| Summary: | With the fast-growing pace of the aging population, there is increasing need to monitor the elderly health status by continuously tracking their vital signs in daily life. Conventional vital signs monitoring devices are usually bulky, expensive and power-hungry, and thus, the usage is limited to within the hospital and is not suitable for daily use in the home environment.
A miniaturized wearable biomedical signal monitoring system is considered as a game changing technology to enable remote monitoring and diagnosis for personalized home healthcare. However, technology features such as the compatibility of readout circuit with different types of biomedical sensors, low power consumption for prolonged continuous usage, high signal resolution and robust performance against environment changes are all key design challenges that remain to be solved.
In this thesis, time-domain resistive/capacitive sensor readout circuit is explored as an alternative solution to the conventional voltage/current sensor readout circuits. The research can be categorized into two parts. Firstly, RC relaxation oscillators, which are employed as frontends of sensor readout circuit, are presented. A novel energy efficient swing boosting technique is proposed in the first oscillator, while a novel voltage-to-delay feedback with switch capacitor swing boosting technique is proposed in the second oscillator. Experimental measurements had demonstrated their effectiveness in suppressing output phase noise while for the second technique, in achieving high output accuracy too. Secondly, a VCO-based Delta-Sigma Modulator (DSM) readout circuit is proposed for bio-signal recording. A VCO-based frontend and system feedback is proposed to replace the conventional amplifier-based frontend with open-loop structure. It can operate with large input range of up to 100-mVpp, as well as with wide signal bandwidth, high energy efficiency and high output resolution.
Despite the work done, the challenges lying ahead are still exciting. These challenges can be viewed in two parts, the integration of proposed frontends into systems, and the creation of even better readout circuits. Plans to deal with them are constantly being generated and more contributions to the subject matter will certainly be made in the near future.
|
|---|