On Fri, 22 Jul 2016 17:07:51 -0500 John Griessen <[email protected]> wrote: > and gnucap says, "non-recoverable convergence failure, reducing (itl4) > newtime=3.300000e-05 rejectedtime=3.300000e-05 > oldtime=3.300000e-05 using=3.300000e-05 tried everything, still > doesn't work, giving up"
What version of gnucap are you using? In older versions, there used to be a bug that sometimes messed up time stepping, causing duplicate time steps, rejecting a step that was actually good and repeating exactly the same time. It had something to do with floating point truncation. I thought it was fixed a few years ago. If it's a recent version, I want to see your circuit. On Fri, 22 Jul 2016 17:07:51 -0500 John Griessen <[email protected]> wrote: > .model 1N4007 D(IS=7.02767n RS=0.0341512 N=1.80803 EG=1.05743 XTI=5 > BV=1000 IBV=5e-08 CJO=1e-11 VJ=0.7 M=0.5 FC=0.5 TT=1e-07 mfg=OnSemi > type=silicon) Some of the parameters here are proprietary extensions in some versions of Spice. As you figured out .. Forward voltage drop of a diode is determined mainly by the parameters IS, N, and RS. If you are fitting it expermentally, and have two data points, choose IS to match the forward voltage at a low current (where the diode exponential is more pure), then RS to match at a high current (where ohmic resistance dominates). As Felix said, the "native" diode model doesn't implement breakdown. There are several alternatives available as plugins in git://git.sv.gnu.org/gnucap/gnucap-models.git These models are what they are, which may be better or worse than the "native" "modelgen" model. _______________________________________________ Help-gnucap mailing list [email protected] https://lists.gnu.org/mailman/listinfo/help-gnucap
