The simulation did not converge
The solver could not settle on a starting point or could not step through time. What causes each, and what to change.
Britt Espinosa
Last Update 15 days ago
Convergence failures show up as one of two titles in the Status tab of the Info panel, depending on whether the solver failed at the starting point or partway through a transient run.
The messages you may see
Under the title The circuit has no stable starting state:
The simulator could not find a consistent starting voltage and current for every node.
Under the title The circuit changes too abruptly to calculate:
The simulator was forced into steps that were too small to finish this transient analysis.
Technical details for either one usually mention convergence, gmin stepping, source stepping, an iteration limit, or a timestep that became too small.
No stable starting state
Every analysis begins by solving for the operating point, so this failure blocks even a transient run. The product's steps are:
- Check component values, polarities, and power connections.
- Make sure every analog section has a path to Ground.
- For a transient run, try enabling "Use initial conditions" only when the circuit has intentional startup behavior.
Work through them in that order. Reversed electrolytics, a diode fed the wrong way, a supply wired to the wrong rail, and a resistor entered as 1 instead of 1k are the values that most often make the operating point unsolvable. The path to ground check is worth doing even when no ground error appeared, since an isolated section can still be reachable through a capacitor that blocks DC. See No ground in the circuit.
The third step names "Use initial conditions". The control that does this sits under Advanced in Simulation Settings and is labeled Enable UIC. Turn it on when your circuit has deliberate startup behavior, such as a precharged capacitor or an oscillator that needs a kick. Leave it off otherwise, because it tells the solver to skip the operating point rather than solve it.
Steps that became too small
This one appears after the run has started. The steps are:
- Check ideal switches, diodes, and sources for instantaneous changes.
- Add realistic resistance or transition time where appropriate.
- Try a smaller maximum timestep only after checking the circuit connections.
An ideal switch closing in zero time, a pulse source with no rise time, and a diode with no series resistance all ask the solver to follow an edge with infinite slope. Give the source a rise and fall time, add a few milliohms in series with switches, and add the small parasitic resistance a real part would have.
The issue carries a Simulation settings action that opens the panel directly, which is where the maximum timestep and the transient end time live.
If neither change helped
Cut the circuit down. Run the smallest section that includes the suspect part, confirm it solves, then add the rest back one stage at a time. When you find the stage that breaks it, copy the technical details from the issue in the Status tab and send them to support with the circuit name.