The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Body Effect
https://designers-guide.org/forum/YaBB.pl?num=1131605603

Message started by vijay on Nov 9^th, 2005, 10:53pm

Title: Body Effect
Post by vijay on Nov 9^th, 2005, 10:53pm

Hi,
I simulated a single NMOS in saturation.The saturation current is 2.13mA.For the same setup I introduced a volatage at the bulkand source at gnd in my NMOS.Ideally the current should decrease as Vth increases.But the simulation results speak:-
1.When there was no body effect Id = 2.13mA
2.The current increases to 4mA max, when the bulk volatge is = 700mV and after that it started decreasing.FInally at Vbulk=3V the current reduced to a very low value.

What has caused that sudden increase in the saturation current?Is it that that Isubstrate is not considered in normal mos current equation?
Why my device follwed the theory of Id-decrease with Vbulk after 700mV and not immediately.

Thanks in advance.
V

Title: Re: Body Effect
Post by uncle_ezra on Nov 11^th, 2005, 12:14am

I believe wat u r seeing is usage of bulk-driven MOS. Let me summarise.

Normally NMOS bulk is connected to the most negative supply in order to avoid forward-biasing the PN-junction.But it is known that slightly foward-biasing Vbs decreases Vth, hence u see the increase the current. However, if Vbs > Vdiode ~ 0.7V you end up forward biasing the PN junction.

For more information about bulk-driven MOS go to
http://amesp02.tamu.edu/~sanchez/607-bulkdriven-2005.pdf

Hope this helps ! Let me know if you need a clearer explanation

Title: Re: Body Effect
Post by vijay on Nov 13^th, 2005, 9:20pm

Thanks,
The link is of good help.My actual doubt for the setup was i needed an explanation for the following setup:-

the set up is :- single stage amplifier is before a differential amplifier. The single stage amplifier (CS) has its source and bulk connected to ground to have no body bias effect.
The next stage differential amplifier would in that case have its bulk already connected to ground,as in a n-well process all nmos share the same substrate.
My question is in that case there would be a body bias effect in the DA?If the source and bulk in the DA is connected together and then connected to a tail current source.
If the single stage CS has its bulk grounded then in fabrication the bulk of the DA would also be at ground.
So what is the preferred configuration:-
1.Source and bulk connected together to a tail current source?
2.Since the bulk would in the process of fabrication be connected to ground we consider the bulk at ground and then simulate?
Thanks in advance.
V

Title: Re: Body Effect
Post by uncle_ezra on Nov 13^th, 2005, 10:54pm

My usual practice is to connect NMOS bulk to gnd unless I'm using a triple well process. So even though in a DA the current source may be at AC gnd I dont connect the source and bulk together due to process, temperate and voltage variation.

Title: Re: Body Effect
Post by vijay on Nov 14^th, 2005, 1:01am

Thank you very much.
Regrads,
V.

Title: Re: Body Effect
Post by ice_wind on Nov 24^th, 2005, 6:58pm

uncle_ezra wrote on Nov 13^th, 2005, 10:54pm:

That is right. By the way, if connect the source and bulk together, then a cap will exist between this node and ground, for high frequency it is not a good news.