I had a professor who insisted that if you design a digital controller for a physical system the only legitimate test would be against an analog simulation of the system -- that you were asking for trouble if you used a digital simulation. I don't remember his reasoning but I was skeptical at the time. I complained that even then, back in the early 80s, analog computers were obsolete and you couldn't get repairs for them but he had in mind a sort of Moore's law for analog computing, pointing out that op amps were available at Radio Shack for pennies and you could always build your own. All of us in the class were supposed to learn to program the things but there weren't even enough patch wires to go around. I think the only student who actually got the analog simulation to work was some quiet Chinese guy. I turned in a digital simulation and I think everyone else just blew it off.
Later in grad school I had access to the things for just tinkering around. Some attempted projects required several linked together but for all that I never actually got anything working just right.
08-27-2013, 10:22 PM
Threaded Mode