# Spice model, can't figure it out

I am using LTSpice with the following NMOS model from the library included in the famous book 'Sedra & Smith, Microelectronic circuits' :

*       Level-1 Model for the 0.5-um NMOS Transistor (Part NMOS0P5)
*       (created by Anas Hamoui & Olivier Trescases)
.model  NMOS0P5 NMOS(Level=1 VTO=0.7 GAMMA=0.5 PHI=0.8
+       LD=0.08E-06 WD=0 UO=460 LAMBDA=0.1 TOX=9.5E-9 PB=0.9 CJ=0.57E-3
+       CJSW=120E-12 MJ=0.5 MJSW=0.4 CGDO=0.4E-9 JS=10E-9 CGBO=0.38E-9
+       CGSO=0.4E-9)


Well, to analyze my simulations, I try to make my pencil and paper analysis to notice possible differences. However, I need to know several things about my MOS, and I can't find the factor ' K = mu_n * C_ox '. Where should I find it ? Under which name ? I see that often other MOS models for SPICE include this factor in the model definition, which I can't find in the one I am using.

For information, this model, under SPICE, works perfectly well.

Thanks a lot for your help, Ryl

• According to this page the default is KP=2E-5. So I would try adding that parameter to the model and seeing if the result changes. – mng Oct 10 '12 at 19:28

## 2 Answers

It took some digging, but I found some definitions:

$\mu_n$ is the electron mobility, which is $1360 \frac{cm^2}{Vs}$

$C_{ox}$ is the capacitance per unit area of the oxide layer, which is $\dfrac {\epsilon_{ox}}{t_{ox}}$

${\epsilon_{ox}} = 3.45 \times 10^{-11} \frac{F}{m}$

$t_{ox}$ is the thickness of the oxide layer.

The only unknown appears to be the thickness of the oxide layer, which is already in your model (TOX = 9.5E-9), so you should be good to go.

There are some of method to calculate: $$k=\mu_n\times C_{ox}$$

Example: parameter extraction for more accuracy: Calculate: Vth & K of MOSFET 