwidth and power. In this lecture specific attention is given to successive approximation and sigma delta conversion. Some practical examples will illustrate the concepts.
|
|
Dynamic Offset-Cancellation Techniques
Kofi Makinwa, TU Delft, the Netherlands
In modern CMOS processes, device mismatch typically results in offset voltages of several millivolts. But many sensor interfaces require much lower offset levels. By using dynamic offset cancellation techniques such as auto-zeroing and chopping, however, microvolt levels of offset can be routinely achieved. In this lecture, an introduction to the theory of auto-zeroing and chopping will be given, and the pros and cons of both techniques will be discussed. Examples will be given of the use of auto-zeroing and chopping in sensor interfaces with residual offsets as low as 50nV.
|
|
CMOS-Compatible Microfabrication
Reinoud Wolffenbuttel, TU Delft, the Netherlands
A CMOS chip containing both the sensor and electronic circuits is the ultimate level of integration of a Smart Sensor System. However, wafer-level co-fabrication requires full compatibility between the dedicated processing of the sensors and the main process flow for the circuits. Some transduction effects, such as utilization of the temperature dependence of CMOS components, come for free with the mainstream CMOS process. Other effects intrinsic to operation of CMOS components, such as photo-electric effects in diodes, can be exploited using a non-conventional mask design. However, in the general case departures from the mainstream CMOS process are required for including sensor functionality. Any additional sensor-related processing makes the CMOS process non-standard and care should be taken to ensure proper operation of the circuits. Compatibility requirements forces to strategic decisions: on the economic viability of the on-chip integration in the intended application, but also on whether to postpone sensor-related processing until after completion of the CMOS processing or to interrupt the main process flow. Compatibility of process steps also results in limitations, such as: acceptable materials, etchants (for cleanroom re-entrance) and conditions (maximum anneal for circuit operation). The various approaches for CMOS-compatible integration are discussed, with an emphasis on CMOS-compatible micromachining using examples of successfully integrated single-chip microsystems.
|
|
Smart Inertial Sensors
Michael Kraft, KU Leuven, Belgium
Accelerometers and gyroscopes are one of the most successful MEMS sensors. The lecture will briefly present their underlying principles, and then focus on the state-of-the-art as there is still considerable effort going on to increase their performance and functionality. Key to high performance is the inclusion of micromachined sensing element in a force-feedback, closed loop control system. The approach based on electro-mechanical sigma-delta modulator has proven to be very successful in recent years. Such a control system yields a digital output enabling the digital processing of the sensors’ output, hence allowing the design of smart inertial sensors.
|
|
Integrated Hall Magnetic Sensors
Pavel Kejik, EPFL, Lausanne, Switzerland
The lecture starts with a brief introduction into the Hall effect and Hall elements. Then the problems and good practices in the realization of integrated Hall magnetic sensors will be reviewed. The main issues are offset, temperature cross-sensitivity, switching noise, and drift related to the packaging stress. By combining Hall elements with well-adapted interface electronics some of the problems can be dramatically reduced. It will be shown that integration of magnetic flux concentrators on the sensor chip will further decrease the equivalent magnetic noise and offset.
|
|
Smart Temperature Sensors
Kofi Makinwa, TU Delft, the Netherlands
Smart temperature sensors are everywhere! They are used in CPUs for thermal management, in DRAMs to control refresh rates, and in MEMS frequency references for temperature compensation, to name but a few high volume applications. In this lecture, the operating principles of smart temperature sensors will be explained, their main sources of inaccuracy identified, and suitable remedies, at the device, circuit and system levels, described. To further illustrate these concepts, the design of state-of-the-art temperature sensors with inaccuracies in the order of 0.1°C will be presented.
|
|
Implantable Smart Sensors for Advanced Medical Devices
Tim Denison, Medtronic Neuromodulation, USA
The use of physiological sensors is a key enabling technology for implementing ‘smart’ implantable systems. For example, electrocardiograms (ECG) are well established for measuring the intrinsic activity of the heart, and algorithms based on the ECG help to initiate stimulation therapy in the presence of an abnormal beat in modern pacemakers. The role of physiological sensing continues to grow as technology evolves and can be applied to resolving unmet clinical needs. The practical implementation of chronic physiological sensors presents numerous challenges. In particular, sensors that go in the body have strict requirements on reliability, stability and safety. Additional challenges arise with the constraints placed on an implantable design. These constraints include low supply overhead and limited current drain, as chronic sensors must often limit their power dissipation to microwatt levels in order to have acceptable implant longevity. This tutorial will highlight recent physiological smart sensor prototypes that provide robust performance within the constraint of an implantable system. Case studies will include “reflex concepts” implemented with accelerometers, as well as prototype seizure monitors and prosthetic brain-machine interface technologies utilizing precision chopper amplifiers.
