Vds至少要比Vdsat大多少才保险呢？

MOSBJT · 发表于 2013-7-10 11:43:38

马上注册，结交更多好友，享用更多功能，让你轻松玩转社区。

您需要登录才可以下载或查看，没有账号？注册

×

本帖最后由 MOSBJT 于 2013-7-10 11:46 编辑

一般来说，Vds至少要比Vdsat大多少才保险呢？求高手指教！！！

yeyepig101 · 发表于 2013-7-10 19:17:18

Vds-Vdsat越大表明Transistor 离linear region越远。一般可以这么理解，一般都是在Gate和source形成一个channel，也就是Vgs-Vth>0，Vgd-Vth<0。这样子的话这个transistor就在saturation region。如果你Vds-Vdsat<0的话，就是Vgs-Vth>0而且Vgs-Vth>0 就是在Gate和Drain也有有一个channel。这样子就在linear region。至于要多大保险。你可以用monte carlo simulation 测试一下。 mismatch和process variation 会改变Vth。如果Vds-Vdsat 太小的话可能会让transistor在linear region。

feynmancgz · 发表于 2013-7-10 20:23:10

回复 2# yeyepig101

难道是ye局？

jamesccp · 发表于 2013-7-11 16:44:29

格雷书介绍大概100mV以上吧.
个人觉得工艺和实际电路不同，可能差异比较大。但有电压空间的话，取大点安全

JoyShockley · 发表于 2013-7-11 18:06:10

If the transitor appears in the small signal model of an amplifier and is required to contribute to the DC gain of the amplifier, the best Vds is about 2 or 3 times greater than Vov. In my study about 2id/gm methodology, I find the intrinsic gain of a transistor only relates to the L and the ratio K of Vds and Vov (or V* mentioned in EE240 from UCB), and only when the ratio reaches to 2 to 3, there is a certain intrinsic gain of a transistor.

Fig. 1 shows the relationship between the intrinsic gain and the ratio K as well as L, which could prove the above arguments. All data are mesured in TSMC 130nm technology.

Note that the expression of DC gain of a amplifier could convert into the expression of the correspondent intrinsic gain of Mosfets, which could be found in EE214 from stanford.