I try to learn from my mistakes, and I've been getting quite an education this harvest. So far this year I've learned:
-when calibrating/programming automatic header height control systems, its possible to set their lower limit too "high." If you do, the operator will then have to fight the automatic system shutting off or going out-of-range any time a rock or old plow ridge jams the head beyond that limit.
-that I should have stood firm last winter and summer, and resisted skimping on repairs to save customers money. There was a lot of pressure to cut corners and, "Fix only what absolutely needs to be fixed, 'cause there ain't much money in farming this year." This fall I've been chastised by a few of those same folks: "Why didn't you fix that when you had it in the shop last summer?"
-I need to quit continuity-testing wires, and start load-testing wires. In short, a worn, frayed wire with only two intact strands of wire will show continuity and low impedance if I simply do a continuity check. But that wire will show low voltage if I load test it with a small light bulb or other "load." That can make a huge difference when you're trying to diagnose problems with mouse-chewed wires that aren't completely severed.
-I need to trust computers more. In most cases, if a combine's self-diagnosing computer says that "Circuit 249 is powered when it shouldn't be," then there's a good chance that circuit 249 is somehow getting voltage when it shouldn't. Start diagnosing the problem by seeing if there really IS power on circuit 249 when there shouldn't be.
-I need to trust computers less--sometimes hunches pay off. If you're getting all sorts of odd, unrelated computer warnings from a half-dozen computers on a combine, and the only thing they have in common is the main power supply, before you start ripping apart dozens of electrical circuits and computers, it doesn't hurt to play a hunch to see if a wire to the alternator came loose.
Like I said, I learn from my mistakes, and I've been getting quite an education this fall.