D.C. van Duyn

Integrated thermopile sensors

Abstract This paper is about integrated silicon thermopiles and their applications in silicon sensors. After a short description of the thermoelectric effect and its use in silicon thermopiles, some attention is devoted to the design of micromachined structures for implementing thermal sensors. The various sensing principles based on thermal effects are discussed next. Finally, an impression is given of some of the recently developed silicon-thermopile sensors which implement these sensing principles.

Small-size vacuum sensors based on silicon thermopiles

Abstract This paper describes the development of integrated silicon thermal vacuum sensors with much reduced sizes, compared with previously described versions. After a short expose of the physics of thermal vacuum measurement, the general sensor structure is elucidated. Then the novel structure is described, seven times smaller but three times more sensitive than previous versions. The experimental results show a sensitivity of 12.4%/Pa for nitrogen and a pressure range of 1 mPa to 10 kPa absolute pressure. The sensor can be mounted in a small TO-5 housing.

A novel accelerometer based on a silicon thermopile

An accelerometer based on a silicon thermopile in a cantilever beam structure is presented. The principle is that the motion of a cantilever beam, induced by the accelerations of the chip, results in temperature variations. The mathematical model for both the thermal sensitivity and the thermal time response reasonably conformed to the experimental results. It is deduced that the sensitivity can be increased by choosing a smaller gap; the sensitivity can be 1.3% per unit g, which compares with capacitive accelerometers; and the thermal time constant is in the range of 10 to 18 ms, according the theoretical model, which is consistent with the experiments.<<ETX>>

Design considerations for the thermal accelerometer

Abstract This paper presents some design and optimization aspects of the thermal accelerometer. The design and optimization strategy is based on a one-dimensional model describing both the thermal and mechanical properties of the device. It is shown that this model is reasonably in agreement with experiments carried out on some prototype devices. The performance optimization criteria are based on the ‘maximally flat amplitude response’ from which the optimum design parameters can be calculated for a given bandwidth.

Information transduction in solid-state transducers: A general thermodynamic systems approach

Abstract In this paper we review the concept of information transduction, especially in the case of solid-state transducers. We conclude that in general, information transduction must be viewed as a coupling between the states of two systems. The physical concepts of state and coupling are discussed. We also briefly touch upon philosophical and semantic aspects of the concept of information. The above point of view is applied to a solid-state transducer system using general thermodynamic principles and applying the above point of view locally. Based on thermodynamics, we arrive at a unified set of thermodynamic state variables suitable for characterizing the process of information transduction in solid-state transducers. Moreover, the result of this approach strongly supports the ‘sensor-effect cube’ as proposed by S. Middelhoek and A. C. Hoogerwerf (Sensors and Actuators, 10 (1986) 1–8).

Finite-element modelling of thermoelectric materials and devices

Abstract This paper deals with the modelling of materials and devices in which electric and thermal effects and their cross effects (thermoelectric effects) are important. A phenomenological model based on the thermodynamics of irreversible processes (TIP) and its mixed-hybrid finite-element implementation are presented. It is argued that TIP provides a unified approach to sensor modelling. Mixed-hybrid finite elements are used to resolve the thermoelectric interaction terms accurately. A numerical test example is given to demonstrate the feasibility of this approach.

Modeling and simulation of solid-state transducers: the thermal and electrical energy domain

Abstract This paper deals with a basic strategy towards the (efficient) numerical modeling of semiconductor micro-structures in which electric, thermal and thermoelectric effects are important. In particular, we discuss a very efficient method of O(n) for solving the nonlinear semiconductor thermoelectric model equations, where n is the total number of elements in the grid. The method is based on the multigrid method in combination with the mixed finite element discretization method and a nonlinear Vanka-type relaxation method. For a simple, but representative test problem we demonstrate the feasibility of this approach by showing that (nearly) mesh-size independent convergence rates can be obtained.

Finite element modeling of sensor: thermoelectric effects

The author considers the modeling of materials and devices in which thermal, thermoelectric, and electric effects are important. A basic model and its implementation are discussed. The model consists of coupled sets of elliptic partial differential equations describing the relations between the intensive and extensive thermal and electric state variables. For the implementation a finite element method is used employing mixed-hybrid finite elements in order to accurately resolve the interaction terms between the electrical and thermal domains. To enhance robustness a continuation strategy has proven to be necessary.<<ETX>>The author considers the modeling of materials and devices in which thermal, thermoelectric, and electric effects are important. A basic model and its implementation are discussed. The model consists of coupled sets of elliptic partial differential equations describing the relations between the intensive and extensive thermal and electric state variables. For the implementation a finite element method is used employing mixed-hybrid finite elements in order to accurately resolve the interaction terms between the electrical and thermal domains. To enhance robustness a continuation strategy has proven to be necessary. >