> > The "max-write-to-each-single-cell" is still valid.
> > See paramater D120 in most (all?) data sheets.
> > 
> > So even if refreshed, you can not refresh a specific
> > EEPROM cell more time then the D120 paramater says.
> > A refresh is just a normal write and counts againts D120.
> > 
> > > I thought 1M-and-then-no-more was a bit weird.
> > 
> > Why did you think it was weird ?

> Because I didn't know EEPROM actually died. Plus, the fact that a
> logger that writes data to EEPROM every second will die in a few days.

As others have said, there are other ways to deal with that.
- Use Ramtron devices.
- Use standard RAM and copy to the EEPROM on a power loss (must
  have power loss detection and some caps or batteries to power the
  PIC during the EEPROM write).

> I hadn't read the paragraph you cited (who wrote it? It seems I don't
> have that post).

I'm not sure anymore what paragraph that was.

> I thought you needed to refresh a frequently used cell every D120
> cycles,...

Since a "refresh" is nothing but a normal write, a "frequently used cell"
would *never* have to be refreshed, by definition.

> and any less-used cells (that are affected during writes
> to other cells) every D124,...

Right.

> and doing a refresh would reset both counters. It would've been nice.

The D120 counter (the "max writes" counter) is incremenet for each
individual cell at each write, no matter if it's a normal "application 
write"
or a "refresh write". And it's never reset, just as you said.

The D124 counter is incremented for each used cell whenever *another*
cell is written. It's reset on refresh (or an "application write", of course).

> That was my inexperience speaking. Thanks for clarifying this, I
> never would've thought it was such a problem.

There is definitly a need of some kind of "whate paper" on this issue :
-)

Happy New Year !
Jan-Erik



-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist