Yahoo Groups archive

AVR-Chat

Index last updated: 2026-04-28 22:41 UTC

Thread

RE: [AVR-Chat] RE: ATmega88PA EEPROM Memory Life

RE: [AVR-Chat] RE: ATmega88PA EEPROM Memory Life

2012-12-13 by Tim Mitchell

----Original Message----
Show quoted textHide quoted text
From: AVR-Chat@yahoogroups.com
[mailto:AVR-Chat@yahoogroups.com] On Behalf Of Bob Paddock
Sent: 13 December 2012 02:44 To: AVR-Chat@yahoogroups.com
Subject: Re: [AVR-Chat] RE: ATmega88PA EEPROM Memory Life

> I've always gone conservative and assumed EEPROM wears by
> pages.  Can 
> someone point to a definitive Atmel answer?

As far as I understand it, and I was told by an Atmel FAE who may or may not have known what they were talking about, Eeprom wears by bits. When you write a bit from 1 (erased) to 0 (programmed) then that's one cycle for that bit. If the bit pattern of the stored number doesn't change the state of that bit, then it isn't a write cycle for that bit.

And the magic 100,000 figure is not a real-life tested value, it is just a very pessimistic calculation. A bit like the quoted 50,000 hour life for LEDs. You will always get more than 100,000 writes, but in a critical application you can't rely on it. I have seen results on the web of tests people have done where the endurance on their tested devices is over 10 million writes and counting.

-- 
Tim Mitchell

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.