Quote Originally Posted by nocturne View Post
I am interested in your answer, and anyone else who would like to respond.
When it doesn't "work".

I'm guessing you won't let me get away with that sleight of hand.

But what "works" is very situation specific. If a theory yeilds enough information to solve the problems I am interested in, then it is good enough.

In fact, my training guides me to pick the "simplest" models that could possibly work, and if I they yeild invalid results, then I go back and revise.

I am also now saying that theories should be "Design-Oriented", in that they should be constructed to help make decsions.

I just recently got some CDs from ardem associates to help me with my work as a circuit designer, and I have to say it is money well spent.

I will vote for R.D. Middlebrook for the best Electrical Engineering Theorist of all time. He is the Don Knuth of E.E.s.