Skip to content

Pulsim Device Model Reference

This document provides detailed information about all device models supported by Pulsim, including their equations, parameters, and usage guidelines.

Table of Contents

  1. Passive Components
  2. Sources
  3. Semiconductor Devices
  4. Switching Devices
  5. Magnetic Components
  6. Thermal Models
  7. Device Library

Passive Components

Resistor

Symbol: R Equation: V = I × R

The resistor is modeled as a linear conductance G = 1/R in the MNA matrix.

MNA Stamp: 

     n+    n-
n+ [ +G   -G  ]
n- [ -G   +G  ]

Parameters: | Parameter | Symbol | Unit | Range | Description | |-----------|--------|------|-------|-------------| | value | R | Ω | > 0 | Resistance |

Capacitor

Symbol: C Equation: I = C × dV/dt

Pulsim uses the companion model approach for time integration.

Backward Euler Companion Model: 

I_eq = (C/dt) × V_new - (C/dt) × V_old
G_eq = C/dt

Trapezoidal Companion Model: 

I_eq = (2C/dt) × V_new - (2C/dt) × V_old - I_old
G_eq = 2C/dt

Parameters: | Parameter | Symbol | Unit | Range | Description | |-----------|--------|------|-------|-------------| | value | C | F | > 0 | Capacitance | | ic | V₀ | V | any | Initial voltage |

Inductor

Symbol: L Equation: V = L × dI/dt

Backward Euler Companion Model: 

V_eq = (L/dt) × I_new - (L/dt) × I_old
R_eq = L/dt

Trapezoidal Companion Model: 

V_eq = (2L/dt) × I_new - (2L/dt) × I_old + V_old
R_eq = 2L/dt

Parameters: | Parameter | Symbol | Unit | Range | Description | |-----------|--------|------|-------|-------------| | value | L | H | > 0 | Inductance | | ic | I₀ | A | any | Initial current |

Sources

DC Voltage Source

Symbol: V Equation: V = V_dc

Implemented using a voltage source MNA stamp with an additional equation row.

MNA Stamp: 

      n+    n-    I_v
n+  [  0    0    +1  ]
n-  [  0    0    -1  ]
I_v [ +1   -1     0  ] = V_dc

Pulse Voltage Source

Waveform Definition: 

        V2  ____________________
           /|                  |\
          / |                  | \
         /  |                  |  \
    V1 _/   |                  |   \_____
       |    |                  |   |
       td   tr      pw         tf
       |<-------- period ---------->|

Parameters: | Parameter | Symbol | Unit | Description | |-----------|--------|------|-------------| | v1 | V₁ | V | Low voltage level | | v2 | V₂ | V | High voltage level | | td | t_d | s | Delay time | | tr | t_r | s | Rise time | | tf | t_f | s | Fall time | | pw | t_pw | s | Pulse width | | period | T | s | Period |

Sinusoidal Voltage Source

Equation: 

V(t) = V_o + V_a × sin(2π × f × (t - t_d) + φ) × e^(-θ(t-t_d))

Parameters: | Parameter | Symbol | Unit | Description | |-----------|--------|------|-------------| | vo | V_o | V | DC offset | | va | V_a | V | Amplitude | | freq | f | Hz | Frequency | | td | t_d | s | Delay | | theta | θ | 1/s | Damping factor | | phi | φ | deg | Phase |

PWM Source

Generates a pulse-width modulated signal for switching control.

