Hi @Amro Tork, not sure if you are the right person to ask but I have some questions regarding those mismatch parameters in the

sm141064.ngspice

. On line 47019, the following parameters define the mismatch properties of `nfet_03v3`:

.param
+ par_vth=0.007148  
+ par_k=0.007008  
+ par_l=1.5e-7
+ par_w=-1e-7
+ par_leff='l-par_l'
+ par_weff='par*(w-par_w)'
+ p_sqrtarea='sqrt((par_leff)*(par_weff))'

My understanding is that

par_vth

(essentially

A_vth

) accounts for the mismatch of `Vth`; but I am not 100% sure of

par_k

, what is this

k

in particular? Is it the same as

\beta=\muCox(W/L)

(which is the notation convention some people follow), or is it the same as

K

defined in this paper (https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=572629) which is a parameter that accounts for the change of

Vth

due to the presence of a source-body voltage (also see the attached picture)? Thanks.

Hi @Chris, I think I'm the right person yes. The parameters here in the model card are filling the values for BSIM4 version. Documentation could be found here: https://bsim.berkeley.edu/models/bsim4/ Naming convention of the model parameters were made by the foundry. I doesn't necessarily map to what you see in literature. The right way to do this is to trace where this parameter is used and how it affects the BSIM 4 equations. Tracing the use of

par_k

, I could see that it used in a variable called

var_k

which seems to be the variance for some variable, then

var_k

is used in

mis_k=agauss(0,var_k,1)

. As you see it's used to generate a random value for variable

mis_k

from a Gaussian distribution that is required for MC simulations. After I went to check that in the code, it seems it's not used anywhere in the model accordingly it doesn't have any effect on the simulation. It should change something in the BSIM4 model parameters if it will affect the transistor behavior. That said, that looks like a bug to me. Good catch. If you want more information about the BSIM model parameters, you could find it at the link above. Appreciate if you could open an issue for it on : https://github.com/efabless/globalfoundries-pdk-libs-gf180mcu_fd_pr/tree/main

@Amro Tork I've filed the issue ticket. Thanks.

Thanks @Chris