Corona discharges occur in high voltage electrical equipment in case of defects and damage, while ultraviolet(UV) light generated during discharge. High resolution imaging in the solar-blind UV bands has a lot of applications in corona discharge detection. A ultraviolet imager based on 320×256 solar-blind AlGaN focal plane arrays (FPA) was designed that work even in the sunlight, because the Cut-off wavelength of the AlGaN FPA is 280nm. The UV image signal processing system based on FPGA is composed of various function modules include the voltage bias, sequence drive, A/D data acquisition, non-uniformity correction, video transformation. Due to FPGA-based data acquisition and realtime image processing technology, the UV imager can operate at a rate up to 100 frame/s. The results show that the simulation high voltage ultraviolet image can be obtained by the UV imager. The image non-uniformity correction performed is one-points correction method to realize background subtraction. And the images show good uniformity and contrast. The UV image of the alcohol burner flame can be detected by the Ultraviolet Imager. Imaging quality was discussed which can be determined by signal-to-noise ratio (SNR), the integration time, the optics f/number and so on. The best imaging conditions were analyzed and the imaging system was designed and setup. The conclusion is proved that the ultraviolet imager based on solar-blind AlGaN FPA provides a new method for corona discharge detection of high voltage power transmission and distribution system.
A Readout Integrated Circuit (ROIC) for GaN ultraviolet (UV) focal plane array (FPA) working in “solar-blind” band is studied in this paper. It has a format of 320×256 and a pixel pitch of 30μm. This circuit can operate both in integrating-while-reading (IWR) and integrating-then-reading (ITR) mode with the frame rate higher than 100fps. It is common that trade-offs always exist between chip power consumption and performances in integrated circuits design. In order to get high injection efficiency with small area and low power, A novel low-power capacitive-feedback trans-impedance amplifier (CTIA) with snapshot mode is designed for the proposed circuit. The smallest operational current of CTIA is only 10nA for 5V power supply. The total power consumption of ROIC is reduced significantly to 45mW with the ultra-low-power pixel. By adopting the 0.35μm 2P4M mixed signal process, the high-performance CTIA architecture can make two gain selections which charge capacities are 3.4Me - and 0.16Me - per pixel with 2.5 V output range. According to the experimental results, this circuit works well under 5V power supply and achieves 8MHz pixel-data-transmission rate.
A novel column-stage structure of readout integrated circuit (ROIC) for GaN ultraviolet (UV) focal plane array (FPA) working in “solar-blind” band is proposed. The column stage has better drive capability, higher dynamic range, stable bias current and low impedance. The noise voltage of the column readout stage is discussed, which has been reduced by small-current driving, column-stage sample and hold and the technology of divided-output-bus. This research on low-noise ROIC is designed for weak-current UV FPA. It is designed, simulated and laid out using the 0.35um 2P4M CMOS 5V process. The clock rate operates at 8MHz. The simulation input current sets 0.01nA. The output swing is 2.6V and power consumption is 40 mW according to the measurement results.