<Task Type> Properties Specials <Task Name> Properties Specials
In the Solver Specials dialog, some special parameters can be defined, that affect the performance of the circuit simulation engine. Often, these parameters should be left as they are since they are initialized CST Design Studio in a way that proved useful for most circuits. However, if convergence, accuracy, or speed is not as good as expected, the user might fine-tune these parameters to improve the overall performance.
The following frames are available for all kinds of circuit simulation tasks (S-Parameters, Operating point (DC), AC, Transient, Spectral lines, Mixer, and Amplifier):
Parallelization frame
Contains settings that control the degree of parallelization within the circuit simulation task.
Parallelization: The user can select between three settings:
1. Maximum: The number of physical CPU cores is used for parallelization, if they are available.
2. Up to: The user can specify, how many physical CPU cores are to be spent on parallelization.
3. None: No parallelization is applied. Only one physical CPU core is used.
Nonlinear solver frame
Contains settings that control the nonlinear iteration in the solution process of (nonlinear) circuits.
Tolerance: Specifies the maximum accepted error current. It is defined as the sum of all currents to any particular node during the Modified Nodal Admittance (MNA) Analysis.
Radio buttons: By these, the user can select between absolute (measured in terms of current units) and relative (in relation to the current impressed by external sources) tolerances.
Max. number of Iterations: It specifies, after how many iterations the nonlinear solver terminates unsuccessfully, when the desired tolerance on the error current is not reached. As a general rule, highly nonlinear and large circuits tend to require more iterations.
Auto: By activating this check box, the simulator applies appropriate settings for the nonlinear solver. Other entries within the frame are ignored.
Circuit condition frame
Contains settings by which the condition of the circuit equations might be improved, especially, if there are topological problems.
Minimum conductance: It is used in a twofold way. Firstly, minimum conductances (also called GMIN) are spanned to ground, from nodes, that have been identified as being isolated. This prevents the circuit equations from becoming numerically singular. Secondly, ideal current sources are given an inner conductance of GMIN.
Minimum resistance: It is used in a twofold way. Firstly, when resistive circuit elements have a resistance smaller than the minimum resistance (also called RMIN) they are treated internally as nullors to avoid very large entries in the MNA matrix, which might destroy the condition of the circuit equations. Secondly, ideal voltage sources are given an inner resistance of RMIN.
Auto: By activating this check box, the simulator applies appropriate settings for GMIN and RMIN. Other entries within the frame are ignored.
The following frames do only appear for Transient simulation tasks:
Time steps frame
Contains settings, by which the number and size of time steps during transient simulations can be controlled.
Time stepping: It specifies the mechanism which is used to control the time steps. If it is on 'Automatic', the simulator starts with an adaptive scheme with an appropriate range of time steps. On non-convergence, the range of time steps is adjusted. If this proves unsuccessful, a fixed time stepping scheme is applied. If the setting is on 'Adaptive', time steps are varied within limits to achieve a sufficiently accurate solution. If the setting is on 'Fixed', fixed time steps are used throughout transient simulation.
Time step: It specifies the time step used in transient simulation. It can only be modified for non-automatic time stepping. For adaptive time stepping, the user specifies a maximum time step (the minimum time step chosen appropriately by the simulator).
Excitation frame
Contains settings, by which the frequency and time resolution of dispersive blocks in the circuit can be specified. On automatic time stepping, the range of time steps is also influenced by this setting.
Maximum frequency estimator: It characterizes the method, by which the maximum frequency to which dispersive circuit elements are evaluated, is determined. Generally, the spectrum of dispersive circuit elements is truncated beyond a maximum frequency fmax, that is calculated from all excitation signals. For the 'Energy-Based' estimator, excitation signals are truncated in frequency such, that the energy of the truncated signal in relation to the original signal is larger than a given threshold over a period. For the 'Transitiontime-Based' estimator, the calculated fmax is related to the smallest rise- or fall time of all excitations. The 'Automatic' estimator chooses an appropriate fmax , depending on the types of excitations that are present in the transient task. Finally, the user can specify fmax explicitly by selecting the 'Manual Fmax' setting.
Energy threshold: It can only be specified, if the 'Energy-Based' estimator has been enabled. Excitation signals are truncated in frequency such, that the energy of the truncated signal in relation to the original signal is larger than a given energy threshold over a period.
Maximum resolved frequency: It can only be specified, if the 'Manual Fmax' estimator has been enabled. It directly specifies the maximum frequency fmax at which the spectra of dispersive circuit elements are truncated.
Time integration frame
Contains settings, by which time integration method and the associated solution accuracy in transient simulation can be specified.
Applied Method: Contains a choice between three different time integration methods: Gear (2nd order) , Adams (2nd order) and Backward Euler (1st order). The 1st order method is not running in combination with adaptive time stepping.
Relative accuracy: Defines the maximal tolerable local truncation error of the time integration method. If the estimated local truncation error of the solution at a new time step is above the chosen accuracy limit, the solution is rejected and the time integration step is repeated with a reduced time step. As a consequence this accuracy setting may have a great influence on the overall simulation time. Note that this setting does only apply if an adaptive time integration scheme is employed.
Miscellaneous frame
Contains settings that are not directly related to the transient simulation but to postprocessing of the transient simulation results.
Number of FD samples: Defines the number of samples to be used to represent all calculated frequency domain data of this task.
Ok
Stores the current settings and leaves the dialog box.
Cancel
Closes this dialog box without performing any further action.
Help
Shows this help text.
See also