Parameters: | Parameter | Symbol | Unit | Description | |-----------|--------|------|-------------| | frequency | f_sw | Hz | Switching frequency | | duty | D | - | Duty cycle (0-1) | | phase | φ | deg | Phase offset | | dead_time | t_dead | s | Dead time between transitions | | inverted | - | bool | Invert output |

Dead Time Handling:

For complementary switches (e.g., high-side and low-side in half-bridge): 

S_high: |----ON----|     |----ON----|
S_low:       |----ON----|     |----ON----|
        ^dead^      ^dead^

Semiconductor Devices

Diode - Ideal Model

Symbol: D Model: ideal

Equations: 

OFF state (V_ak < 0):  I = 0, G = G_off = 1e-9
ON state (V_ak >= 0):  I = V_ak / R_on, G = 1/R_on

Parameters: | Parameter | Symbol | Unit | Default | Description | |-----------|--------|------|---------|-------------| | ron | R_on | Ω | 0.001 | On-state resistance | | roff | R_off | Ω | 1e9 | Off-state resistance |

Diode - Shockley Model

Symbol: D Model: shockley

DC Equation: 

I_d = I_s × (e^(V_d/(n×V_t)) - 1)

where V_t = kT/q ≈ 26mV at 300K

Junction Capacitance: 

C_j = C_jo / (1 - V_d/V_j)^m   for V_d < FC×V_j
C_j = C_jo × (1-FC)^(-1-m) × (1-FC(1+m) + m×V_d/V_j)   for V_d >= FC×V_j

Diffusion Capacitance: 

C_d = τ_t × dI_d/dV_d = τ_t × I_s/(n×V_t) × e^(V_d/(n×V_t))

Parameters: | Parameter | Symbol | Unit | Default | Description | |-----------|--------|------|---------|-------------| | is | I_s | A | 1e-14 | Saturation current | | n | n | - | 1.0 | Emission coefficient | | rs | R_s | Ω | 0 | Series resistance | | cjo | C_jo | F | 0 | Zero-bias junction capacitance | | vj | V_j | V | 0.7 | Junction potential | | m | m | - | 0.5 | Grading coefficient | | tt | τ_t | s | 0 | Transit time | | bv | BV | V | ∞ | Breakdown voltage | | ibv | I_BV | A | 1e-10 | Breakdown knee current | | fc | FC | - | 0.5 | Forward-bias capacitance coefficient |

Linearization for Newton-Raphson: 

I_d ≈ I_d0 + G_d × (V_d - V_d0)
G_d = dI_d/dV_d = I_s/(n×V_t) × e^(V_d0/(n×V_t))

MOSFET - Level 1 Model

Symbol: M (NMOS/PMOS)

Operating Regions:

Cutoff (V_gs < V_th): 

I_ds = 0

Linear/Triode (V_gs >= V_th, V_ds < V_gs - V_th): 

I_ds = K_p × (W/L) × [(V_gs - V_th) × V_ds - V_ds²/2] × (1 + λ × V_ds)

Saturation (V_gs >= V_th, V_ds >= V_gs - V_th): 

I_ds = (K_p/2) × (W/L) × (V_gs - V_th)² × (1 + λ × V_ds)

Parameters: | Parameter | Symbol | Unit | Default | Description | |-----------|--------|------|---------|-------------| | vth | V_th | V | 2.0 | Threshold voltage | | kp | K_p | A/V² | 20e-6 | Transconductance parameter | | lambda | λ | 1/V | 0.01 | Channel-length modulation | | w | W | m | 100e-6 | Channel width | | l | L | m | 10e-6 | Channel length |

Capacitances: | Parameter | Symbol | Unit | Description | |-----------|--------|------|-------------| | cgs | C_gs | F | Gate-source capacitance | | cgd | C_gd | F | Gate-drain (Miller) capacitance | | cds | C_ds | F | Drain-source capacitance |

Jacobian Elements for Newton-Raphson: 

∂I_ds/∂V_gs = g_m = K_p × (W/L) × V_ds                    (linear)
∂I_ds/∂V_gs = g_m = K_p × (W/L) × (V_gs - V_th)           (saturation)

∂I_ds/∂V_ds = g_ds = K_p × (W/L) × (V_gs - V_th - V_ds)   (linear)
∂I_ds/∂V_ds = g_ds = λ × I_ds                              (saturation)

Power MOSFET - Simplified Model

For power electronics simulation, a simplified model using datasheet parameters is more practical.

