In the Plasma, Time Periodic interface it is possible to sustain a plasma using power and voltage sources. In both cases a periodic electric potential is applied at a terminal. If a power source is chosen the voltage amplitude of the periodic signal is numerical found to obtain a given power absorbed by the plasma. As an example, when using the Metal Contact feature to drive a discharge with a fixed power the following expression and constraint are used:

The constraint on the power Prf is used to compute the voltage amplitude Va such that a fixed amount of power is absorbed by the plasma. In most cases it is numerically more stable to drive a plasma with a power source.

Solving a problem directly for a high power can be numerical difficult. It is strongly recommended to start a simulation with a very low power (a few watts for an industrial reactor) and increase the power to the desired value using the Auxiliary sweep in the Study Extensions section in the Study node. This strategy is used in several models in the application library. In an initial phase of the implementation of a model, a two-study version of the power sweep can be used: first, a study is run at low power, after, its solution is used as initial condition in another study that is used to increase the power. This approach saves the time of running the low power step while polishing the model.

When using the Auxiliary sweep the default option for the continuation solver is Last parameter. When using this option the solver tries to predict the solution for the next parameter. If power is used as a parameter and with a slow increase in the power (a few watts) it is possible to obtain solutions in just a few iterations. However, in some cases the variation in the parameter needs to be unreasonably small in order for the continuation solver to work well. In those cases it is suggested to use the No parameter option, and choose the option Yes in Reuse solution from previous step. When doing so, the solver does not use a prediction for the next step but it still uses the solution from the previous step as initial condition. With this option it is possible to increase the power with much larger steps, but more iterations are needed.