利用sentaurus device不收敛

利用sentaurus device进行器件大信号仿真时要么直接不收敛，要么经过几个周期后不收敛，不知道你们又没有过类似的仿真?这种情况要怎么破啊?

Its not a problem with sentaurus but its iterative solver use less norm of error in newton iterative solver. See how the matrix is formed. Don't get pissed off so soon, there is always good things to learn by solving big problem.

当仿真不收敛的时候，可能需要重新调整网格的分布。这点要求挺好的，需要对体系内部结构的很多物理特性进行预判。最后，不收敛挺正常，device仿真有时就是要寻找收敛的情况。