Geoffrey_Coram
|
I haven't seen an example using $temperature, but I don't see any reason that that wouldn't be allowed. The original paramset proposal had the idea that one might pick a model based on a design switch -- eg, if you are doing digital or analog simulation of the block.
Usually, though, the model has its own temperature-update equations, and one specifies parameter values for the nominal temperature (tnom or tref). Sometimes, there will be subcircuit equations involving temperature. I wouldn't expect to need a paramset -- unless you're trying to pick one model for room-temperature operation, and a completely different model for super-cold (superconducting?) or super-hot (industrial) temperatures.
|