Product Documentation
Spectre Circuit Simulator Components and Device Models Reference
Product Version 23.1, June 2023

BSIM-IMG Model Equations

Bias Independent Calculations

Physical Constants

Physical quantities in BSIM-IMG are in M.K.S units, unless specified otherwise.

Effective Channel Width, Channel Length

Binning Calculations

Front and Back Gate Workfunction Calculation

Terminal Voltages and Pre-Conditioning

Terminal Voltages and Vdsx Calculation

Back Gate Biasing Effect

Short Channel Effects

Vt Roll-off

Drain Induced Barrier Lowering (DIBL)

Vt Roll on/off at moderate channel lengths

Vt Roll on/off at moderate channel lengths and high Vds

Body Doping Effects

Body Doping Effect

Surface Potential Calculation

Surface Potential is computed solving the Poisson's equation in silicon body:

where

and

The first equation above is simplified, replacing vtm by nvtm and neglecting the second exponential term on right side. Further approximating   s2 to be

The simplified Poisson’s equation obtained is:

This can rewritten as:

where

Solving the above equation using an iterative Halley’s algorithm for a maximum of four iterations stopping when the error in surface potential solution is below 1 nV. Halley's algorithm is a second order Householder's method and can be written as:

Simplifying the first equation assuming the back-gate to remain always in weak inversion, reducing it to become a function of only the front-gate parameters:

The above equation is solved using two iterations of Halley's algorithm and can be written as:

Once x satisfying is known, s1 is obtained from x using:

The equation is solved once at the source side, with Vch = 0, and once at drain side, with Vch = Vds, to obtain the surface potentials at the two ends of the channel.

Drain Saturation Voltage

The drain saturation voltage model is calculated after the source-side surface po-tential ( s) has been calculated. Vdseff is subsequently used to compute the drain-sidesurface potential ( d).

Electric Field Calculations

Calculate the Drain Saturation Voltage

If Drs=0, then

If Drs is not equal to 0, then

Calculate Average Field, Potential, and Charge

Average Field, Potential, and Charge

Mobility Degradation

The mobility model is based on the BSIM4 model.

Output Conductance

Channel Length Modulation

Output Conductance due to DIBL

Moc is multiplied to Ids in the _nal drain current expression.

Veolocity Saturation

Current Degradation Due to Velocity Saturation

The following formulation models the current degradation factor due to velocity saturation in the linear region. It is adopted from the BSIM5 model.

Drain Current Model

C-V Model

Channel Length Modulation

Assign Variables

Parasitic Resistance and Capacitance Models

BSIM-IMG models the parasitic source/drain resistance in two components, a bias dependent extension resistance, and a bias independent diffusion resistance.

The parasitic capacitance model in BSIM-MG includes a bias-independent outer fringe capacitance, a bias-dependent inner fringe capacitance, a bias-dependent overlap capacitance, and substrate capacitances.

Bias-dependent Extension Resistance

RDSMOD=0 (Internal)

RDSMOD=1 (External)

RDSMOD=2 (Internal and Geometry dependent)

Bias-independent Diffusion Resistance

Overlap Capacitance Model

Outer Fringe Capacitances

Inner Fringe Capacitance

Source/drain to Substrate Capacitances

Impact Ionization and GIDL/GISL Model

Impact Ionization Current

Gate-Induced-Drain/Source-Leakage Current

Gate Tunneling Current

Gate to channel current (Igc)

Gate to source/drain current (Igs, Igd)

Self Heating Model

Figure -1 R-C network for self-heating calculation

Thermal resistance and capacitance calculations

Noise Modeling

Following noise models are available in BSIMIMG 102.6.1:

Flicker Noise Model

In BSIMIMG102.6.1, the flicker noise model is same as FNOIMOD=1 in BSIM4. The unified physical ficker noise model is smooth over all bias regions. The physical mechanism for the ficker noise is trapping/detrapping-related charge fluctuation in oxide traps, which results in fluctuations of both mobile carrier numbers and mobilities in the channel. The unified flicker noise model captures this physical process. In the inversion region, the noise density is expressed as:

The channel length reduction due to channel length modulation and given by:

Charge density at the source side is given by:

Charge density at the drain side is given by:

and N* is given by:

where CIT is a model parameter from DC IV.

In the strong inversion region, the noise density is written as:

In the subthresold region, the noise density is given as:

The total flicker noise density is given by:

Thermal Noise Model

For TNOIMOD=0 (Charge-based model), the noise current is given by:

where Rds(V ) is the bias-dependent LDD source/drain resistance, and the parameter NTNOI is introduced for more accurate fitting of short-channel devices. Qinv is the total inversion charge in the channel. Qs,intrinsic, Qd,intrinsic are intrinsic charges at source/drain ends.

Gate Current Shot Noise

Resistor Noise Model

If RDSMOD=1, then

If RGATEMOD=1, then


Return to top
 ⠀
X