| Description |
Since the late 1950s, scientists have been working toward realizing implantable devices that would directly monitor or even control the human body's internal activities. Sophisticated microsystems are used to improve our understanding of internal biological processes in animals and humans. The diversity of biomedical research dictates that microsystems must be developed and customized specifically for each new application. For advanced long-term experiments, a custom designed system-on-chip (SoC) is usually necessary to meet desired specifications. Custom SoCs, however, are often prohibitively expensive, preventing many new ideas from being explored. In this work, we have identified a set of sensors that are frequently used in biomedical research and developed a single-chip integrated microsystem that offers the most commonly used sensor interfaces, high computational power, and which requires minimum external components to operate. Included peripherals can also drive chemical reactions by setting the appropriate voltages or currents across electrodes. The SoC is highly modular and well suited for prototyping in and ex vivo experimental devices. The system runs from a primary or secondary battery that can be recharged via two inductively coupled coils. The SoC includes a 16-bit microprocessor with 32 kB of on chip SRAM. The digital core consumes 350 μW at 10 MHz and is capable of running at frequencies up to 200 MHz. The integrated microsystem has been fabricated in a 65 nm CMOS technology and the silicon has been fully tested. Integrated peripherals include two sigma-delta analog-to-digital converters, two 10-bit digital-to-analog converters, and a sleep mode timer. The system also includes a wireless ultra-wideband (UWB) transmitter. The fullydigital transmitter implementation occupies 68 x 68 μm2 of silicon area, consumes 0.72 μW static power, and achieves an energy efficiency of 19 pJ/pulse at 200 MHz pulse repetition frequency. An investigation of the suitability of the UWB technology for neural recording systems is also presented. Experimental data capturing the UWB signal transmission through an animal head are presented and a statistical model for large-scale signal fading is developed. |
