Abstract
In the formal evaluation tool for sleep-disordered breathing, patients need to go to sleep lab in hospital to do the overnight test. However, the multiple test channels attached to the patient are uneasy. In addition, the whole process needs a professional technician to operate and mark the events after the test. All the procedure is labor-intensive. In this research, we try to develop an unattended healthcare system to monitor the sleep respiratory disorders at home. The sensor channel must be minimized and the setup has to be simple. By using wireless transmission and signal processing in the back end of the system, the recorded data can be analyzed by computer. This proposal is expected to provide daily respiratory information to assist the medical system for evaluation. One of the front-end sensors in this prototype platform is accelerometer, which is used to detect the thorax and abdomen respiratory efforts. In this study, a CMOS-MEMS capacitive accelerometer and interface circuits with a resolution applicable to the platform are presented. Chopper stabilization is used to avoid the low frequency noise associated with the active devices. The technique also converts the capacitance change of the capacitive accelerometer to a voltage change at output. Since CMOS-MEMS has the advantage of integrability with circuits, the whole system hardware, including accelerometer, interface circuits, A/D converter, wireless transmission circuits and digital circuits are integrated into a SoC. The single proof-mass dual-axis accelerometer is implemented in TSMC 0.18um 1P6M CMOS process with in-house post processing. The size of MEMS structure is 625 μm × 633 μm. The power consumption is 976 μW while operates at 1.8 V power supply. The measured sensitivities of X-axis and Y-axis are 31.7 μV/G and 0.2 mV/G, respectively. Noise level of X-axis and Y-axis are 3.16 mG/√Hz and 1 mG/√Hz.