Research on the test methods of charge accumulating TDICMOS detector
Liang Nan, Zhang Feiran, Cai Shuai, Li Bo, Li Tao     
Beijing Institute of Space Mechanics and Electricity, Beijing 100094, China

Overview: With the increasing demand for high resolution, high-speed transmission and miniaturization in the field of space remote sensing, the original TDICCD detector has been unable to meet the demand on circuit volume and power consumption. At the same time, with the improvement of CMOS technology, TDICMOS detectors based on charge accumulation came into being in recent years. The detector succeeds in inheriting the charge transfer principle of TDICCD detector in charge transfer mode. Therefore, the TDICMOS detector with low power consumption, high integration and charge accumulation can solve these problems.

As detector is the core component of video circuit, the performance parameters of TDICMOS detector itself have always been an important part of video circuit. Even by reducing the quality of remote sensing satellite image, it can directly affect image interpretation. Therefore, this paper focuses on the test methods of TDICMOS detector parameters. Because AD circuit, sequential circuit and interface circuit are integrated in the detector, the detector is essentially different from the traditional TDICCD and CMOS detector with digital accumulating, no matter in process or in structure. Therefore, many original methods for testing the performance parameters of the detector cannot be applied to the TDICMOS detector based on charge accumulating.

Firstly, this paper compares the TDICMOS architecture with TDICCD. Then, this paper concludes that the test methods of charge-DN factor and full-well charges, charge transfer efficiency and readout noise need to be redesigned. After that, the key parameter of charge-DN factor is effectively extracted by fitting curve about the graph based on image DN value and noise square, and then the full-well charges can be obtained. At the same time, this paper analyzes the noise model of TDICMOS. By the method of controlling the integration time and exposure, the linear relationship curve between the integration time and image noise is drawn, and the readout noise can be obtained which is independent of photon incidence and dark charge. Subsequently, a reverse driving sequence is proposed to realize the reverse movement of charges in the isolation line, so that the charge transfer efficiency can be obtained effectively. This paper also builds a TDICMOS test system for testing the three different detector parameters. The experimental results are also given in this paper and prove the correctness of the new test methods and achievable in engineering. The methods described in this paper provide an important basis for the engineering application of TDICMOS test in the future.

Supported by the National Key Research and Development Program of China (2016YFB0500802)