How to find value of cox and kn ???

[Neal Modi]

I have following data. How can I find value of cox and Un and Kn ???

* p18 model card
.MODEL NMOS NMOS
+ VERSION = 3.1
+ LEVEL = 49 NOIMOD = 1 TNOM = 2.70E+01
+ TOX = '4.1E-9/proc_delta' XJ = 1.00E-07 NCH = 2.33E+17
+ VTH0 = '0.36+vt_shift' K1 = 5.84E-01 K2 = 4.14E-03
+ K3 = 1.01E-03 K3B = 2.20E+00 W0 = 1.00E-07
+ NLX = 1.81E-07 DVT0W = 0.00E+00 DVT1W = 0.00E+00
+ DVT2W = 0.00E+00 DVT0 = 1.73E+00 DVT1 = 4.38E-01
+ DVT2 = -3.70E-04 U0 = '260*proc_delta*proc_delta' UA = -1.38E-09
+ UB = 2.26E-18 UC = 5.46E-11 VSAT = 1.03E+05
+ A0 = 1.92E+00 AGS = 4.20E-01 B0 = -1.52E-09
+ B1 = -9.92E-08 KETA = -7.16E-03 A1 = 6.61E-04
+ A2 = 8.89E-01 RDSW = 1.12E+02 PRWG = 4.92E-01
+ PRWB = -2.02E-01 WR = 1.00E+00 WINT = 7.12E-09
+ LINT = 1.12E-08 XL = -2.00E-08 XW = -1.00E-08
+ DWG = -3.82E-09 DWB = 8.63E-09 VOFF = -8.82E-02
+ NFACTOR = 2.30E+00 CIT = 0.00E+00 CDSC = 2.40E-04
+ CDSCD = 0.00E+00 CDSCB = 0.00E+00 ETA0 = 3.13E-03
+ ETAB = 1.00E+00 DSUB = 2.25E-02 PCLM = 7.20E-01
+ PDIBLC1 = 2.15E-01 PDIBLC2 = 2.23E-03 PDIBLCB = 1.00E-01
+ DROUT = 8.01E-01 PSCBE1 = 5.44E+08 PSCBE2 = 1.00E-03
+ PVAG = 1.00E-12 DELTA = 1.00E-02 RSH = 6.78E+00
+ MOBMOD = 1.00E+00 PRT = 0.00E+00 UTE = -1.50E+00
+ KT1 = -1.10E-01 KT1L = 0.00E+00 KT2 = 2.19E-02
+ UA1 = 4.28E-09 UB1 = -7.62E-18 UC1 = -5.57E-11
+ AT = 3.30E+04 WL = 0.00E+00 WLN = 1.00E+00
+ WW = 0.00E+00 WWN = 1.00E+00 WWL = 0.00E+00
+ LL = 0.00E+00 LLN = 1.00E+00 LW = 0.00E+00
+ LWN = 1.00E+00 LWL = 0.00E+00 CAPMOD = 2.00E+00
+ XPART = 5.00E-01 CGDO = 6.98E-10 CGSO = 7.03E-10
+ CGBO = 1.00E-12 CJ = '9.8e-4/proc_delta' PB = 7.34E-01
+ MJ = 3.63E-01 CJSW = '2.4e-10/proc_delta' PBSW = 4.71E-01
+ MJSW = 1.00E-01 CJSWG = 3.29E-10 PBSWG = 4.66E-01
+ MJSWG = 1.00E-01 CF = 0.00E+00 PVTH0 = -7.16E-04
+ PRDSW = -6.66E-01 PK2 = 5.92E-04 WKETA = 2.14E-04
+ LKETA = -1.51E-02 PU0 = 3.36E+00 PUA = -1.31E-11
+ PUB = 0.00E+00 PVSAT = 1.25E+03 PETA0 = 1.00E-04
+ PKETA = 6.45E-04 KF = 4.46E-29

 

[Mahmoud_Sofy]

there are two options:
1- create simple testbench, knowing Vgs, W, and L, and Id will be calculated in the simulation, then you can substitute in the square law and find out (approx.) u.Cox
2- you can find out the equations for the model you are using, hence you can calculate the parameters you need (e.g. you can check this https://nitkkr.ac.in/docs/MOSFET_LEVELs.pdf)

 

[dick_freebird]

You have TOX, COX is epp0*eppSi*W*L/TOX
(though there's optical TOX and electrical TOX,
in most WAT parameters, and for FET models
you would use the electrical TOX while optical
is used for oxide growth process control).

k' is U0*Cox. You'd have to go dig in the PDK
include chain to find how '260*proc_delta*proc_delta'
resolves for U0, then use COX as above.