body effect decreases output resistance of source followers?

Status
Not open for further replies.
S

surreyian

Member level 3
Member level 3
Joined
Feb 10, 2006
Messages
63
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,796
Hello all,

In razavi's design of analog CMOS integrated circuit, chapter 3, it states/shows that body effect decrease the output resistance.

But I find it puzzling. I bias the NMOS source with a certain voltage, eg 0.3V.
Setup 1: Base and Source are connected together to 0.3V bias. (No body effect)
Setup 2. Source is connected to 0.3V whilst base is connect to ground. (with Body effect)

Id=(Vgd-Vth)^2K.
Based on the equation, setup 2 Vth will be greater than setup 1, hence reducing Id, hence increase output resistance. Am I right in the analysis?
If that is the case how does body effect reduces output impedance?

Appreciate all help in helping me understanding the fundamental.
 

surreyian said:
In razavi's design of analog CMOS integrated circuit, chapter 3, it states/shows that body effect decrease the output resistance.
Click to expand...

No. It states "Interestingly, the impedance seen at the source of M1 is lower when body effect is included." This statement refers to Fig. 3.8 (a), i.e. to the input resistance of the source follower load transistor (which is named M2 in Fig. 3.9).
 

If you simply draw the small signal model and add the gmb part which represents the body effect, you will find the output resistance is Ro=ro||1/gm||1/gmb. Absolutely, it's been reduced.

Intuitively, body effect can be regarded as the fact that the channel current is controlled by not only the GATE but also the back gate(BODY or BULK) of the device, i.e., another current branch is introduced in the model of calculating output resistance of the COMMON DRAIN STRCUTURE, so the output resistance must be reduced compared with the case when VBS=0.
 

Status
Not open for further replies.

Similar threads