Hello all I have a question about transistor regio...
# analog-design
e
Hello all I have a question about transistor region of operation in ngspice In the below picture there is a parameter called (Vdsat) is this Vov?
c
Roughly speaking, vdsat is the drain-to-source voltage above which the drain-to-source current is (almost) independent of the drain-to-source voltage vds. However, this parameter is inherited from old MOS models (level 2 or 3) and today's models (BSIM3 or BSIM4) are more sophisticated and more complicated. You'd have to look up the detailed model description (I don't know if they are included in the ngspice manual). I'm not familiar with Vov.
Anyhow, how did you convince ngspice to show you the list of transistor operating point parameters? I know how to get them in Cadence/Spectre and LTspice, but not yet in ngspice.
e
ok I understande you, so I can check the region by check that Vds > Vgs - Vth or not
c
Yes, if you set the operating point of your transistor above threshold.
e
Using command
show
πŸ‘ 1
or Vds > Vdsat ?
c
No, Vgs>Vth
e
to ensure sat ?
not just ON
c
To a first approximation, MOS transistors turn on only if Vgs>Vth. However, some crazy people operate MOS transistors at Vgs<Vth ("weak inversion"). It turns out that the relative change in drain current as function of Vgs (gm/id) is actually higher than if the MOS transistor is officially on (in "strong inversion")
The specifics of vdsat in the particular case you mention are (probably) explained in way more detail than you would ever want to know in section 5.6 of this manual.
e
@Christoph Maier ok many thanks
c
If you search for vdsat in the ngspice manual, you may find more useful information.
πŸ‘ 1
a
vdsat is same as Vov, so either you say Vds > (vgs - vth or vdsat or vov )
πŸ‘ 1
e
@ArunAshok Ok many thanks
πŸ‘ 1
l
Most designs don't use the overdrive voltage anymore, as it is a simple approximation for the saturation voltage (Vov = Vgs-Vth). This is still valid for transistors operating in strong inversion, where the quadratic model is valid. There is the FET weak inversion operation, which happens approximately when Vds < Vth-100 mV, and the drain current increases exponentially with VGS. For this operation, VDSAT is approximately 4 to 6 Ο†T (100 to 150 mV). See, if Vdsat = Vov = Vgs-Vt, it could be negative! What really happens is that Vdsat is approximately 100 mV for weak inversion and increases until it reaches strong inversion, which happen when Vgs is about Vt+150 mV, then Vov is valid. This region between the weak inversion and strong inversion is called moderate inversion. BSIM4 is partially based in the EKV model, which has its own formula for Vdsat, which can be shown as even lower than 4Ο†T. https://www.researchgate.net/publication/283772239_An_Analytical_MOS_Transistor_Model_V[…]_and_Dedicated_to_Low-Voltage_and_Low-Current_Applications
πŸ‘ 2
e
Thanks @Luis Henrique Rodovalho
v
You hit on something. This is one of the few parameters where there is some controversy as to its name. Different authors have their own preferences. I've mostly heard people call it 'vee dee sat', that is, vdsat like SPICE calls it. Some prefer 'vee dee esse sat', that is, Vds-sat. Then come those who reject the wise ways of SPICE all together and refer to it as 'the overdrive voltage', Vov, or 'the excess bias', Veb. All of these terms are referring to the same thing, however.
πŸ‘ 1
e
Thanks @Vern
j
Vdsat means the drain-source voltage above which the device is considered to be operating in its saturation region as opposed to its triode region, hence the suffix 'sat' for saturation.
πŸ‘ 1