BIRD 12, non-linear ramps, response to Jon from Kellee

          Kellee and others,

          The intent of the Bird is to present ramps that deviate
          significantly from the straight line currently supported in
          V1.1, such as those exhibited by controlled slew rate
          devices. The bonus that Jon pointed out is that the ramp
          curve itself provides a way of verifying the simulation,
          simply by simulating the model into the load used to derive
          the ramp curve in the first place. Additional ramps could
          be added to provide golden waveforms into other loads, which
          can be specified using the other elements in the Bird.

          By the way, I strongly hope to get this Bird, as modified by
          Forum discussion, into V2.0, since it overcomes the one
          glaring artificial linear element in the original spec. The
          strength of IBIS is that it allows second order effects to
          be simulated, but the single-value ramp rate limits its
          ability to work with a whole class of devices we are
          currently struggling to model.

          Will

From: Kellee Crisafulli
To: Jon and IBISians

From: Jon
As far as different simulators getting different results.
The correct answer is the input data itself (which is the cool thing
about this approach). If you don't get that you are wrong.
Getting close to that is the old competition game.
jonp

I think I didn't present my point well enough. I understand that
the data would be used to validate the model for a given load.
I was hoping that the data would also be used to produce simulation
results. It then becomes modeling information not just validation
information. And the question is does the model work for other load
conditions!

Is the intent to use this data only to validate models?

Have a great day... Kellee
Received on Wed Apr 27 23:27:02 1994