parameters extraction using hspice

[endyson]

anyone who can help me explain the plot below?

I am using hspice extracting parameters of nmos transistors, I set the transistor worked in the saturated region, fixed the gate voltage and sweeped the drain voltage, but got result shown below, the upper is the drain current(Y axis) versus drain voltage (X axis) ,the plot below is the derivative of the upper, Y is the derivative X is the drain voltage, transistor size is w=2u l=0.35u.

anyone who can explain way the derivative of the current decrease and then increase as the drain voltage growing?
listed here my netlist,
nmos parameters
***************************************title
.lib 'sm046005-1f.hspice' typical
***************************************library
m1 d g gnd gnd nmos_3p3 w=2u l=0.35u

***************************************
vg g gnd dc 1
vd d gnd dc 3.3
***************************************
.dc vd 1 3 0.01
.options post
.end
[/img]

 

[endyson]

nobody knows?

 

[erikl]

... explain why the derivative of the current increases and then decreases again as the drain voltage is growing?

I suspect: inadequate modelling, which doesn't take into account weak inversion (sub-threshold).

 

[standup]

I supposed that the curve is absolutely normal for the standard model. The effect of channel length modulation and Vth which will decrese with Vds will make Ids increase with the Vds, while the effect of hot carrier will degenerate the Ids. As a result, the two powers will fight for the major status, the result of derivation will change with the Vds

 

[endyson]

I got it, the sub-threshold current increase in exponential fashion initially, and linearly because of channel length modulation, and then came in to the saturation.
thank u for your reply!