热度 21
2013-4-4 14:09
1612 次阅读|
0 个评论
For special projects, I keep an inexpensive MP3 player handy as I'd rather not tie up or risk my smartphone. Recently, when I pushed the "off" switch, the player wouldn't turn off: in other words, it wouldn't shut up because it wouldn't shut down. It was small-scale version of those scary movies where the killer robot or machine is unstoppable, until someone, somehow, figures out how to get to the unit and shut off the power. I assumed the switch wasn't making proper contract. No big deal, I cracked open the case to see what I could do (one of the virtues of it being a cheap, "don't care about it" player), found the switch on its PC board, and jumpered across the terminals. It still wouldn't quit. Aha ... the problem was apparently not with the switch itself, but some software/hardware soft-failure glitch whereby the unit didn't recognise or respond to the switch closure. I thought to disconnect the battery, but it was soldered in, so that avenue of easy termination was closed off. So I just let the unit run down its battery all the way, then recharged it, which forced a software reset (as I expected it would). I have always been uncomfortable with presumed "on/off" switches and buttons that really aren't, but which actually are soft functions. This is where the system processor must recognise a switch closure, and put the unit into a quiescent sleep mode (and do the opposite to wake up). That is not the same as cutting off the power source. It assumes that everything is working just right, and the switch closure will be sensed and acted upon. I know that these soft functions including power controls are now standard on many products, but they still making me feel a little creepy. Obviously, a stuck MP3 player which won't turn off is not a crisis. But a motor, vehicle, or instrument that won't turn off: well, that's another story. Sometimes you can get to the power source and kill it, by pulling the plug or battery connector, but sometimes you can't. Or perhaps you could, but it would be too late. Many years ago, I worked for a company that made industrial machinery, powered by big electric motors or hydraulic pumps. For many years, everything was controlled by true hardware, such as switches and relays. As software and processor-based controls came into their own, we switched the controllers over, of course. But a big, red, unambiguous, mushroom-topped STOP switch which interrupted primary power was still key part of every system, for obvious reasons. Have you ever had a system which you couldn't get to turn off or stop-but needed to? How did this experience affect your design approach? Bill Schweber EE Times