The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Semiconductor Devices >> meaning of vdsat
https://designers-guide.org/forum/YaBB.pl?num=1101762622

Message started by heaton on Nov 29^th, 2004, 1:10pm

Title: meaning of vdsat
Post by heaton on Nov 29^th, 2004, 1:10pm

guys,
someone told me that vdsat is nothing but vgs-vth. however, my dc operating point report of a 0.18u nmos shows that vgs=515.5m, vth=492.9m, where vdsat=81m, a lot larger than vgs-vth. i'm really confused, any advice would be appreciated.
thanks.

Title: Re: meaning of vdsat
Post by extractor_man on Nov 29^th, 2004, 10:57pm

Vdsat is your maximum drain source voltage
A value limited mainly by hot carrier injection in short channel MOS transistor.

Title: Re: meaning of vdsat
Post by crhu on Dec 21^st, 2004, 4:04pm

Hi, heaton,
What region does your transistor work? You need to
make sure that the transistor works at staturation/pinch
off region.

Vdsat, (aka veff (effective voltage) and delta v), is used to
check if the transistor works at saturation.

Vdsat is calculated with Ids and w/l, sqrt((2*Ids)/(kp*w/l)).
You can derive it from the saturaton current equation w/o channel length
modulation:
Ids = kp/2*(w/l)*(Vgs-vth)^2
Vdsat = vgs -vth @ saturation

If vgs > vth, and vds > vdsat, the transistor is in saturation.
In this case, vdsat=vgs-vth if channel length modulation
is omitted.
If vgs > vth, and vds < vdsat, in linear region.
In this case, vdsat != vgs-vth, because in linear
region there is no square law bt. Ids and (Vgs-vth).

Title: Re: meaning of vdsat
Post by heaton on Dec 24^th, 2004, 11:08am

crhu,

My vds is around 400mv, larger than vdsat.

Title: Re: meaning of vdsat
Post by crhu on Dec 27^th, 2004, 6:04pm

Hi, heaton,
May I know which simulator you are using, Hspice or Spectre?
If you can upload your netlist, I can have a look into it.

Thanks.

--Charlie

Title: Re: meaning of vdsat
Post by Geoffrey_Coram on Jan 10^th, 2005, 2:46pm

I believe Vdsat = vgs - vth is an approximation from a time of much simpler MOS models. A glance at the BSIM3 source code shows

Vdsat = EsatL * Vgst2Vtm / (Abulk * EsatL + Vgst2Vtm)

where Vgst2Vtm = (vgs - vth) - 2 * (thermal voltage), EsatL is computed from the VSAT parameter and the effective length, and Abulk has some other dependencies.

Even in MOS level 3, we have vdsat = (vgsx-vth)/Fbody, so the body effect can change vdsat from the simple approximation.

Title: Re: meaning of vdsat
Post by crhu on Jan 11^th, 2005, 1:26pm

Hi,Geoffrey and Keaton,
I went through the mosfet manuals (BSIM3v3 and BSIM4v3)
and you are right that vdsat is not equal to (vgs-vt)
since level3 as small geometry is considered.

I checked it too in a testcase and it shows that
vdsat != vgs -vth
//file: hspice.lis
...
element 0:m1 0:m2 0:m3 0:m4 0:m5
model 0:nch 0:nch 0:pch 0:pch 0:nch
region Saturati Saturati Saturati Saturati Saturati
id 1.2300u 1.2300u -1.2300u -1.2300u 2.4601u
ibs 0. 0. 0. 0. 0.
ibd 0. 0. 0. 0. 0.
vgs 632.2317m 632.2317m -792.4483m -792.4483m 600.0000m
vds 739.7834m 739.7834m -792.4483m -792.4483m 267.7683m
vbs 0. 0. 0. 0. 0.
vth 622.6295m 622.6295m -676.6599m -676.6599m 586.9457m
vdsat 46.5314m 46.5314m -87.5725m -87.5725m 47.4368m
...

Thanks.

--charlie

Title: Re: meaning of vdsat
Post by yilmazerc on Jan 12^th, 2005, 6:14am

hi all
i think ican add something physical
vdsat=vgs-vth is valid for long channel devices

vdsat is smaller for short channells because of the velocity saturation due to the high electric field along the channel