Yes, Rob, maybe I were stuck in the way I did it with a small EEPROM. If the life expectency (still being discussed) get to be rather large, successful approximation definitely will speed thing up.

Any other suggestions out there ?

Erik