top of page
Picotest

SPICE models need correlation to measurements

Many SPICE models are available online that were created by IC and semiconductor manufacturers. Unfortunately, availability doesn’t necessarily equal usability. The majority of these models have no documentation, or detailed description of the model’s capabilities. More importantly, the limitations aren’t discussed, at least not in detail.

Device manufacturers don’t explain the structures of the macro-modeled subcircuits and they don’t provide a roadmap about what behavior each element controls. Rarely is there data correlating the model performance to bench measurements or datasheet performance. This can leave you wondering what the model is and isn’t good for and under which operating conditions is it valid.


There is so much at stake and yet, despite the lack of model correlation, these models are often presumed to be correct and used without validation because they are provided “by the manufacturer.” Some simple simulations reveal this to be a dangerous assumption.

Because the results of a simulation are totally dependent on each and every SPICE model employed, you must consider two fundamental concerns, both of which apply to the part models and to the circuit model as a whole. First, does the SPICE model exhibit the characteristics needed for the analysis being performed? For example, if you’re performing a stability analysis on a power supply circuit, does the output impedance of the voltage regulator vary appropriately? Second, how accurately does the characteristic match the part’s or circuit’s performance over the entire operating load current and input voltage range? For example, while the regulator’s output impedance might vary over frequency, is it within initial tolerances?

11 views0 comments

Recent Posts

See All

Comments


bottom of page