Search

A definitory feature of MEMS devices is the coupling at microscale between multiple energy domains. Reduced order macro-models (ROM) can be built using either signal flow or energy flow paradigms. While easier conceptually, reducing the device to a dynamical system defined by a set of differential equations, the so-called causal signal flow paradigm is nevertheless less portable, because the macro-models embed as well the excitation, or boundary conditions. The energy flow paradigm, on the other side, is an acausal modeling methodology that reduces the physical system to an equivalent generalized, Spice-like, network, based on across-through (power complementary) variables. The circuit-like representation makes it easier to interface with read-out or actuation electronics systems, for complete microsystems design, simulation and optimization. Besides existing standard behavioral modeling languages like VHDL-AMS and Modelica, MathWorks Simscape language has evolved in the past years to the point where it offers an alternative for the implementation of advanced MEMS macro-models; its seamless interface with Matlab and Simulink offers a strong framework for developing complex hybrid microsystems models, that mix Simulink (signal flow) models with Simscape (energy flow) ones. We have built a generic Simscape library for parameterized MEMS pressure sensors for two distinct categories, piezoresistive and capacitive sensing. The behavior of Simscape macro-models (e.g. sensitivity and frequency response) was validated against models of the same devices extracted in Coventor MEMS+ (a dedicated software for the design and simulation of MEMS devices) and finite element analyses (FEA).

This paper describes the development of a microinstrumentation system in silicon containing all the components of the data acquisition system, such as sensors, signal-conditioning circuits, analog-digital converter, interface circuits, sensor bus interface, and an embedded microcontroller (MCU). The microinstrumentation system is to be fabricated using the Multi-Chip-Module (MCM) technology based on a chip-level infrastructure. A standard silicon platform is the floorplan for individual smart sensor die attachment and an on-chip local sensor bus interface, testing facilities, optional compatible sensors (such as thermal sensors). The microinstrumentation system is controlled by a MCU with several modes of low-power operation (inclusive stand-by mode). As the intended application requires a huge amount of data-processing, a RISC-type MCU architecture is to be used. The MCU communicates with the front-end sensors via a two-line (clock and data lines) intramodule sensor bus (Integrated Smart Sensor bus). The sensor scan rate is adaptive and can be event triggered. This upgraded version of the ISS bus allows: service and interrupt request from the sensors, test and calibration facilities. However, the additional functionality requires a third line. The MCU also controls the power consumption and the thermal budget of all system. This paper also presents three applications for the microinstrumentation system: condition monitoring of machines, an inertial navigation system and a miniature spectrometer., STW - Project DEL55.3733. TUDelft. Junta Nacional de Investigação Científica e Tecnológica - Praxis XXI-BD/5181/95.