4 ms·
I remember reading that story. IIRC, the bug was was that long running jobs would fail. They eventually traced it to electrical interference generated by a loos
by epmatsw 13y ago
I remember reading that story. IIRC, the bug was was that long running jobs would fail. They eventually traced it to electrical interference generated by a loose floor panel that would move when the user paced over it. Also can't find the link though :(
- jameskilton 13y agoA friend of mine who works in the embedded space recently had a very similar bug. The component would crash when he stood up from his chair. Apparently the hardware connections were so sensitive that tiny bursts of static electricity through the air were enough to disrupt the whole setup.
- jjoonathan 13y agoIf you put an LED, resistor, and MOSFET in series and then attach the gate to a wire, you can often switch the LED on and off by waving your hand around the wire. If you have an oscilloscope, hook it up to the resistor and see if you can find a position of your hand that will switch the transistor at 60hz :-)
- sp332 13y agoStatic electricity can generate hard X-rays http://www.nytimes.com/2008/10/28/science/28xray.html?_r=0 http://www.nytimes.com/2008/10/28/science/28xray.html?_r=0 (in a vacuum at least).
- tobiasu 13y agoThis is quite common when working on unshielded electronics. Standing up from a chair, depending on clothes and chair fabric, generates a huge amount of static. Even just walking past can be enough. Proper hardware development labs have antistatic carpet etc., but in embedded software development, one often doesn't have this luxury and has to be aware.
- jvanenk 13y agohttp://www.youtube.com/watch?v=87jnCABBaeY http://www.youtube.com/watch?v=87jnCABBaeY Overly enthusiastic video demonstrating the issue for those who are curious. :)