Re: Verilog-AMS Committee Meeting Minutes - Oct 5 2006

Marq -
I haven't thought too hard about this, but my understanding
goes like this.  Suppose we have

  I(out) <+ slew(V(in), +1, -1);

During a transient analysis, when V(in) is changing rapidly
and hence the slew function is active, the matrix load for
the (out,in) element will be zero: the output current at
the current timepoint does not depend on the input voltage
at the current timepoint.  (The current depends on the
voltage at the previous timepoint and the timestep.)

When the slew function is not active, the (out,in) matrix
element will be 1.0.

-Geoffrey


Marq Kole wrote:
> 
> Ken, Geoffrey,
> 
> Could you please explain to me what is meant with the current (LRM 2.2) wording?
> 
> If we restrict ourselves to AC small-signal, potentially from an operating point captured from a transient, then moving from a point where the slew function is not slewing to a point where it starts to slew will introduce a discontinuity in the AC behaviour -- instead of the unity transfer function it will suddenly become a zero transfer function. That is, if my interpretation of "zero transmission" is correct.
> 
> I would expect the AC behaviour to be less discontinuous -- more of a low-pass behaviour with a time constant proportional to the slew-rate.
> 
> Cheers,
> Marq