??? 04/09/05 21:24 Read: times |
#91314 - remarks Responding to: ???'s previous message |
First, if the programmer does not recognize the chip, it has nothing to do with FLASH wear; rather, with wear of the programmer (ZIF or connectors). The consequences of FLASH wearout are certainly dependent on particular technology and circuit details and maybe also on the memory cell's "history". As FLASH/EEPROM memory cells are not too dissimilar to EPROMs, the "dead" cells will behave similarly. During the years, I have seen all the possible behaviour, "stiff" cells which could not been programmed or erased; "alzheimer-sick" cells "forgetting" their state, "gambling" cells returning random values. Detrimental for erasable memories is not only the wear, but also various kind of radiation, electrical (over)stress, heat(!). I always try to keep a "fresh" chip for developing and "get rid" of the "used" (but not abused) one to production pieces. Btw. I don't think there is any error correction in common microcontrollers' FLASH (it is in flash cards/disks); and if I remember right, the counter in GALs was in fact implemented in the programmers' software. Jan Waclawek |