Abstract :
A simple high-linearity charge-injection technique for charge-coupled devices (CCD\´s) is described, analyzed, and practical results presented. The technique is of the potential equilibration type and utilizes the formation of a potential barrier and charge-storage site under one CCD electrode adjacent to an input diode. This obviates the need for an input clock, as used in previously reported surface potential equilibration charge-injection techniques, by isolating a fixed quantity of "supply charge" from the input diode and the CCD transfer electrodes. The analog signal is applied to the succeeding CCD electrode and depending upon its amplitude some, or all, of the supply charge is transferred to under this electrode, and by subsequent CCD action through the register. A theoretical analysis of the technique predicts that nonlinearity vanishes as the effects of the depletion region capacitance are eliminated. In practice, however, a small nonlinear effect, being mainly attributed to charge "slopping," does occur, causing a reduction in the useful dynamic range of the device for a particular level of linearity. Experiments have indicated that a useful dynamic range of 47.5 dB is attainable with all harmonic components at least 45 dB below the fundamental with a total harmonic distortion of less than 1.2 percent. Variations in threshold voltages between different input stages does not effect the quantity of charge injected (within certain operating limits), but does result in small variations of the maximum useful dynamic range of the devices. The technique is particularly suited to two-phase CCD\´s because it allows linear charge injection, charge transfer, and reset sensing to be accomplished using only one clock.
