>sub-milliseconds to microseconds Time difference? Of what? - Two quite unstable DCO oscillators? You shall specify not only momentary precision but also long term drift, synchronization frequency as well. For more or less reliable timekeeping I would definitely not use DCO but crystal oscillator and measure frequency difference of two clocks by simply sending periodic (1hz or 10hz) time stamps from one system to another, calculate frequency difference and do the rest in software w/o adjusting oscillator.
↧