GetConvergenceInfo Function

Return convergence data for most recent simulation

Arguments

No arguments

Returns

Return type: string array

Returns a string array providing convergence information about the most recent run. Each element of the array is a list of values separated by semi-colons. The output may be pasted into a spreadsheet program that has been set up to interpret a semicolon as a column separator. The first element of the array lists the names for each column and therefore provides a heading. The following headings are currently in use:
type Node or Device
name Name of node or device that failed to converge count Number of times node/device failed to converge during run
time (first step) Time of most recent occurrence of a 'first step' failure.
required tol Required tolerance for most recent 'first step' failure
actual tol Tolerance actually achieved for most recent 'first step' failure
absolute val Absolute value for most recent 'first step' failure
time (cut back step) Time of most recent occurrence of a 'cut back step' failure.
required tol Required tolerance for most recent 'cut back step' failure
actual tol Tolerance actually achieved for most recent 'cut back step' failure
absolute val Absolute value for most recent 'cut back step' failure
final? Node or device failed on the final step that caused the simulation to abort
top analysis Main analysis mode (Tran, DC etc.)
current analysis Current analysis. Either the same as 'top analysis' or Op
op mode Method being used for operating point. (PTA, JI2, GMIN or SOURCE)
A first step failure is a failure that occurred at the first attempt at a time step after a previously successful step. If a time point fails, the time step is cut back and further iterations are made. Failures on steps that have been cut back are referred to in the above table as cut back steps . Quite often the nodes or devices that fail on a cut back step are quite different from the nodes or devices that fail on a first step. The root cause of a convergence failure will usually be at the nodes or devices that fail on a first step .

It is quite difficult to interpret the information provided by this function. The 'where' script performs a simple analysis and sometimes displays the nodes or devices most likely to be the cause.