SPICE modeling of ultrasound systems improvements and verifications

Abstract: The success of modern electronics is built on the possibility to accurately predict system behavior by the use of simulation tools. This paradigm can be extended to components such as sensors attached to the electronics. The ability to simulate both sensors (mechanical components) and electronics together renders possible effective optimizations at system level, i.e. minimizing size, cost and power consumption. In this thesis the simulation of a combined electronics and ultrasound sensor system is explored. The environment used is compatible with the electronic simulation tool SPICE. Improvements and verifications of existing SPICE models for ultrasound equipment is described, and applied in the design of integrated analog electronics for an ultrasound measurement system. Emphasis is put on the interdependence between acoustic performance and electronics design. The goal is to improve precision in the simulations to a level where real systems can be implemented from simulation results alone. The thesis is divided into introduction and three attached papers. In the introduction, an overview of ultrasound devices, measurement technology and simulation is given. Tools and design flow for analog integrated circuits are discussed. The first paper shows that system simulations can be used to minimize the size of the transistors used to excite an ultrasound transducer, while keeping maximum output ultrasound energy. A design of an ASIC (Application Specific Integrated Circuit) driver stage for piezoelectric crystals is made and performance of the system is predicted using system simulations. Measurements and simulations are compared, showing that the optimum transistor size can be chosen from simulation data with very good precision. The goal with paper number two is to achieve absolute amplitude correctness in PSpice simulations of ultrasonic systems. Previously published models of the ultrasound propagation medium include viscoelastic loss but disregard loss due to diffraction, i.e. beam spreading. This paper presents a method to include diffraction loss in the models. Measurements and simulations have been performed using a pulse echo system in water. Results show that the simulated amplitude of the returned echo differs less than 10% from measured values in both near and far fields. In paper number three, the influence of parasitic electrical components on measurements and simulations is investigated. It is shown that simulation of excitation pulses can be done with very high accuracy if parasitics are taken into account. The coaxial cable which connects the electronics and the transducer represents one of the major parasitic components in the system. As the cable length is varied, pulse echo amplitudes and time delays shift. It is shown that simulations can be used to predict these effects with good accuracy.