rbj said:
yung, i don't get the "Unless" caveat. even if you are very good at analog circuit design, you still need to do some programming. how else do you mathematically model what your doing?
yungman said:
I was a firmware engineer when I first started. I don't like programming and I never have to do programming since.
no one would know this more firsthand than you...
Programming is not the same as using simulation programs.
... but that, i disagree with.
assembling a program, and i mean from blocks, which if you drill down far enough eventually are instructions (so don't mean what an "assembler" does) is conceptually no different than assembling and connecting components for PSpice or LTSpice or whatever variant of Spice you're using. when you lay out the nodes and what goes in between the nodes, you are essentially programming.
i don't mean that a SPICE simulation is equivalent to a full-tilt Von Neumann model (a.k.a. "Turing machine"), but it could be if the elements of the language have primitive enough members in the set (and it might, depends on how many arbitrary bends you can put into the volt-amp characteristics of a component).
think of it this way: you enter in all of the "data" of your PSpice simulation, this data becomes bits in semiconductor memory and
that computer with those particular bits, high and low, is
itself simulated (with a very big program) in PSpice. the fact that a sequence of instructions is the same as a set of data that fully describes the sequence of instructions means that when you program SPICE, you are programming. i 'spose, if i had infinite time, i could wire up an ASCII word-processing program in SPICE, and it would make comparisons of ASCII characters with diodes and BJT switches that become elements in the SPICE program.
entering in data into PSpice for PSpice to chew on is qualitatively no different than programming in instructions for a compiler or interpreter to chew on. it's the same thing. it's the same abstraction.