论文部分内容阅读
基于随机过程中布朗桥运动理论,提出了一种积分非线性误差(INL)和微分非线性误差(DNL)与DAC分段比的数学模型,该模型与蒙特卡罗仿真方法进行了比较验证。结果表明,两者数据结果相差不超过0.5%,但是仿真时间比Monte Carlo仿真减少了20倍。低位采用二进制码编码的数模转换器,INL与DNL随分段比s:(N-s)变化趋势相同。当s:(N-s)<1/2时,INL与DNL较小;当s:(N-s)≥1/2时,INL与DNL随分段比增加略有降低,考虑到芯片面积等因素,选择s=[N/2]为佳。该模型可望用于高速视频解码及其相关领域。
Based on the Brownian motion theory in random process, a mathematical model of integral nonlinearity error (INL) and differential nonlinearity error (DNL) and DAC segmentation ratio is proposed, which is compared with the Monte Carlo simulation method. The results show that the difference between the two data results does not exceed 0.5%, but the simulation time is 20 times less than Monte Carlo simulation. The low order digital-to-analog converter using binary code encoding, INL and DNL with subsection than the s: (N-s) the same trend. When s: (Ns) <1/2, INL and DNL are small; when s: (Ns) ≥1 / 2, INL and DNL decrease slightly with the segment ratio increasing. Considering the chip area and other factors, = [N / 2] is preferable. The model is expected to be used in high-speed video decoding and related fields.