something about corner∠
时间：10-02
整理：3721RD
点击：
for the NMOS case：
unCox-40°25°80°
slow300uA/V^2200uA/V^2150uA/V^2
typical320uA/V^2220uA/V^2170uA/V^2
fast350uA/V^2240uA/V^2180uA/V^2
for a two stage Miller OTA,if we use PMOS transistor as input pair and NMOS transistor as the output stage,will phase margin be worst @fastPMOSslowNMOS corner?if we design a telescopic opamp, GBW and non-dominant pole are only determined by one kind of MOS transistor，their gm will change almost equally,then，will phase margin be less dependent on corner variation?is it really?
unCox-40°25°80°
slow300uA/V^2200uA/V^2150uA/V^2
typical320uA/V^2220uA/V^2170uA/V^2
fast350uA/V^2240uA/V^2180uA/V^2
for a two stage Miller OTA,if we use PMOS transistor as input pair and NMOS transistor as the output stage,will phase margin be worst @fastPMOSslowNMOS corner?if we design a telescopic opamp, GBW and non-dominant pole are only determined by one kind of MOS transistor，their gm will change almost equally,then，will phase margin be less dependent on corner variation?is it really?
corner不一定这么简单
仔细分析起来cox的变化nmos和pmos的变化方向是一致的
迁移率更接近硅的本征参数
另外考虑corner的时候还要考虑bias的corner变化
其实分析起来是很麻烦的一个东东，而且很多器件原理层面的很多没法分析的东西
可以通过这种方法偷点余量出来，玩玩学术可以
但是量产还是把细点