Switched-capacitor track-and-hold amplifier with low sensitivity to op-amp imperfections

Oversampling Successive Approximation Technique for MEMS Differential Capacitive Sensor

“…where C0 is the value of the common mode capacitance of differential capacitive sensor, Ci is the value of integration capacitors, ΔC is the change value of differential capacitive sensor, VR is the reference voltage acting as readout excitation, and σd is the deterioration factor. An important conclusion can be drawn from the equation (14), that is, the limit of the OSSA output is the same as the expected ideal output, regardless of the gain A0 and the deterioration factor σd , i.e., (∞) = →∞ ( + ) = 2∆ (15) This means that, by introducing the OSSA technique, no matter how low the gain of amplifier is, and no matter how large the deterioration factor σd is, given enough steps, the output will always reach the near ideal level.…”

“…Although the main structure of the OSSA amplifier is based on the single-end T/H amplifier in the reference [14], the major improvement over the accuracy has been achieved through the following 3 aspects of the work: developing the fully differential structure, proposing the nested non-overlapping clocks (alone with the generator) and parameter optimization based on the analysis of non-idealities. This improvement aims specifically at the application of the low modulation depth MEMS sensor, which has not been considered in the published references on T/H SC amplifier.…”

“…That means this switched-capacitor readout circuit for differential capacitive sensor in Fig. 1 suffers much severer 1/A error than other forms of switched-capacitor amplifier [14], [15].…”

“…However, a particular phenomenon "1/A error deterioration" will happen in traditional fully differential switchedcapacitor readout circuit with CDS due to the parasitic capacitance introduced by the MEMS manufacturing and packaging process (refers to die-to-die bonding), which affect the accuracy of output severely. Although the predicating techniques and the track-and-hold (or sample-and-hold) techniques had been proposed to solve the 1/A error [14], [15], it is hard to apply them to the readout circuit for MEMS capacitive sensor for "1/A error deterioration" solution. This is because these techniques require well matching of the values between the capacitors in predictive/tracking path and the capacitors in measuring/holding path, which is impractical in the situation where those capacitors are manufactured in different processes (MEMS and CMOS).…”

Bandwidth-Enhanced Oversampling Successive Approximation Readout Technique for Low-Noise Power-Efficient MEMS Capacitive Accelerometer

“…The second method is use of negative impedance (negative-R [19] or negative-C [20]) to compensate the signal charge loss. The third method is to calibrate/correct the gain error with the assistance of the digital circuits [21] [22] or the switchedcapacitor circuits, such as the oversampling successive approximation (OSA) readout technique [9], correlated double sampling technique [23] [24] and correlated level shifting technique [25]- [27], etc. This method is able to provide scaling-friendly solution in modern nanometer CMOS process and it is therefore very popular [28].…”