atmega8 to atmega168 porting problem

laborratte · Joined: 27 Jul 2005 Posts: 299 Location: Berlin

Writing random stuff to EEPROM is definitively a sign of unsufficient voltage.

Mega8 and mega168 are different generations of chips, they have different electrical characterics. The interesting part is, that the newer mega168 has nominal better values (faster EEPROM write, higher LOW border on input pin) but seems not as stable as the old mega8 on low voltages. With 8Mhz you should also not drop below 2,7V for mega168 (same as mega8), 1,8V is for max. 4Mhz (maybe you can reduce clock at all, this can lead to more stable EEPROM write - mmmh - maybe not, EEPROM write mechanism is derived from internal RC...).

If you have no chance to measure the voltagedrop-vs.-time before and behind the diode measure the current consuption (total, as you said 100-200mA) and the current through the diode. Then you can roughly calculate the time, when Vcc reaches the difficult point of 2,7V (about 0,6*C*U/I).

Timer-INT: you can work with nested interrupts. enable interrupt in timer-isr like