On-State: 

V_ds = I_ds × R_ds(on)(T_j)

Temperature Dependence: 

R_ds(on)(T_j) = R_ds(on)(25°C) × [1 + TC1×(T_j-25) + TC2×(T_j-25)²]

Typical TC values for Si MOSFETs: - TC1 ≈ 0.005 to 0.01 /°C - TC2 ≈ 0 (often neglected)

Body Diode:

The intrinsic body diode is modeled as a fast-recovery diode: 

I_d = I_s × (e^(V_d/(n×V_t)) - 1)

With reverse recovery: 

Q_rr = I_F × t_rr / 2
E_rr = Q_rr × V_R

Parameters: | Parameter | Symbol | Unit | Default | Description | |-----------|--------|------|---------|-------------| | rds_on | R_ds(on) | Ω | - | On-state drain-source resistance | | vth | V_th | V | 3.0 | Gate threshold voltage | | ciss | C_iss | F | - | Input capacitance (C_gs + C_gd) | | coss | C_oss | F | - | Output capacitance (C_gd + C_ds) | | crss | C_rss | F | - | Reverse transfer capacitance (C_gd) | | qg | Q_g | C | - | Total gate charge | | tc1 | TC1 | 1/°C | 0.007 | Temperature coefficient 1 | | tc2 | TC2 | 1/°C² | 0 | Temperature coefficient 2 |

IGBT - Simplified Model

On-State Voltage: 

V_ce(sat) = V_ce0 + R_ce × I_c

Parameters: | Parameter | Symbol | Unit | Description | |-----------|--------|------|-------------| | vce_sat | V_ce0 | V | On-state voltage at zero current | | rce | R_ce | Ω | On-state resistance | | vth | V_th | V | Gate threshold voltage | | td_on | t_d(on) | s | Turn-on delay time | | tr | t_r | s | Current rise time | | td_off | t_d(off) | s | Turn-off delay time | | tf | t_f | s | Current fall time | | cies | C_ies | F | Input capacitance | | coes | C_oes | F | Output capacitance | | cres | C_res | F | Reverse transfer capacitance |

Switching Devices

Ideal Switch

Model: 

OFF: R = R_off (very high, ~1e9 Ω)
ON:  R = R_on (very low, ~1 mΩ)

Control Logic: 

V_control > V_th → ON
V_control <= V_th → OFF

Event Handling:

When a switch changes state: 1. Event detected at crossing time 2. Simulation steps back to crossing 3. Matrix is reformulated 4. Integration restarts with new topology

Magnetic Components

Ideal Transformer

Equations: 

V_1 = n × V_2
I_1 = -I_2 / n

where n = N_1/N_2 (turns ratio)

MNA Implementation:

The ideal transformer is implemented using coupled voltage and current sources.

Transformer with Parasitics

Equivalent Circuit: 

    R_p    L_lk_p        n:1        L_lk_s    R_s
o---/\/\/---LLLL---o    | |    o---LLLL---/\/\/---o
                       | |
                    L_m ||
                       | |
o----------------------o    o---------------------o

Parameters: | Parameter | Symbol | Unit | Description | |-----------|--------|------|-------------| | turns_ratio | n | - | Primary to secondary turns ratio | | lm | L_m | H | Magnetizing inductance | | llk_pri | L_lk,p | H | Primary leakage inductance | | llk_sec | L_lk,s | H | Secondary leakage inductance | | rp | R_p | Ω | Primary winding resistance | | rs | R_s | Ω | Secondary winding resistance |

Magnetic Core Runtime Contract (MVP)

For saturable_inductor, coupled_inductor, and transformer, the backend accepts component.magnetic_core with model: saturation|hysteresis and optional core_loss_k / core_loss_alpha / core_loss_freq_coeff.

