there is no official paper. Although nearly every EEprom manufacturer
knows this problem, no one speaks about it
I was told that there are a very few EEprom devices which don't show this
problem. I think it was one type from MicroChip, but I'm not sure.
The problem can easily be reproduced. In the beginning of AVRs we had
this problem with a construction where some external parts supplied a
very low current through the protection diodes, about a few uAmps.
We could measure a voltage around 1.2V with a 50Hz overlay of ~200mV
at the VCC lines of the board. After a few minutes (or an hour?) some
EEprom cells got a random content.
Lesezeichen