Re: About WARNING of ibischk2

Syed,

Syed Huq wrote:
> What I would like to know is since this is a WARNING and not an
> ERROR, how concerned does the model developer need to be ??

When the warning is issued the model developer should determine the
cause of the warning as you have done (i.e. small discrepency due to
measurement technique) and then move on. This is just a warning.

EDA tools:

Viewlogic/Quad tools will use the VI curves as the basis for simulation
with the V/T waveforms as additional information so it turns out that a
small errors like 5% on waveform endpoints will **potentially** only
change simulation results on the order of uVolts (or not at all!) (DC
stable states remain same regardless of V/T waveforms). I think most
EDA tools will operate in a similar manner.

The point of the warning is to let model developer know that something
major may be wrong like:

    They used a TYP VI curve with a MAX VT waveform
    They forgot to include a high curve!!
    They forgot to include a low curve!!
    They used a different load resistor than specified.
    The AC waveform wasn't done stabilizing at end of simulation.
    ...etc...

Any of the above would cause significant error so want to avoid.

Chris Rokusek
Viewlogic.
Received on Mon Apr 6 11:51:28 1998