Re: snapshot/resimulate and hspice
> well, in my experience, the number of iterations necessary
> at each time point for time-point convergence is a function
> of the circuit activity at each time point, and this effects
> the total cpu time required at each time point. otherwise,
> each time point in a transient analysis would cost as much
> cpu time as dc convergence.
Yeah, that's why I say it helps, but not as much as I think
Al reckons ACS should do in the same circumstances.
> for transient analysis, commercial simulators like hspice
> speed up significantly when a circuit is stable, and slow
> down again when activity starts up again. spice3 has this
> behaviour as well, and infact, when it simulates something
> correctly, usually is faster than hspice...
Interesting that you're using hspice. The reason I got into
this lark is 'cos I couldn't stand hspice. Spice3f4 seemed
to work much better if it worked, so I set about trying to
make it work in most cases. I ended up using my own version
of spice3f4 instead of hspice most of the time.
Have you tried ngspice yet ? I think rework 13 has got a few
obvious bugs, but rework 14 should be out soon. I have my own
version of spice3f4 which has a mos level 9 which is mos level
3 with some hspice parameters added. It's set up for win32,
rather than Linux, and it outputs pspice style .dat files.
I can it email to you if you're interested, but you need ESP
to know what parameters are included and how to use it ;-)
Paolo, is rework 14 coming soon ? I think I sent you my stuff
with level 9 MOS and level 2 BJT (dc current in the substrate
diode). Did that ever get in to ng-spice ?
Cheers,
Alan
Partial thread listing: