| ??? 12/23/01 03:17 Read: times |
#17997 - RE: nov-mem getting corrupted |
Erik,
interesting idea about using two supervisor chips... however, the costs are about a buck in volume for that addition. Sometimes the simplest things ca radically reduce such problems. For example, if you have an event triggering an alarm andyou need to write a timestamped EEProm record... ordering can be important. Sometimes you can write the record first before alerting the "outside world." The advantage here is that you can put the micro into a low power configuration first, write and verify the EEProm, then power up and go LOUD&BRIGHT! The advantage here is that a well designed power supply can live off its reservoir caps for awhile in low power mode, and while lacking a supervisor chip to warn you, the low current mode REDUCES the probabilities of such interference greatly. Once you go LOUD&BRIGHT, you are often drawing your maximum design current and the reservour caps will drain real quickly in a interruption of power glitch. And, like the tree falling in the forrest (does it make a sound if no one is around to hear it?) ... does an interrupt really occur if micro is taken out before in can act upon it??? EEProm: Write THEN Validate. If the data is worth paying for, its worth making sure its there. |
| Topic | Author | Date |
| nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted MORE | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
| RE: nov-mem getting corrupted | 01/01/70 00:00 | |
RE: nov-mem getting corrupted | 01/01/70 00:00 |



