Robert James Kaes wrote:

>On Fri, 17 Dec 2004, Michael Rigby-Jones wrote:
>  
>
>>This was covered in a recent thread.  You DO need to perform a refresh
>>of the entire EEPROM if one or more cells are written to more than the
>>number specified in the data sheet.  I don't know the mechanism behind
>>this, but writing to one cell essentialy removes some charge from the
>>others so data will be lost unless all cells are refreshed.  This is
>>explained in the datasheets.
>>    
>>
>
>I thought I understood the refresh requirement, but I guess I was
>unclear about the cycle limit.  The way I read the datasheet was that if
>I refreshed the EEPROM before the cycle limit, it would basically reset
>the cycle count to zero.
>
>However, it would appear that this is mistaken.  The cycle limit is a
>"hard" limit.  Is this correct?  For example, if the data sheet says
>that the cycle limit before refresh is 1M cycles, then the EEPROM can
>_only_ handle 1M cycles before it starts to fail.
>
>So, in my application where the EEPROM value is updated once every two
>seconds, I would see a failure rate after approx. 23 days.
>
>  1M cycles / ((30 cycles/min) * 60min/hour * 24hour/day) =~ 23 days
>
>Is my understanding of how the EEPROM work correct now?  There is no way
>for an EEPROM to work "forever"?
>
>Thank you for your time.
>	-- Robert
>
>  
>
I suggest you use a RAMTRON I2C or SPI device. I've never seen one fail, 
and I've got
a LOT of them in the field doing almost the same thing you are doing.

Microchip's EEPROM arrays seem to be weak sisters. I use it very sparingly.

--Bob

-- 
Note: Attachments must be sent to
 attach@engineer.cotse.net, and 
 MAY delay replies to this message.
        520-219-2363

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