Draw a circuit,
Play through it. Live.

Ideal classic resistor element that is used in every circuit. Fully linear.

Ideal capacitor, the circuit emulates them by integrating with the trapezoidal rule at the sample period.

A reverse-biased electrolytic is reported, not simulated: the builder flags it as an error in the console rather than quietly modelling a part that would be failing on the bench.

Ideal capacitor, the circuit emulates them by integrating with the trapezoidal rule at the sample period.

The Ideal transformer models the turns ratio and nothing else: no bandwidth limit, no losses, and it passes DC.

The Real transformer has magnetising inductance, leakage and winding resistance, all four numbers derived from the turns ratio against an assumed 8 Ω secondary. These simulations are tailored after guitar-specific transformers and emulate the overall functionality without being overly faithful to a particular model.

Neither model emulates core saturation and stays perfectly linear (this is a current weakness of the modelling).

There is also a centre-tapped version, useful for push-pull output stages.

A perfect potentiometer. It has three tapers : linear, logarithmic and reverse logarithmic. Every pot you draw becomes a control in the strip and a parameter the host can automate.

They can be tweaked live, just like a real amplifier, with the knobs contained in the lower strip at the bottom of the plugin.

Two potentiometers with the same label will become ganged. Each potentiometers keeps its own resistance and its own taper, only the shaft is shared. The wiper is drawn on the sheet live, making it easy to see the impact it has on the circuit.

It is perfectly modelled : no wiper resistance and no taper tolerance.

Switches function in the same fashion as the potentiometers. It is implemented as a resistor that changes value (it is not a true break in the circuit). It has the same naming convention : two switches with the same name will be shared and ganged together.

The voltage source is ideal, it fixes a voltage and lets the circuit decide over the current. Supply sag is modelled.

All of the modelled diodes share the same p-n junction maths. It is an exponential I-V curve with maths inspired bt SPICE.

No junction capacitance and no reverse recovery. This is negligible at audio-rate clipping but slightly wrong in a power supply rectifier.

The Early effect and the junction capacitances are modelled for the parts where these values were available. These settings can be disabled in the settings (and they will slightly lower the CPU load).

Transit time is not modelled : f_T does not degrade with current.

JFE transistors. The square law is a simple one. There are no capacitances, no Miller like the BJTs.

Each Op-amp in the editor is assembled from a resistor, two controlled sources, a capacitor and two diodes. The gain stays linear and only the clamp reaches the nonlinear solve. This is a limitation that makes the op-amp affordable for realtime rendering.

What is modelled : finite open-loop gain, one dominant pole, and the rails it clips against. What is not: slew limiting, input bias and offset current, CMRR, PSRR and the higher-order poles a real part has.

Every valve implemented to the editor is fitted to a manufacturer's datasheet. It reproduces plate current, transconductance and plate resistance at the quoted operating point in the datasheet.

There are two modelling techniques used by the engine, Koren and D-Z. Koren is the classic formulation and fits published curves. It is the work of Norman Koren and the engine uses the equations he derived. Dempwolf–Zölzer is fitted to analysis and measurements of three individual 12AX7s from a published paper by Dempwolf and Zölzer.

Interelectrode capacitance is an enhancement that tries to make valve emulation closer to reality. It is slightly expensive, can cause problems at time, and can be disabled if it causes problems or to save CPU.

The pentodes use Koren's equations. The suppressor grid is permanently connected to the cathode.

There are two limitations to their emulation : Screen current does not depend on plate voltage: in a real valve, when the plate collapses into clipping the screen takes the current the plate stops taking, and here it carries on regardless. There are no secondary emission, the kink in the plate curves at low plate voltage where electrons flow backwards.

This model is pure 3/2-power space-charge flow, fitted to real datasheets.