Tons of convergence errors

QSPICE
9

I did send a couple bug-reports (with examples attached) to the official email under help::about

  1. QSpice sometimes finds the wrong operating point in TI models, .ns V(out)=0 can help
  2. QSpice often slows or stops with small timesteps when using TI models, setting a shorter max-timestep in tran often helps
    finds_unstable.qsch (13.3 KB)
    switches3.qsch (10.6 KB)
    switches2.qsch (10.1 KB)

I’m not sure, though, if it is feasible for QSpice to keep compatibility with TI’s PSpice at that level; I suspect the TI engineers will keep making more complicated models to match the details of their products, so long as those models work for them in PSpice.