Re: [ng-spice] Switch problems / solutions . . .


To ng-spice@ieee.ing.uniroma1.it
From Al Davis <aldavis@ieee.org>
Date Fri, 27 Oct 2000 20:17:10 -0700
Delivered-To mailing list ng-spice@ieee.ing.uniroma1.it
In-Reply-To <JOJMALEAILMIEAAA@hotbot.com >
Mailing-List contact ng-spice-help@ieee.ing.uniroma1.it; run by ezmlm
References <JOJMALEAILMIEAAA@hotbot.com >
Reply-To ng-spice@ieee.ing.uniroma1.it

On Fri, 27 Oct 2000, GLAO S. Dezai wrote:
> Could we use  XSPICE event-driven transient simulation core ?

How about the ACS event-driven transient simulation core?

In ACS, all devices have a "review" function that is called after
convergence, on deciding whether to accept a time step.  It returns a
suggestion for a preferred value for the next step, and can force an
exact time in the event queue.


Back to the switch .......

ACS has a voltage controlled resistor, which can have a nonlinear or
time dependent transfer characteristic.  I think this is a better
way.  An ordinary resistor can be time dependent, or nonlinear using
source-like functions.

A future version will be able to make components nonlinear AND time
dependent, using 2 dimensional tables, and there will be a "trigger"
element to get it going.  This is a very high priority because I need
it for IBIS support.


Partial thread listing: