Okay, but who does that?
I can safely say I have never seen a circuit expose 1N914s to microwaves. And anyone who claimed to do so would be seen as retarded.
I see zero value in creating that model.
A much more useful situation is an ordinary power rectifier, where you hit it with a "fast ringy pulse" that's produced from a switching inverter. "Fast" in quotes because it's maybe tens of nanoseconds at the fastest.
In that regime, transport looks like series inductance, but it's nonlinear and non-conservative. An R||L can crudely replicate part of the behavior (like voltage overshoot), but doesn't help with reverse recovery and losses.
There are millions of inverters out there. This is a high value target. It would be fantastic to have an accurate measure of power dissipation, and dynamics, in a simulation of a real, practical circuit.
In the extreme case, one could simulate the electromagnetic and thermal response of the entire device, before ever ordering a prototype.
That's a model that has real value.
Phil, are you not motivated by value? Surely there's nothing worse than cooking up some (academically or computationally) amazing toy, and having it go nowhere because it's some obscure, useless bit of math that no one needs.
It sounds like you /are/ motivated by value, because accurate PD models would be directly useful to you, and of competitive advantage if you had them.
So I don't get your obsession with analytical accuracy. You're being a useless dick about it. If all your PhD has to say is "can't be done", then would you please step out of this conversation while the engineers discuss the "well what if..."?
Please do correct me if I'm wrong about any of these things: the value of an analytical 1N914 model, or a rectifier, or PD, or the reasons for your obstinance.
Cheers,
Tim