Current MVP runtime behavior:

  • Computes per-sample magnetic loss estimate:
  • P_core = core_loss_k * |i_equiv|^core_loss_alpha * (1 + core_loss_freq_coeff * |di_equiv/dt|)
  • Exports canonical virtual channel:
  • "<component>.core_loss" (W)
  • For model: hysteresis, exports deterministic memory-state channel:
  • "<component>.h_state" (bounded [-1, 1])
  • Exports channel metadata:
  • domain="loss"
  • unit="W"
  • source_component="<component_name>"
  • Optional policy:
  • loss_policy: loss_summary appends deterministic loss-summary row "<component>.core" when global loss pipeline is enabled.
  • When thermal is enabled together with losses, magnetic core-loss rows are coupled into thermal telemetry:
  • thermal channel T(<component>.core)
  • thermal summary row <component>.core
  • component_electrothermal row consistency against the thermal channel.

Current limitations:

  • Magnetic thermal coupling currently uses simulation defaults (simulation.thermal.default_rth/default_cth) for magnetic virtual rows.
  • Advanced hysteresis family parameter sets (beyond current bounded state model) are not implemented yet.

Thermal Models

Runtime Thermal RC Model

The current runtime electrothermal path uses a per-device first-order RC model:

P_loss --> [R_th, C_th] --> T_j
              |
           T_ambient

Thermal parameters are configured per component using:

Parameter Symbol Unit Description
rth R_th K/W Thermal resistance (junction-to-ambient equivalent)
cth C_th J/K Thermal capacitance
temp_init T_0 degC Initial device temperature
temp_ref T_ref degC Reference temperature for scaling
alpha α 1/degC Electrical temperature scaling coefficient

Global defaults and ambient are configured in simulation.thermal.

Thermal-port enablement (component.thermal.enabled: true) is supported for:

  • resistor
  • diode
  • mosfet
  • igbt
  • bjt_npn
  • bjt_pnp

Temperature-Dependent Parameters

MOSFET R_ds(on): 

R_ds(on)(T) = R_ds(on)(25°C) × (T/300)^α

where α ≈ 2.0 to 2.5 for silicon

Diode Forward Voltage: 

V_f(T) = V_f(25°C) - α_vf × (T - 25)

where α_vf ≈ 2 mV/°C

Device Library

Pulsim includes a library of pre-defined device models based on manufacturer datasheets.

MOSFETs

Model V_ds R_ds(on) I_d Package
IRF540N 100V 44mΩ 33A TO-220
IRFP460 500V 270mΩ 20A TO-247
IPP200N25N3 250V 20mΩ 64A TO-220
STW48NM60N 600V 70mΩ 44A TO-247

IGBTs

Model V_ce V_ce(sat) I_c Package
IRG4PH50UD 1200V 2.0V 45A TO-247
IKW40N120H3 1200V 1.7V 40A TO-247

Diodes

Model V_r V_f I_f t_rr
1N4148 100V 0.7V 200mA 4ns
MUR860 600V 1.0V 8A 60ns
STTH30R06 600V 0.95V 30A 35ns

Using Library Models

- type: mosfet
  name: Q1
  nodes: [d, g, s, 0]
  use: IRF540N

Or override specific parameters:

- type: mosfet
  name: Q1
  nodes: [d, g, s, 0]
  use: IRF540N
  rds_on: 0.05

Model Selection Guidelines

For Fast Switching Analysis

  • Use simplified models with datasheet parameters
  • Include gate charge model for timing
  • Enable adaptive timestep near switching events

For Loss Calculation

  • Use models with accurate on-state characteristics
  • Include temperature dependence for thermal coupling
  • Use switching energy lookup tables

For EMI Analysis

  • Include all parasitic capacitances
  • Use short timesteps (< 1/10 of rise time)
  • Model PCB parasitics as lumped elements

For Thermal Analysis

  • Use Foster models with 3-4 time constants
  • Enable bidirectional thermal-electrical coupling
  • Run for sufficient time to reach thermal steady-state