|
|
CMOS-based DNA Microarrays
Roland Thewes, TU Berlin, Germany
CMOS-based DNA sensor arrays have gained huge interest in recent years since they provide advantages compared to state-of-the-art commercially available tools for the same purpose using optical principles. In this lecture, an overview is given starting with the general operation principle of DNA microarrays, optical and electronic functionalization techniques, and detection principles using labeling-based and label-free read-out. CMOS integration issues and the related processing requirements are briefly discussed. Emphasis is put on CMOS circuit design requirements in terms of sensitivity, stability, and further parameters depending on the particular application. Examples from the literature are considered to demonstrate opportunities and limitations of CMOS chips applied in that field.
|
|
Guided Simulation on Multi-Domain Modeling
Ger de Graaf, TU Delft, the Netherlands
This guided tutorial is intended to provide each individual participant with hands-on experience in the use of a finite element-based multi-physics simulation tool (COMSOL Multiphysics) that has become indispensable in smart sensor design. The simulation is organised in two inseparable sessions. During the first session the essentials of multi-domain modelling in the coupled thermal-electrical domains are highlighted, first in terms of equivalent lumped components and followed by the representation that is suitable for finite element modelling. Subsequently, the modelling is applied for analysing the thermal behaviour of a free-standing micro-beam with an electrical current applied. The step-wise guided simulation confronts each participant with practical issues, such as meshing, applying proper boundary conditions, use of a material database and post processing of the results. Finally operation as a CMOS-compatible hot-wire anemometer is verified. The second session is dedicated to the design of a bulk-micromachined accelerometer. After the presentation of the mechanical-electrical coupled domains, the equivalent lumped components and the hands-on building of the finite element model, participants are guided through several accelerometer designs aspects, such as bandwidth, higher-order modes of operation and squeeze-film air damping.
|
|
Precision Instrumentation Amplifiers
Johan Huijsing, TU Delft, the Netherlands
To understand the problems of the designers of sensor interface circuits will help the system designer to get the best performance. The principles and features of precision instrumentation amplifiers, key building blocks of many sensor interfaces, will be discussed from a designer’s point of view. Constraints regarding noise, dynamic range, common-mode range will be discussed for circuits made in state-of-the art technology. The case studies include instrumentation amplifiers with offset cancellation, and amplifiers with rail-to-rail voltage ranges.
|
|
References for Smart Sensors
Fabio Sebastiano, TU Delft, The Netherlands
Although often neglected, a reference is always required for any measurements, since measuring basically involves comparing the physical parameter of interest to a known quantity. Consequently, it is a fundamental component in any smart sensor, which can even limit the performance of the whole system if not properly designed. In this lecture, an overview of references for smart sensors will be given with specific focus on references that can be implemented on chip in a standard CMOS process. An overview of references available in CMOS will be presented, including resistance, capacitive, voltage, current and frequency references. We will discuss the need for references in smart sensors and their requirements, and explore through practical examples and case studies how the performance of integrated references can meet those requirements.
|
|
Multi-Electrode Capacitive Sensors
Gerard C.M. Meijer, TU Delft, the Netherlands
A systematic approach towards the design of reliable smart low-cost high-performance capacitive sensors is presented. The basis problems and their solutions of both the physical and the electrical signal processing are discussed. The examples concern capacitive sensors in position detectors, liquid-level detectors and personnel detectors.
|
|
Smart Acoustic Sensors
Michiel A.P. Pertijs, TU Delft, the Netherlands
Acoustic waves can be used to perform a wide variety of measurements, such as flow sensing, ranging and medical imaging. This lecture first introduces the basic operating principles of acoustic sensors and then focuses on the opportunities opened up by combining transducers and integrated electronics to form smart acoustic sensors. This combination is key for the realization of ultrasonic devices that employ transducer arrays with large numbers of elements, e.g. for 3D medical imaging, and for miniaturized, low-power devices. The basic operating principles of piezo-electric and capacitive ultrasound transducers and key interface circuits such as LNAs, pulsers and beamformers will be discussed. A miniature ultrasound probe for 3D medical imaging will presented as a case study.
|
|
CMOS Image Sensor
Albert Theuwissen, Harvest Imaging, the Netherlands
Today, image sensors are present in a wide variety of applications, such as picture taking, video capture, medical imaging, scientific instrumentation and machine vision. Image sensors are used as one of the key input devices for highly automated systems, such as self driving cars or order picking robots. Most image sensors are built in CMOS technology, because it allows to optimize the image sensor for the required specifications and to implement the required functionality in a power- and cost-efficient way. This presentation will give an overview of CMOS image sensors and pixels, readout circuit architectures, manufacturing technologies and key image sensor specifications. New applications are demanding specific requirements to the image sensor, of which some examples will be elaborated.
|
 |
