This paper presents a switched capacitor interface circuit for a monolithic three-axis capacitive micro-electro-mechanical system (MEMS) accelerometer. The MEMS sensor and the interface circuit of our system-in-package-type MEMS accelerometer are 3-D stacked to optimize integration density with a small package footprint. The proposed fully integrated interface circuit includes a capacitance-to-voltage converter followed by a Δ Σ analog-to-digital converter (ADC). To optimize system power and area, the programmable-gain functionality is embedded to the second-order Δ Σ ADC without any stability degradation. The offset and 1/f noise of the fully differential interface circuit are mitigated by a correlated double sampling technique. The rest of the low-frequency error from system mismatch is also suppressed by calibration using fine metal-oxide-semiconductor capacitor array. The measurement results of our MEMS accelerometer show a 13.7-b maximum effective resolution with the 197-μg/√Hz noise floor in a conversion time of 1 ms with a maximum nonlinearity of 1.09%. Implemented in a standard 0.18-μm CMOS technology, the fabricated chip consumes only 247-μA current from a 3.3-V supply, and 37-μA current from a 1.8-V supply.

This paper describes a high accurate humidity sensing system. The humidity sensor and the interface circuit of our system-in-package (SiP) are 3-D stacked to optimize integration density with small foot print. A single reconfigurable ADC is utilized for both 3-D stacked humidity and on-chip temperature sensors. An incremental ΔΣ ADC with a reconfigurable ΔΣ modulator is employed with dual input stages. With the combination of chopping, correlated double sampling (CDS), dynamic element matching (DEM), and system-level chopping techniques, high accuracy sensing system is achieved. The interface circuit, implemented in a standard CMOS 0.18 μm CMOS technology, occupies an area of 1.24 mm2. The SiP sensing system dissipates only 146.5 μA and 25 μA from 3.3 V supply and 1.8 V supply, respectively. Its humidity inaccuracy is 0.15 %RH with a resolution of 0.012 %RH, while a temperature inaccuracy is 0.44 °C (3σ) with a resolution of 0.015 °C.

Medical practioners usually measure dorsiflection of the human joint angle for medical assessment purpose. In order to measure the joint angle, a more or less standard mechanical or electro-mechanical goniometer is used. But mechanical construction has many problems. This research is originated from the lack of user-friendly, versatile and not expensive apparatus which are commercially available. In this paper, we present new design concepts of goniometer utilizing the light refraction characteristics of optical fiber in order to measure angular displacement of the human joints. In order to form an asymmetrical beam profile at the end of fiber tips, we cut the end of fiber tip at an angle of 35 degrees. Phototransistor is used to measure beam intensity as a function of optical fiber angle with respect to optical axis of fiber. Thus the sensed signal represented an indirect measure of curvature angle. The result showed that asymmetrical beam profile provide wider linear relationship than symmetrical one. Eventually, overall performance of the proposed sensor is quite suitable for measurement of the human joint angle

High-resolution read-out integrated circuits (ROICs) are often used in DC measurement systems such as bridge transducers. Since these typically output small-amplitude signals with a bandwidth of a few hertz, a highly linear gain ROIC is required with a resolution above 20b [1,2]. In previous work, high-precision instrumentation amplifiers (IAs) followed by analog-to-digital converters (ADCs) were reported [1-4]. However, IA topologies based on switched-capacitor (SC) or multiple operational amplifiers (opamps) suffer from a poor power-noise tradeoff due to noise folding [2], or the number of amplifiers [3], respectively. Current-feedback IAs (CFIAs) are more power-efficient, but require highly linear feedback resistors [4,5]. This drawback can be addressed by capacitively-coupled IAs (CCIAs), which are even more power efficient [1].