Go to www.awrcorp.com
Back to search page Click to download printable version of this guide.

Philips Level 903 Geometric N MOSFET Model: MOS9



MOS9 is an implementation of the Philips MOS model 9 level 903 device. The model is intended for the simulation of circuit behavior with emphasis on analogue applications. It provides an excellent description of the electrical characteristics in all relevant regions of transistor operation. The underlying model is based on the gradual-channel approximation with a number of first order corrections for small size effects, which allows to describe transistor behavior over a wide range of channel lengths and widths using just one parameter set. The consistency is maintained by using the same carrier-density and electrical-field expression in the calculation of all model quantities.


Name Description Unit Type Default
ID Element ID Text M1
*TNOM Parameter extraction temperature DegC 21
*TEMP Device temperature DegC 21
*NFLAG Noise flag   Noise On
*LER Effective channel length of reference transistor um 1.1
*WER Effective channel width of reference transistor um 20
*LVAR Difference between actual and programmed poly-silicon gate lengths um -0.22
*LAP Effective channel length reduction per side um 0.1
*WVAR Difference between actual and programmed filed-oxide opening um -0.025
*WOT Effective channel width reduction per side um 0
*VTOR Threshold at zero back-bias of reference transistor V .73
*STVTO Threshold voltage temperature of coefficient (VK^-1)   -0.0012
*SLVTO Threshold voltage length dependence first coefficient (Vm)   -1.35e-7
*SL2VTO Threshold voltage length dependence second coefficient (Vm^2)   0
*SL3VTO Threshold voltage length dependence third coefficient V  
*SWVTO Threshold voltage width dependence coefficient (Vm)   1.3e-7
*KOR Low back-bias body factor (V^0.5)   0.65
*SLKO Low back-bias body factor length dependence first coefficient (V^0.5m)   -1.3e-7
*SL2KO Low back-bias body factor length dependence second coefficient (V^0.5m^2)    
*SWKO Low back-bias body factor width dependence coefficient (V^0.5m)   2e-9
*KR High back-bias body factor (V^0.5)   0.11
*SLK High back-bias body factor length dependence first coefficient (V^0.5m)   -2.8e-7
*SLK2 High back-bias body factor length dependence second coefficient (V^0.5m^2)    
*SWK High back-bias body factor width dependence coefficient (V^0.5m)   2.75e-7
*PHIBR Strong inversion surface potential V 0.65
*VSBXR Transition voltage for the dual-k factor model V 0.66
*SLVSBX VSBXR length dependence coefficient (Vm)   0
*SWVSBX VSBXR width dependence coefficient (Vm)   -6.75e-7
*BETSQ Gain factor for the infinite square transistor (AV^.2)   8.3e-5
*ETABET Gain factor temperature dependence exponent   1.6
*LP1 Characteristic length of first profile um  
*FBET1 Relative mobility decrease due to first profile    
*LP2 Characteristic length of second profile um  
*FBET2 Relative mobility decrease due to second profile    
*THE1R Gate induced field mobility reduction coefficient (V^-1)   0.19
*STTHE1R THE1R temperature dependence coefficient (V^-1K^-1)   0
*SLTHE1R THE1R length dependence coefficient (V^-1m)   1.4e-7
*STLTHE1 THE1R length dependence temperature coefficient (V^-1mK^-1)   0
*GTHE1 THE1R scaling rule selector 0:old, 1:new    
*SWTHE1 THE1R width dependence coefficient (V^-1m)   -5.8e-8
*WDOG Characteristic drawn gate width, below which dogboning appears um 0
*FTHE1 THE1R geometry dependence coefficient    
*THE2R Back-bias mobility reduction coefficient (V^-.5)   0.012
*STTHE2R THE2R temperature dependence coefficient (V^-.5K^-1)   0
*SLTHE2R THE2R length dependence coefficient (V^-0.5m)   -3.3e-8
*STLTHE2 THE2R length dependence temperature coefficient (V^-0.5mK^-0.5)   0
*SWTHE2 THE2R width dependence coefficient (V^-0.5m)   3e-8
*THE3R Lateral field mobility reduction coefficient (V^-1)   0.145
*STTHE3R THE3R temperature dependence coefficient (V^-1k^-1)   -0.00066
*SLTHE3R THE3R length dependence coefficient (V^-1m)   1.85e-7
*STLTHE3 THE3R length dependence temperature coefficient (V^-1mK^-1)   -6.2e-10
*SWTHE3 THE3R width dependency coefficient (V^-1m)   2e-8
*GAM1R Drain induced threshold shift for large gate drive coefficient (V^(1-nDS))   0.145
*SLGAM1 GAM1R length dependence coefficient (V^(1-nDS)m)   1.6e-7
*SWGAM1 GAM1R width dependence coefficient (V^(1-nDS)m)   -1e-8
*ETADSR GAM1R Vds dependence exponent   0.6
*ALPR Channel length modulation factor   0.003
*ETAALP ALPR length dependence exponent   0.15
*SLALP ALPR length dependence coefficient (m^nALPHA)   -0.00565
*SWALP ALPR width dependence coefficient   0.00167
*VPR Channel length modulation characteristic voltage V 0.34
*GAMOOR Zero gate drive drain induced threshold shift coefficient   0.018
*SLGAMOO GAMOOR first length dependence coefficient (m^2)   2e-14
*SL2GAMOO GAMOOR second length dependence coefficient    
*ETAGAMR GAMOOR back-bias dependence exponent   2
*MOR Subthreshold slope factor   0.5
*STMO MOR temperature dependence coefficient (K^1)   0
*SLMO MOR length dependence coefficient (M^0.5)   0.00028
*ETAMR MOR back-bias dependence exponent   2
*ZET1R ZET1R length dependence exponent   0.42
*ETAZET ZET1R length dependence exponent   0.5
*SLZET1 ZET1R length dependence coefficient (m^nGAMMA)   -0.39
*VSBTR MOR and GAMOOR Vsb dependence limiting voltage V 2.1
*SLVSBT VSVTR length dependence coefficient (Vm)   -4.4e-6
*A1R Weak-avalanche current factor   6
*STA1 A1R temperature dependence coefficient (K^-1)   0
*SLA1 A1R length dependence coefficient um 1.3
*SWA1 A1R width dependence coefficient um 3
*A2R Weak-avalanche current exponent V 38
*SLA2 A2R length dependence coefficient (Vm)   1e-6
*SWA2 A2R width dependence coefficient (Vm)   2e-6
*A3R Weak-avalanche drain-source voltage factor   0.65
*SLA3 A3R length dependence coefficient um -0.55
*SWA3 A1R width dependence coefficient um 0
*TOX Gate-oxide layer thickness um 0.025
*COL Gate overlap capacitance per unit channel width (Fm^-1)   3.2e-10
*NTR Thermal noise coefficient (J)   2.44e-20
*NFMOD Flicker noise model selector. 0:old, 1:new   0
*NFR Flicker noise coefficient (V^2)   7e-11
*NFAR First flicker noise coefficient (V^-1m^-4)    
*NFBR Second Flicker noise coefficient (V^-1m^-2)    
*NFCR Third Flicker noise coefficient (V^-1)    
*L Lay-out drawn channel length um 1.5
*W Lay-out drawn channel width um 20
*MULT Number of devices in parallel   1
*THE3MOD Flag for THE3R clipping    

