A comment that Specification D120 must be adhered to for EEPROM to not require refreshing.
It appears that long term constants that are infrequently changed need to be refreshed.
Can anyone explain this?
How 'long' before a refresh is needed?
Is a refresh reading out the EEprom data and then rewriting it?
I think what it's saying is that if you have variables which you are updated very frequently, and variables which are not updated frequently, then you need to refresh the ones which don't update frequently. If all of your variables are frequently refreshed, or all of them are rarely refreshed, then you shouldn't need to worry.
It says on the same table that the EEPROM data retention is >40 years so maybe just refresh once every 40 years
Yes refreshing is just reading and rewriting.
I could be wrong on what I think they're saying, though.
Byte endurance... as I have understood... if you chip runs at normal temperature.. -40 > +85. 1 million write cycles otherwise if you run at high temperatures > 85.. then expect the byte location's endurance to only 100 thousand write cycles.. ( still an awful lot though ).. Gobbledok is correct though... if you write a byte to EEprom..it will be there 40 years later ( unless its over written ).