Post History
This depends, of course, on what you are trying to achieve with the model. Remember that a model is a simplification for purpose of analysis, because the full physics is either too complicated or ...
Answer
#1: Initial revision
This depends, of course, on what you are trying to achieve with the model. Remember that a model is a simplification for purpose of analysis, because the full physics is either too complicated or contains too many unknowns. By necessity, some real-world details are always left out of any model. What you are really asking is how much detail is required. Again, that obviously depends on the circumstance and the reason for using the model. For example, if you just want to make sure the correct bits are going back and forth in a IIC bus message, modeling each line as a digital signal is fine. In fact, including more details would make the model less useful since they only add extra work without further insight. On the other hand, if you want to simulate the timing of each line, then including the bus capacitance and pullup resistors is necessary. Signals are no longer digital, but exponential decays towards their steady-state values. But the exponential decay model is just another simplification. It doesn't take into account transmission line effects, how those vary by frequency, effect of a ground plane, or not, propagation delay, etc. If you are getting into this detail for IIC bus signals, then you are way too close to margins and are pushing limits that may well be passed on some fraction of real units. However, this level of detail is necessary for some high speed digital signals, like accessing modern high speed memories.