This paper proposes pixel process techniques to reduce the charge transfer time in high speed CMOS image sensors. These techniques increase the lateral conductivity of the photo-generated carriers in a pinned photodiode （PPD） and the voltage difference between the PPD and the floating diffusion （FD） node by controlling and optimizing the N doping concentration in the PPD and the threshold voltage of the reset transistor, respectively. The techniques shorten the charge transfer time from the PPD diode to the FD node effectively. The proposed process techniques do not need extra masks and do not cause harm to the fill factor. A sub array of 32 x 64 pixels was designed and implemented in the 0.18 #m CIS process with five implantation conditions splitting the N region in the PPD. The simulation and measured results demonstrate that the charge transfer time can be decreased by using the proposed techniques. Comparing the charge transfer time of the pixel with the different implantation conditions of the N region, the charge transfer time of 0.32 μs is achieved and 31% of image lag was reduced by using the proposed process techniques.
This paper presents a 12-bit column-parallel successive approximation register analog-to-digital con- verter （SAR ADC） for high-speed CMOS image sensors. A segmented binary-weighted switched capacitor digital- to-analog converter （CDAC） and a staggered structure MOM unit capacitor is used to reduce the ADC area and to make its layout fit double pixel pitches. An electrical field shielding layout method is proposed to eliminate the parasitic capacitance on the top plate of the unit capacitor. A dynamic power control technique is proposed to reduce the power consumption of a single channel during readout. An off-chip foreground digital calibration is adopted to compensate for the nonlinearity due to the mismatch of unit capacitors among the CDAC. The prototype SAR ADC is fabricated in a 0.18 μm 1P5M CIS process. A single SAR ADC occupies 20 × 2020μm2. Sampling at 833 kS/s, the measured differential nonlinearity, integral nonlinearity and effective number of bits of SAR ADC with calibration are 0.9/-1 LSB, 1/-1.1 LSB and 11.24 bits, respectively; the power consumption is only 0.26 mW under a 1.8-W supply and decreases linearly as the frame rate decreases.
为了降低传统增量型Σ-ΔADC在同精度情况下的量化时钟周期数,提高转换速率,提出了1种采用粗细量化的2步式增量放大型ADC.该ADC采用SAR ADC先进行6位粗量化,再采用增量型Σ-ΔADC进行8位高精度位的细量化,通过数字码拼接完成最终量化结果.同时引入了1种增益自举C类反相器技术,有效地降低了供电电压和整体功耗.该ADC使用0.18μm标准CMOS工艺进行了电路实现,在1.2 V供电电压,1 MHz采样频率、10 k S/s的转换速率的情况下,达到了81.26 d B的信噪失真比(SNDR)和13.21位的有效位数(ENOB),最大积分非线性为0.8 LSB.并且该ADC的整体功耗为197μW,可用于低电压低功耗的仪器测量和传感器等领域.
This paper presents a novel compact memory in the processing element （PE） for single-instruction multiple-data （SIMD） vision chips. The PE memory is constructed with 8×8 register cells, where one latch in the slave stage is shared by eight latches in the master stage. The memory supports simultaneous read and write on the same address in one clock cycle. Its compact area of 14.33 μm^2/bit promises a higher integration level of the processor. A prototype chip with a 64×64 PE array is fabricated in a UMC 0.18 μm CMOS technology. Five types of the PE memory cell structure are designed and compared. The testing results demonstrate that the proposed PE memory architecture well satisfies the requirement of the vision chip in high-speed real-time vision applications, such as 1000 fps edge extraction.
