THE RESISTIVE-GATE charge transfer device area image sensor' is characterized by a specific method of signal charge transport from the photosensitive matrix to the video output. In the image sensor to be described, resistive gate' controlled CTD channels are used for column transport. Figure 1 shows a lay-out of a 4x4-element configuration. Vertical bulk channels, indicated by dashed lines, are covered by poly-Si resistive-gate electrodes which are interconnected at the top (RG-) and bottom ( R G t ) of the matrix. An appropriate voltage across these electrodes will cause an electric drift field along the depleted channels, so that electrons injected into the channels will be transported downwards.Integrator electrodes, whjch traverse the matrix in the horizontal direction, are positively biased so that potential wells are present between the channels. The regions between the integrator electrodes arc used as sensing elements and charge carriers, generated by incident light, are collected in the adjacent potential wells. During readout, the integrator electrodes are sequentially selected by a vertical scan generator. The charge packets under a selected line are transferred simultaneously into the corresponding resistive gate channels and through them to buffer regions, controlled by a common electrode B; transfer gates TG1 and TG2 control the charge transfer to and from the buffer regions. After recollection, the charge packets are then transferred into the PCCD3 output register which serially shifts out the video information. In practice, the maximum time taken to transport signal charge to the buffer will be less than half the video line time interval, which permits effective point anti-blooming. After each line readout, excess carriers from any point of the matrix are transported through the same channels and extracted by anti-blooming drains at the lower end of each column. Furthermore, the sensing area of a picture element is not covered by an electrode, resulting in good sensitivity to blue light. cuit, incorporating the column structure, illustrate these two features. Figure 2 shows the charge transport time Ttr as a function of the voltage VRG across the resistive gate and for a transport length of 2700pm. The data are in good agreement Experimental results on a linear resistive gate sensor test cir-Charge Transfer Principle", Jap. Journal of Applied Physics, 'van Santen. Vol. 20, p. 177-181; 4Hoffman. K., "Surface Charge Transport with an MOS 1 9 7 7 .with theory4. Figure 3 shows a typical spectral response characteristic, indicating a quantum yield o f about 30% for h = 400nm. The absolute rwponsivity figures are related to thc photo-sensitive part of thr picture clement arca. element resistive gate arca sensor has been designed and fabricated. Figure 4 shows a microphotograph o f the circuit which contains 200 columns at 28pm ccnters, resulting in a horizontal matrix dimension of 5.6mm. Each column, having a length of 4.2mm, is provided with an anti-blooming structurc and contains 300 separate sensing/...
