There is no potential between the micro's ground and the 7805 ground and the adaptor's output ground. I think this is what you meant.
"potential" in this respect is not measured with a voltmeter, it is measured with a storage scope. I have seen very short spikes of great magnitude between "ground" and "ground". Make sure your stuff is grounded in one place only, if that is impossible, optoisolate.

By the two clocks biting each other, do you mean the two crystals , one for the micro and the other for the RTC ?
As an example, yes. That is the reason you need to shield them and the clock traces from each other. My latest example of 'biting clocks' was a 20kHz oscillator for a chopper-stabilized amplifier that I had to replace with a T2OUT signal.

. This post would have been completed in a quarter of the time if i had not to go back and change all the u's into you's.....
Your effort is greatly appreciated.

Erik