热度 20
2015-4-23 22:04
1433 次阅读|
0 个评论
I graduated college in 1979 (yeah, I know.... makes me about 35, right?). Back then our college Comp Sci department had less processing power than a modern toaster. There was talk, probably apocryphal, of a PDP-11 somewhere in the building, but we in the Physics department cut our teeth on a PDP-8, which had less processing power than a modern slice of toast. Nonetheless, we made do, learning machine code, assembly language and FORTRAN, booting up the machine with a strip of paper tape and entering our programs via punched cards. My first job post-college was working on flight simulator control loading systems at Link-Miles in Lancing on the south coast of England, for the princely sum of 4000 pounds a year. Plus shift differential, of course, flight simulation being a 24/7 kind of business. Flight simulators were all unique, depending on the particular avionics fit of the aircraft we were simulating. The control loading was overwhelmingly analog. At age 35, my memory isn't what it was, but essentially a load cell, located at the base of the control yoke, translated pilot inputs into a force, which was then integrated twice (using opamps and capacitors, of course) to produce both velocity and position terms. These then drove a hydraulic servoactuator which moved the yoke to provide feedback modeled on real aircraft behavior. Along the way, inputs were added into the equation representing flight surface loading, cable stretch, stiction, and so on. All in all, a classic old-school control loop. Old School board layout (Image source: Wikipedia ) The whole control loop was implemented on wire-wrap boards. If you needed to add another term to an equation, you would manually wire in a new socket or two, add a couple of op amps and a few passive components, then test it out. If it worked, you'd mark up the schematic, turn it into the drafting department and you were done. Quality control? Functional safety? Hahahahahaha. Need some logic functions? Crack open the Texas Instruments orange 7400 book. One of these days I'll find a use for some of those bit-slice parts. Analog? I always started with the Burr-Brown data book, because it contained almost magical functional blocks like analog dividers and log amps that went down to DC. Never found a use for those either, sadly. To quote Ellen DeGeneres - my point...and I do have one, is this: back then, working as a professional EE was like being a home hobbyist on steroids. Much of what we did, at least on the hardware side, wasn't too far removed from what you might do tinkering at home in your garage. And just like at home, we felt an almost parental involvement in the final product - perhaps because we designed, built, tested and approved our own (admittedly minor) piece of it, with no committees, peer review boards, or any of that nonsense those indispensable components. Contrast that to the current situation. The increasing complexity of current engineering applications has led to each project being shared among dozens, if not hundreds, of people distributed worldwide. And I think we've lost something along the way in terms of job satisfaction. I'm sorry, but being a small cog in the vast machine that's the 21st-century cross-functional team, doesn't give me quite the same, well, buzz. What say you?