PMOS逻辑
计算机科学
物理
量子力学
电压
晶体管
作者
Bo-An Tsai,Wei-Cheng Chu,Yu‐Chih Chang,Yiheng Chen,Chien-Fu Chen
标识
DOI:10.1109/ipfa53173.2021.9617283
摘要
It is well known that when PMOS is stressed at I bmax or I gmax , its I dsat will first increase and then decrease, known as the turn-around effect [1],[7]. In this paper, we will first explain the differences in I dsat changes in the three stress modes of PMOS (I bmax , I gmax , V g =V d ), then compare the difference between I dsat increase percentages under I bmax and I gmax stress. Secondly, in regards to the significant increase of I dsat percentage under I gmax stress, the results of this paper is seldom mentioned in other papers, such as the appearance of I gmax peak in relation to V d , and I dlin always having an earlier turn-around compared to I dsat . Finally, we discuss the occurring time and saturation percentage of I d 's turn-around at I gmax stress with different geometric factors, especially in its correlation with the narrow width effect [5]. The results show, i) the holes tunneling into the oxide near the source side by ΔV gs and the hot electrons (generated by HC impact) injecting into the oxide near the drain side by ΔV gd compete with each other, bringing about the different I dsat changes of the three stress modes; ii) the horizontal and vertical electric fields affect the impact point of HCI and the peak point (the highest probability) of the hot electrons injecting into the oxide (we abbreviated as PPHEIIO), effecting the larger increase of I dsat percentage under I gmax stress compared to I bmax stress. This brief explanation is seldom mentioned in other papers; lastly, the energy carried by the holes and the channel lattice scattering affect the energy of the hot electrons, which affected the turn-around effect for PMOS stressed at I gmax . At the same time, besides "interplay between interface trap generation and trapping of carriers by oxide traps" mentioned by other papers being the only explanation for earlier turn-around of I dlin as to I dsat , the increase in R on seems to be an important influencing factor.
科研通智能强力驱动
Strongly Powered by AbleSci AI