Qspice not working properly with BSIM4 Models

I know that Qspice (like LTspice) is not really meant for simulating integrated circuits but I nevertheless tried with the BSIM4 models I had for a TSMC process. Micro-cap and ngspice yield predictable results when using the model files (Id for a 1u/100n NMOS is ~100uA) but qspice is giving me 10s of amperes of current ! And when i tried using BSIM3 model files for an older CMOS process, qspice is behaving well…so it seems the problem is specifically with BSIM4 models.

From the output log it seems, qspice is ignoring some parameters regarding resistance and well-proximity effect (the latter shouldn’t cause any problems cuz iirc LTSPICE doesn’t take it into account either) here is the log for reference:

Subcircuit expansion time: 1.58275
Warning: BSIM4: Ignoring unknown model parameter "RGEOMOD" in RGEOMOD=1 RBPB=50 RSHG=15.953 RBPS=50 RBPD=50 NGCON=1 BINUNIT=2 PARAMCHK=1 MOBMOD=0 CAPMOD=2 IGCMOD=1 IGBMOD=1 DIOMOD=1 RDSMOD=1 RBODYMOD=0 RGATEMOD='1*RGFLAG' PERMOD=1 ACNQSMOD=0 TRNQSMOD=0 TEMPMOD=0 TNOM=25 TOXE='TOXN' TOXM=2E-009 DTOX=2.4905E-010 EPSROX=3.9 WINT=0 LINT=2.8625E-10 LL=0.0050965 WL=0 LLN=-0.875 WLN=1 LW=0 WW=0 LWN=1 WWN=1 LWL=0 WWL=0 LLC=0.0050965 WLC=0 LWC=0 WWC=0 LWLC=0 WWLC=0 XL='-1.46E-08+DXLN' XW='1.6E-08+DXWN' DLC='2.8625E-10+DDLCN' DWC=0 XPART=1 TOXREF=3E-009 DLCIG=2.5E-009 VTH0='0.31703+DVTHN' K1=0.24753 K2='-0.0041123731+DK2N' KVTH0=19E-10 PKVTH0=0 WKVTH0=5E-8 LKVTH0=-0.4E-8 LK2='1.8164781E-009+DLK2N' WK2='-2.2983394E-009+DWK2N' PK2='2.4569193E-016+DPK2N' K3=-3.98252 K3B=2.24786 W0=0 DVT0=1.08116 DVT1=1.18773 DVT2=0.015579 DVT0W=0 DVT1W=0 DVT2W=0 DSUB=0.5 MINV=-0.46494 VOFFL=0 DVTP0=4.20003E-007 DVTP1=0 LPE0=6.10104E-009 LPEB=4.55849E-008 XJ=1.1E-007 NGATE=3.1592E+020 NDEP=6.3687E+017 NSD=1E+020 PHIN=0 CDSC=0 CDSCB=0 CDSCD=0 CIT=-0.00014804749 LCIT='4.6261635E-010+DLCITN' WCIT='1.4753784E-009+DWCITN' PCIT='-2.6857591E-016+DPCITN' VOFF='-0.12931886+DVOFFN' LVOFF='-2.6411604E-009+DLVOFFN' WVOFF='2.3858227E-010+DWVOFFN' PVOFF='8.400871E-016+DPVOFFN' NFACTOR='0.6+DNFACTORN' ETA0='0.12702749+DETA0N' LETA0='5.8414043E-013+DLETA0N' WETA0='2.2435567E-008+DWETA0N' PETA0='-5.3897131E-019+DPETA0N' ETAB=-0.06480454 LETAB=-2.8262198E-014 WETAB=7.08999E-008 PETAB=2.3109453E-020 U0='0.022320968+DU0N' LU0='-2.24302E-010+DLU0N' KU0=-2.9E-8 PKU0=2E-14 WKU0=1.2E-7 LKU0=2.2E-7 WU0='-1.4555707E-009+DWU0N' PU0='-2.3976379E-016+DPU0N' UA=-1.4967634E-009 LUA=-2.7226435E-017 WUA=1.0438898E-017 PUA=1.521437E-024 UB=2.0634228E-018 LUB=1.0345374E-026 WUB=-2.3715874E-026 PUB=-2.6904759E-032 UC=7.91364E-11 LUC=7.1196346E-018 WUC=4.720788E-018 PUC=-8.3670699E-024 VSAT='101530+DVSATN' A0='0.40094308+DA0N' LA0='-3.4388967E-009+DLA0N' KVSAT=1 WA0='-4.4175841E-007+DWA0N' PA0='1.2490971E-013+DPA0N' AGS='6.478622+DAGSN' LAGS='-2.7723817E-007+DLAGSN' WAGS=-2.2840519E-006 