PIC16F690 EEprom currupt after aprox. one year? SOS :-(

Thread Starter

ph_elec

Joined Sep 24, 2016
23
Having C code to look at would be good, but something generated automatically by this Actum software might be very hard to read.

It occurred to me that maybe the EEPROM is getting hit by electrical effects. Apparently there is an on-chip charge pump which produces a voltage higher than Vdd, and perhaps that's vulnerable to a hostile environment where the rest of the chip is safer. But I'd still look very hard for anything that might be writing to the EEPROM when you don't expect it.

I wonder if there's any way that Microchip could figure out the nature of the damage to the EEPROM, if you sent them the damaged chips. But probably that's a kind of service that they won't do unless you pay them.
Hi John,
Thanks! I have asked Actum Solution to make a C-code out of my project. I have made many of this curciut, and use a standard 5V regulator, and also why after one year, if there should be some noisy / sparks or something, it should happend more random. Normalle I use PIC16F689, and nearly same code, and never seen problems. Let see what happend. Actum Solution have the whole project, and they are VERY godd specialist. I will of cource be back. Regards Per
 

dl324

Joined Mar 30, 2015
16,846
It occurred to me that maybe the EEPROM is getting hit by electrical effects. Apparently there is an on-chip charge pump which produces a voltage higher than Vdd, and perhaps that's vulnerable to a hostile environment where the rest of the chip is safer.
It is possible for cells adjacent to ones being programmed to also be "programmed", and for some "hot electrons" to make it to the floating gate during read (called read program). But those conditions don't occur with high enough frequency to affect the entire array to permanently. Reprogramming would still be possible.

EEPROM lifetime is defined to be when the first bit stops working. For the entire array to stop working it's something other than wearout in the array. Maybe some weak link in the program circuitry.
 

Thread Starter

ph_elec

Joined Sep 24, 2016
23
It is possible for cells adjacent to ones being programmed to also be "programmed", and for some "hot electrons" to make it to the floating gate during read (called read program). But those conditions don't occur with high enough frequency to affect the entire array to permanently. Reprogramming would still be possible.

EEPROM lifetime is defined to be when the first bit stops working. For the entire array to stop working it's something other than wearout in the array. Maybe some weak link in the program circuitry.
Hi
Reprogramming not possibel :-(
 

dl324

Joined Mar 30, 2015
16,846
Reprogramming not possibel
The memory cells wouldn't all fail at the same time. For the entire memory array to fail at the same time, it has to be something other than the typical write wearout scenarios.

But the distinction is moot. If you need to write to the memory array, the device is defective.
 

John P

Joined Oct 14, 2008
2,025
This is why I asked how an EEPROM fails if the number of writes is excessive. Apparently it fails one cell at a time, depending on which cells are used the most. So if this EEPROM is "dead all over" it seems to me that the problem must be elsewhere. If that's the case, it's not Actum's fault. But I'd check for too many write operations also!
 

joeyd999

Joined Jun 6, 2011
5,237
This is why I asked how an EEPROM fails if the number of writes is excessive. Apparently it fails one cell at a time, depending on which cells are used the most. So if this EEPROM is "dead all over" it seems to me that the problem must be elsewhere. If that's the case, it's not Actum's fault. But I'd check for too many write operations also!
For this reason, I always employ write leveling with the ability to lock out bad bytes. This extends eeprom life dramatically.
 
Top