Hi all,
There is no λ(channel length modulation coefficent) in the model file. And λ need to be known in the hand calculation. How to obtain the λ? Please help!!!
What I did for determining lambda is to determine the slope of the IDvsVDS curve of the transistor in saturation. Now, draw a line through a point of the IDvsVDS curve with the determined slope and determine the intersection with the ID = 0 curve. This results in the point (0, -1/lambda).
This is easily seen from the next level 1 model equation for a transistor in saturation:
ID = 1/2*mu*Cox*W/L*(VGS-VT)^2*(1+lambda*VDS).
lambda is function of L and VGS, so their might be some averaging involved.
It's true, deviations from experiments aren't very accurate. Although lambda is a very common parameter for equations in gain estimation of an opAmp. So lambda does have a practical meaning.
If you have a better idea and more accurate means to determine small signal output impedances, feel free to share your knowledge.
well ,, you can make sweep by simulation ,,and measure the slope of drain current in the saturation region,, this vakue will be 1/ro and ro=1/(lamda*Id)