overdrive voltage (Vgs-Vth)?

[patriot]

overdrive voltage

In 0.18um CMOS process, with supply voltage of 1.8V. How much of the overdrive voltage to assure the MOS in saturation? Is it reliable of 80mV?

 

[PaloAlto]

vth voltage

It sounds a little bit scarce, however, why don't you make a plot of the characteristic curves of the MOS transistor and see where you are?

Make a sweep of Vds fro different Vgs and plot the Id put a line 80mV over vdsat and see how saturated it is

 

[jeffsky520]

overdrive voltage edaboard

I think the answer is 'no'. But as of where MOS is used, 80mV overdrive voltage is acceptable, like as input differential pair.

 

[sutapanaki]

ee240 v*

80mv could be enough for saturation, depending on the value of your Vds. However, most probably your transistor will be in weak inversion or just out of weak inversion into moderate inversion. 80mv is not good for matching, for distortion and frequency behavior. However, it provides good current efficiency since you get the maximum gm for the current you have through this transistor.

 

[patriot]

overdrive voltage for pmos transistor

It sounds a little bit scarce, however, why don't you make a plot of the characteristic curves of the MOS transistor and see where you are?

Make a sweep of Vds fro different Vgs and plot the Id put a line 80mV over vdsat and see how saturated it is

In some books, i see that Vgs-Vth should be at least 200mV? Is it a experiential value?

 

[PaloAlto]

voltage vgs

that's the value I normally try to have, yes 200mV

 

[yschuang]

voltage overdrive of a pmos

200mV is ok and it is ok if Vgs-Vth<200mV. Usually Vgs-Vth=200mV is for reasonalbe spedd and swing. For more details, visit EE240 course materials from berkeley. You can search them in edaboard.

 

[LAS.xs]

nmos transistor voltages vgs

80mv
Tdevice maybe works in the subthreshold region. And the model given by the foundry is not accurate.

 

[patriot]

vgs-vth,vds

200mV is ok and it is ok if Vgs-Vth<200mV. Usually Vgs-Vth=200mV is for reasonalbe spedd and swing. For more details, visit EE240 course materials from berkeley. You can search them in edaboard.

thanks! yes, it is also 200mV in Sansen's book Analog design essentials.
BTW: there is another parameter of Vdsat in Spectre, Shall we assure that Vds>Vdsat and Vds>Vgs-Vth.
Is Vdsat always less than Vgs-Vth in saturation?