Hi Peter,

12 MHz micro may run fine until 16 MHz, but in RTC its not just the running, its the exact time base that is effected due to micro failing at the edge.

> Have you really the exact same software burned into both ?

Yes I burned exactly the same code as well as exactly the same crystal in both micros.

I am running it from internal flash only, no external RAM / ROM used.

Yes I have taken care of the interrupt response time in my code.

The problem in my RTC was not due to overlooking the interrupt response in the code because, in my code the time base is 0.5 seconds during which interrupt occurs 6 times. So with 12 MHz crystal it should give an error of 12 uSec. per second. i.e. 1.0368 seconds in one day of 24 hours.

But only by changing the micro from 89c51 to 89c2051 my RTC was giving an error of more than a second per hour. Which is much much larger than would be caused by ignoring the interrupt response in the code.

I still beleive that the problem was due to running the micro on its speed limit.

Babar/