Monte-Carlo Noise in the Time Domain

Noise in transient simulations can be added as pseudo-random voltages and currents at each timepoint. The noise has a Gaussian amplitude distribution. The noise can be frequency- or bias-dependent, as appropriate for each component. All components that generate noise in a linear or harmonic balance simulation generate the appropriate noise, with the following exceptions:

• S-parameter data from files that include a noise parameter block with NFmin, Rn and Sopt
• Behavioral components for amplifiers and mixers
• Noisy2Port component
• NoiseCorr noise correlation component
• SDD
• Mextram504 and Hicum transistor models

The noise is added to the signal, so there is no easy way to separate the signal from the noise. The full, nonlinear circuit equations are applied to this composite signal of random voltages and currents, so no small-signal assumptions about the relative size of the noise are required.

The Noise bandwidth parameter (under the Transient Noise tab) enables the generation of noise in transient analysis. Because transient analysis uses a variable timestep method, the noise must be limited to a bandwidth of less than 1/(2*MaxTimeStep). If this parameter is left blank or specified as zero, no noise is generated.

The Noise scale parameter is used to multiplicatively scale all of the noise sources in the circuit. This can be useful when the noise levels are very low, as this allows them to be increased so they are visible above the numerical noise in the simulation. However, if the noise is increased too much, it can change the nonlinear operation of the circuit.

The noise is generated by a random number generator. It will produce a different sequence of random numbers each time the simulation is run. If a repeatable sequence is required, it can be obtained by setting the simulator variable __randseed to an integer value with a schematic equation. For example,

__randseed=12345 (two underscores precede randseed)