The error can be solved by decreasing the minimum timestep (default: 1e-12) by setting dtmin option:
.option dtmin=1e-13 Still, simulation using backward Euler method gives better results and runs faster due to the lack of ringing. Is there any method to "stabilize" the simulation using trapezoid method? -r. On Nov 30, 2007 2:03 AM, Leandro Marsó <[EMAIL PROTECTED]> wrote: > 2007/11/29, a r <[EMAIL PROTECTED]>: > > When using a TSMC 0.25 MOSIS models (level=49) setting the time step > > below 0.1ns like below: > > > > .tran 0.01n 40n > > > > causes following problems. > > > > storage element step control error:Cgs.Mn1.xgate 1e-12 > > using Euler, disabling time step control > > > > With timestep of 0.1ns the simulation runs fine but the numerical > > ringing is very visible. > > > > Any idea? > > > > -r. > > Hello -r, > be carefull with the units in the length and width of the channel, it > is in meter units. If it is not the problem, you could try to isolate > the problem with a small circuit and paste it and send it to the list. > good luck! > Leandro > > > _______________________________________________ > Help-gnucap mailing list > [email protected] > http://lists.gnu.org/mailman/listinfo/help-gnucap > _______________________________________________ Help-gnucap mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-gnucap
