A capacitive fingerprint sensor with low-temperature poly-Si TFTs

Abstract

Recent studies of a direct-contact type fingerprint sensor are published [1,2]. Compact, thin-package sensors are needed for remote terminal use. However, as these fingerprint sensors are produced by the standard CMOS process, the chip cost is high. Production costs using low-temperature poly-Si that could produce ICs directly on soda glass substrate would be lower than those using CMOS. A capacitive fingerprint sensor with low-temperature poly-Si TFTs is reported [3]. There charge saved between the finger surface and sensor plate is transferred to external CMOS amplifiers with two TFTs in the sensor pixel. They realize a 200x200 fingerprint sensor with 100mm pitch. However this method has problems. First, the S/N is worse due to the parasitic capacitance of the scanning line. Second, many high-precision amplifiers are needed to amplify the weak output signal from each pixel, so silicon ICs are used. Third, it is not easy to increase the resolution to 500dpi demanded by users, because two TFTs are used in a pixel. The capacitive fingerprint sensor reported here employs a sensor plate and only one TFT for amplifying charge.Figure16.4.1 shows the structure of the sensor cell. The cell is composed of the sensor plate and a pixel transistor for amplification. The sensor plate is connected to the gate of a pixel transistor. The perception capacitance C f and the parasitic capacitance CGS of a pixel transistor are connected to a gate electrode of transistor Tra in series. Perception capacitance Cf acts between the finger surface and the sensor plate. The detection principle is given below. It is assumed that the surface potential of a finger is 0V, and that VOUT, VSROa and VSROb are 0V as initial conditions. When VSROa becomes Va, the gate electric potential VGa rises by capacitance combination and becomes selfconsistently V Ga =C GS ×V a /(C GS +C f ). Because the gate electrical potential V Ga depends on the perception capacitance C f , the finger surface information is converted into gate electrical potential V Ga . Furthermore V Ga is converted into a drain current I Pa determined by the characteristic of Tra. The parasitic capacitance of an output line, C OUT, is charged by IPa, so the output voltage VOUT rises. When the sampling time of TS has passed, the output voltage can be sampled by turning on Trc, that is, VOUT . =.IPa×TS/COUT. The fingerprint information is then obtained by analyzing the voltage waveform of VOUT.Since the pixel transistors are not isolated, there is a fear of IPa leakage to other pixels (for example, to VSROb through transistor Trb). In an actual sensor array, many of those other pixels flow the leak current, so that the total leak current is large and the S/N of the output signal degrades. To solve this problem, advantage is taken of the high threshold voltage characteristics of the low temperature poly-Si TFTs, which is considered a disadvantage in circuit design. In Figure 16.4.1, assume that the gate electrical potential V Gb is 0V. If the electric potential V HRn reaches a value equal...