* indicates a secondary parameter

The default and restrictions on the range of parameter values is in full compliance with the model definition.

Operating Point Information

The following letter pairs have been used to identify the NL branches: ds, bs, bd, gs and gd

Parameter Description
pwr (W) Power.
gm (S) Transconductance .
gmb (S) Bulk transconductance .
gds (S) Output conductance .
vgt1 (V) Threshold voltage including backbias effects.
vgt2 (V) Effective gate drive including backbias and drain effects.
vds1 (V) Saturation limit.
vdss1 (V) Saturation voltage at actual bias.
cgg (F)
cgb (F)
cgd (F)
cgs (F)
cbg (F)
cbb (F)
cbd (F)
cbs (F)
cdg (F)
cdb (F)
cdd (F)
cds (F)
csg (F)
csb (F)
csd (F)
css (F)

. Here g,d,s abd b correspond to the gate, drain, source and substrate terminals, respectively. These are used to identify branch related operating point information, i.e., branch voltages, currents, etc.

Implementation Details

This model is mapped into HSPICE as a NMOS M-device with parameters LEVEL set to 50, which corresponds to the Philips MOS model 9 level 902 device, whose parameter set is a subset of the one of the level 903 device. Changing any of the parameters not supported by HSPICE into something different than its default value will result in a warning being issued by the simulator. For the complete set of equations see Ref. [1]


This element does not have an assigned layout cell. You can assign artwork cells to any element. See “Assigning Artwork Cells to Layout of Schematic Elements” for details.


[1] R.M.D.A. Velghe, D.B.M. Klaassen, and F.M. Klaassen, "Unclassified Report NL-UR 003/94".

Legal and Trademark Notice