PAGS='8.3507413E-014+DPAGSN' A1=0 A2=1 B0=0 B1=0 KETA=-0.1 DWG=0 DWB=0 PCLM='-0.12099972+DPCLMN+0.00201949*SS_FLAG-0.00269265*FF_FLAG+0.00215412*SF_FLAG-0.00215412*FS_FLAG' LPCLM='4.230048E-008-4.22598E-10*SS_FLAG+5.63463E-10*FF_FLAG-4.50771E-10*SF_FLAG+4.50771E-10*FS_FLAG' WPCLM='8.0404227E-008-2.02272E-08*SS_FLAG+2.69696E-08*FF_FLAG-2.15756E-08*SF_FLAG+2.15756E-08*FS_FLAG' PPCLM='-1.7082899E-014+DPPCLMN+4.23274E-15*SS_FLAG-5.64365E-15*FF_FLAG+4.51492E-15*SF_FLAG-4.51492E-15*FS_FLAG' PDIBLC1=0 PDIBLC2=0.018624584 LPDIBLC2=-1.110685E-009 WPDIBLC2=4.0846591E-010 PPDIBLC2=1.4174614E-016 PDIBLCB=-0.03716 DROUT=0.5 PVAG=0.7 DELTA=0.01 PSCBE1=4.24E+008 PSCBE2=1E-020 FPROUT=0 PDITS=0.44984 PDITSD=1 PDITSL=264000 RSH=16 RSW='45+DRDSWN' RDW='45+DRDSWN' PRWG=0 PRWB=0 WR=1 ALPHA0=3.3929E-010 ALPHA1=0.0034 BETA0=5.808 AGIDL=1.5797E-008 BGIDL=1.4527E+009 CGIDL=1.4892 EGIDL=0.81976 AIGBACC=0.013432 BIGBACC=0.0065981 CIGBACC=0.39315 NIGBACC=4.8 AIGBINV=0.024468 BIGBINV=0.012141 CIGBINV=0.0855 EIGBINV=1.1 NIGBINV=3.8771 AIGC=0.010744 BIGC=0.000824385 CIGC=-0.087762 AIGSD=0.0093701 BIGSD=0.0012397 CIGSD=0.081703 NIGC=1.5 POXEDGE=1 PIGCD=3.11 NTOX='NTOXN' XRCRG1=12 XRCRG2=1 CGSO='CGON' CGDO='CGON' CGBO=0 CGDL='CGLN' CGSL='CGLN' CLC=1E-007 CLE=0.6 CF='CFN' CKAPPAS=0.6 CKAPPAD=0.6 ACDE=0.6 MOIN=7.6133 NOFF=2.4067 VOFFCV=-0.12779 KT1='-0.25635598+DKT1N' LKT1='-6.7706745E-009+DLKT1N' WKT1='-1.5116486E-008+DWKT1N' PKT1='1.9315376E-015+DPKT1N' KT1L=0 KT2='-0.035543769+DKT2N' LKT2='-1.567168E-009+DLKT2N' WKT2='6.566886E-009+DWKT2N' PKT2='-1.4608393E-015+DPKT2N' UTE=-1 UA1='3.8485974E-009+DUA1N' LUA1='-1.6206912E-016+DLUA1N' WUA1='-9.2334584E-016+DWUA1N' PUA1='3.7887376E-023+DPUA1N' UB1='-4.343838E-018+DUB1N' LUB1='2.1211141E-025+DLUB1N' WUB1='1.2073009E-024+DWUB1N' PUB1='-6.6353287E-032+DPUB1N' UC1=4.932709E-12 LUC1=-6.7227147E-019 WUC1=-1.0399171E-017 PUC1=6.9546911E-024 PRT=0 AT='120004.69+DATN' LAT='-0.0027872146+DLATN' WAT='0.013700549+DWATN' PAT='-3.3224603E-009+DPATN' JSS=1.35E-07 JSD=1.35E-07 JSWS=2.5E-13 JSWD=2.5E-13 JSWGS=2.5EDevice read time: 3.06881
Z:\home\koustubh\eda\TestTSMC65.qsch
Warning: No WPE as none of SCA, SCB, SCC, SC is given and/or SC not positive.
Warning: Source conductance reset to 1.0e3 mho.
Warning: Drain conductance reset to 1.0e3 mho.
Note: QSPICE has BSIM4.8.1.  Model NCH.9 specifies an earlier version: BSIM4.5
Warning: No WPE as none of SCA, SCB, SCC, SC is given and/or SC not positive.
Warning: Source conductance reset to 1.0e3 mho.
Warning: Drain conductance reset to 1.0e3 mho.

Summary: When using a proprietary BSIM4 model library, I am getting 10s of amperes of drain current for a gate voltage sweep from 0-1V for a 1u/100n MOSFET, which is definitely not possible for the CMOS process (ig the device will burn out)