Mixed-Signal Platforms for MEMS and MOEMS Sensor Conditioning

Thanks to advances in integrated technology complexity of System on a Chip has increased very rapidly in recent years while requirements of short time-to-market, high performances and reliability have been increasing. Moreover new applications based on MEMS and MOEMS devices, are currently spreading in a great variety of applications from automotive to optical fields. The interest in these devices is due to low power consumption, low manufacturing cost (that stems directly from batch fabrication and the possibility to exploit the infrastructures already developed for integrated circuit fabrication) and small dimensions, which allow their integration with read out electronics on the same chip or die. However MEMS and MOEMS devices require more complex read out interfaces. In order to face all these requirements and challenges, companies have reviewed typical design flows, developing new different approaches such as the Universal Sensor Interface and the Platform Based Design. However these two approaches suffer from some drawbacks. The former approach offers high flexibility at the expense of overall performances, the latter requires time-consuming architectural space exploration. In this dissertation the ISIF platform, developed by the University of Pisa in collaboration with SensorDynamics AG, is presented as an effective solution to the traditional Platform Based Design drawbacks. Two case studies are presented to show the effectiveness of the proposed design flow: the development of a water flow monitoring system with a focus on the design of the high-voltage driver that has enabled the ISIF platform to actuate the MEMS hot-wire anemometer, and the design of a new platform, named SD4K, able to interface the latest generation of MEMS and MOEMS devices. The SD4K is targeted to the development of a projection system based on a scanning micromirror and a laser source system. This research activity has required developing a high level model of the micromirror for the proper design of the read out electronics. Once the model is realized, it is possible to perform high-level system simulations that take into account the effect of micromirror non-linearities (such as the non-linear frequency response) and thus perform a more accurate and careful design of the sensing and driving stages and of the compensation algorithms implemented by software routines. Finally the dissertation presents the design of the high-voltage driver for the actuation of the scanning micromirror chosen for the laser-based projection display application.