Position sensing for small, indoor arenas

This is the *easiest* way to bring a system to its knees in zero time -- figure out what people are ASSUMING as invariants... then, violate one of them! And, when they invariably respond with "You aren't supposed to *do* that!", just grin and say, "Oh, yeah? *MAKE* me...!"

Reply to
D Yuniskis
Loading thread data ...

... or the motor is controlled incorrectly.

Anecdote warning!

Some years ago my code nearly caused the sliding cover of an optical telescope to drop off the dome of the telescope building. The cover was about 3 meters by 6 meters in size and was driven from the telescope control computer with two binary commands, "open" and "close". One night, the astronomers had told the computer to open the cover, but before it had fully opened, some fog came up and the astronomers immediately told the computer to close the cover (to avoid condensation on the telescope mirror). The computer switched off the "open" command and switched on the "close" command. The door kept on opening... because the motor was of a type that has to be given time to really stop before it can change direction. Switching immediately from "open" to "close" simply kept the motor running in the opening direction.

The cover hit the "fully open" limit switch, which was hard-wired to stop the door. However, it was wired to interrupt the "open" command only, so that the "close" command still worked. Thus the limit switch had no effect, and the cover kept on opening. Luckily the site engineer was present and was able to stop the cover by cutting power and propping it up with some timber. Damage was slight.

It turned out that the dome, the cover, and the motor electronics had been bought off the shelf. The cover was designed to be controlled manually from a hand-held box with two press-and-hold push-buttons, "open" and "close". However, the box had a sliding lid over the buttons so that, after pressing the "open" button, say, the user had to release the button and slide the lid to its other position before the "close" button could be pushed. This gave the motor time to stop and prepare to change direction.

Unfortunately, this mechanically implemented delay was omitted from the specs when the telescope cover was put under computer control. And it was not tested. And my face was red.

--
Niklas Holsti
Tidorum Ltd
 Click to see the full signature
Reply to
Niklas Holsti

As a child, I had a similar encounter with a (Ford) Station Wagon. It was the first (?) year the rear tailgate had the option of opening as a traditional "drop down" tailgate *or* a conventional

*door*.

It had an electric window in it -- a delightful toy for a kid to play with! But, the window would not "go up" with the door open. Nor would the door *open* unless the window was down.

In order to understand the mechanism, I opened the window, then opened the tailgate "as a door". Using my tiny hands (remember, I'm a little kid), I managed to force the two latches into the "closed" position -- to trick the door into thinking it was closed. Then, turned the key in the switch (window control) and grinned ear-to-ear as the ~4 ft wide sheet of glass glided up out of the tailgate.

*Literally*.

It seems there is no real "limit switch"; instead, the mechanism counts on the window being obstructed at the end of its track to prevent it from going further. With the door *open*, that obstruction was no longer in place. So, the window raised itself *out* of its track...

My folks were not happy -- the car was three days old at the time...

Reply to
D Yuniskis

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.