Home Forums Bug Reports Convergence Problems with Mosfet models

Viewing 2 posts - 1 through 2 (of 2 total)
  • Author
  • #667

    I was wanting to simulate a phase-frequency-detector circuit that uses N-MOS and P-MOS, but ran into a brick-wall with the simulator giving minimum-time-step errors.  I think this might be a convergence problem.  This occurs even on the simplest NMOS inverter (one-transistor) circuit making the mosfet circuit modeling unusable.  Any workarounds or suggestions?  Attached is an example schematic.  I tried changing some of the mosfet model parameters (Used the Typical Values from Semiconductor Spice Modeling with SPICE textbook) but I could not get anything to work for a transient simulation.

    Jess Stuart

    You must be logged in to view attached files.

    In your MOSFET model, the parameter “Ld” is much too big. If it’s greater then half the channel length “L”, the effective length becomes negative, which is, of course, unphysical.

    Anyway, the problem here is that the trapezoidal method can’t handle steps quite well. Please choose the integration method GEAR3. I hope that works better.

Viewing 2 posts - 1 through 2 (of 2 total)
  • You must be logged in to reply to this topic.