Remarks on time-energy uncertainty relations