Does anyone know Early voltage ectraction in Cadence?

Status
Not open for further replies.
D

dtzounakos

Member level 3
Joined
Mar 16, 2007
Messages
56
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,662
Hi!
I want to ask if anyone knows how to extract accurately Early Voltage in Cadence using a command for model library!

Thanks!
 

Do you mean the real transistor, to parameterize the model?
Or do you mean how to display the model VA value?
 

How to display-plot VA versus e.g:VGS (for mosfet transistor)!
 

dtzounakos said:
How to display-plot VA versus e.g:VGS (for mosfet transistor)!
Click to expand...

Yes, it should be possible by using some kind of performance analysis.
The Early voltage VA is determined by the slope of the characteristic
curve ID vs VDS and is defined by the following equation:

slope m=ID/(VA+VDS) for each VGS parameter.

Therefore, your simulation program must find the slope (differential quotient) of the several curves at the operational points defined by the triple (ID,VDS,VGS).
Then, the program must be able to solve the above equation for VA and display VA=f(VGS) for a fixed pair (ID,VDS).
 
Reactions: anhnha

    A

    anhnha

    Points: 2
    Helpful Answer Positive Rating
So the right procedure is:

a)DC Simulation (VDS sweep variable)
b)parametric analysis (VGS variable: 0.1 V step size)
c)plot VA=IDS/deriv(ids vs vds)
c)from this step i don't know how to get the desired VA for a specific VDS value
and then plot VA vs VGS!!

Can you tell me analytically or show me snapshots step-by-step how to plot

VA vs VGS? Is there any command on calculator?( "cross" command?)

Thanks!!!
 

I am not familiar with your simulator engine - therefore, I cannot explicitely list the relevant steps.
But a general question: What do you expect from this analysis ?
My recommendation is instead: Take one characteristic curve ID=f(VDS) and try to expand the slope of the curve to the left (asymptote to negative VDS values) until this line crosses the horizontal axis. Then you have the Early voltage VA which with a good approximation is independent on the VGS voltage because all these lines meet in the vicinity of VA.
 
Reactions: anhnha

    A

    anhnha

    Points: 2
    Helpful Answer Positive Rating
Because i want t plot VA vs GmoverID!I want to be sure about VA expression!

And VGS changes VA!

If you don't know, it's ok.Thanks for your advise!
 

VA=f(Rout) and Rout=f(VGS).

So, VA=f(VGS)!

Try to plot VA versus VGS for a constant VDS and you will see the change of VA.
 

Re: Early voltage dependency on VGS

dtzounakos said:
VA=f(Rout) and Rout=f(VGS).

So, VA=f(VGS)!

Try to plot VA versus VGS for a constant VDS and you will see the change of VA.
Click to expand...
The Early voltage doesn't change essentially with VGS, see the PDF below (there, the "official" BSIM designation is used: VM, the channel length modulation voltage, instead of VA = Early voltage).
VM=VA depends very much on the channel length L, however.
 

Choose the transistor.
Right click on it.
select "Edit Pspice Model"
In the bottom you can read "Vaf= 100" for example
That is the Early voltage.
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…