Body effect in a NMOS transistor

[vistapoint]

For a NMOS transistor,

When VB=VS, there is no body effect. But why η is not = 0 in this case?

So when we ignore body effect, does it imply gmb = 0?

gmb=gm*γ/(2√2ΦF+VSB)

What if VSB<-2ΦF?



And is it true that:

When VB increases, η should increase, and gmb increases.

When VB decreases, η should decrease, and gmb decreases.

 

[srivatsan]

gmb mosfet body effect

Please note that if VSB is zero, then there is no body effect, therefore you cannot find the derivative of the body effect equation. Even if you do it, it should be zero.

Solving the I-V equation of the MOSFET with channel elgnth modulation and/or body effect parameters requires the use of numerical analysis. Equations are provided for quick grasp of their probable working. Thats it; not for precise answer. Solving such equations require matlab or mathcad, but the benefit is not that much. Therefore, assuming gmb=0 means you are finding "near" answer.

regarding -2Φf question, please look at basic semiconductor theory. There are lots of them, book from leblebci is a goodplace to start.

Owing direct relationship between VSB and body factor, that statement is just fine. Never work with VB alone, work with VSB.

Hope it helps.
srivats

 

[amir_se]

body effect in a mos transistor

The derived equation for gm is assumed to be in body effect existing condition, then it is not correct when vsb = 0

 

[konqueror]

body effect desugn equation

there is also a mod with 2Φf+ Vsb i mean it is
√ |2Φf+ Vsb| so it does not matter if Vsb < -2Φf
the same eqn should apply

 

[ezt]

when to ignore body effect

I agree with srivatsan about VSB. you can't do the derivatives when the variable is always zero(here VSB.)
and about Vsb<-2Φf. I think when this condition occurs, the diode junction between bulk and source goes forward and there would be a great leakage current. in simple, nothing goes right then.

Regards,
EZT