AC Simulation Description

When an AC small-signal simulation is run, the system first computes the DC operating point of the circuit. Whenever a linear simulation such as a linear AC simulation requires a single-point DC bias simulation to be run first, it is referred to as a bias-dependent linear simulation. The most common example is the case of a linear amplifier that uses a biased transistor as the active element. The DC bias simulation is executed automatically and transparently (unless an error causes the DC simulation to fail to converge).

Following the DC bias simulation, the simulator linearizes all nonlinear devices about their bias points. A linearized model captures the small incremental changes of current due to small incremental changes of voltage. These are the derivatives of the transistor model equations, which are evaluated at the DC bias point. Nonlinear resistors and current sources are replaced by linear resistors whose values are set by the small signal conductance dI/dV. Current sources that depend on voltages other than the voltage across the source are replaced by linear dependent current sources dI ₁/dV ₂. Nonlinear capacitors are replaced by linear capacitors of value dQ/dV.

The resulting linear circuit is then simulated over the specified frequency range. Small-signal AC simulation is also performed before a harmonic-balance (spectral) simulation to generate an initial guess at the final solution.

Use the AC controller to:

• Perform a swept-frequency or swept-variable small-signal linear AC simulation.
• Obtain small-signal transfer parameters, such as voltage gain, current gain, transimpedance, transadmittance, and linear noise.

Simulation can be performed repeatedly while sweeping some parameter. If changing these parameters affects the DC operating point, the DC operating point and linearized circuit will be recomputed at each step.

Note If the circuit has only one AC source, it is often convenient to set its magnitude to one and its phase to zero. In this way, the small-signal transfer function is computed directly.

Enabling Frequency Conversion

Traditional small-signal AC analysis is truly linear in the sense that frequency conversion effects do not occur. In RF system simulation, however, it is common to have frequency-translating mixer components that have approximately linear RF-to-IF conversion characteristics under small-signal RF drive. By enabling frequency conversion (also known as FCAC), you can perform system-level small-signal analyses on such systems. As is the case in standard AC analysis, a noise option is also available.

At the beginning of each FCAC simulation, a so-called frequency map is established. This map specifies the frequencies present at the various circuit nodes, and is based on the frequencies of the sources and the types of behavioral mixer components present in the network. Each node in the network can have only one frequency associated with it. Consequently, each behavioral mixer component can model frequency conversion to either the upper or the lower sideband, but not to both simultaneously.

Sources most often used for FCAC analysis include the V_1Tone, I_1Tone, and P_1Tone components. The frequency used by the source is given by the Freq parameter. If a multitone source is used, the frequency is specified by the Freq[1] parameter. When no frequency is explicitly specified, voltage and current sources default to the global value of the freq variable, while port sources simply become passive. Small-signal amplitudes used for FCAC analysis are given by Vac, Iac, and Pac parameters for voltage sources, current source, and ports, respectively.

Note It is not possible to use FCAC analysis accurately with user-constructed circuit-level mixer blocks (such as diode mixers, Gilbert cell mixers, and the like).