On 04/27/2014 05:08 PM, William Levak wrote: > On Sat, 26 Apr 2014, Gerrit Heitsch wrote: > >> An EPROM that loses its contents due to age reverts back to $FF for >> that cell. > > Not necessarily. It could lose the data one bit at a time. It could also > be just on the edge of being 1 or 0, which could result it getting > different results every time you read it. Of course it's on a per bit basis. But when the programming is gone, the bit is always a '1' and never a '0'. So if you read a '0' from a cell, it's the correct value. The problem starts when you read a '1'. Is it a '0' that suffered bit rot or is it supposed to be a '1'? And the flipping bits that give different readings for every dump? That's why I suggested the 1N4148 to lower Vcc. That will stabilize such weak cells again, allowing you to read the correct data. Gerrit Message was sent through the cbm-hackers mailing listReceived on 2014-04-27 16:02:40
Archive generated by hypermail 2.2.0.