For example, cyclohexane chair and boat conformations are produced from rotations around sigma bonds, but staggered and eclipse rotamers of e.g. propane are made in the same way. Where is the difference?
Neither of these are not rotamers - the energy barriers to rotation about sigma bonds in cyclohexane and butane
(note that propane does not have staggered/eclipsed conformers) have relatively low barriers to rotation - i.e. they interconvert rapidly at room temperature.
At the other end of the scale is compounds like BINAP, which has two conformers that can be
isolated - the barrier to rotation is so high that one conformer would have to be be heated strongly for a long time in order to convert it to the other.
Rotamers interconvert slowly at room temperature, such that you might be able to see them as seperate species on the timescale of NMR for example, but they interconvert too rapidly at room temperature to be isolated. Amides and carbamates can quite often show this behaviour:
e.g.
and
So: rotamers are conformers which interconvert slowly at room temperature, can usually be distinguished by e.g. NMR, but normally interconvert too rapidly for isolation. Atropisomers interconvert so slowly that they can (theoretically) be separated and will not undergo appreciable interconversion at room temperature. They are both special types of conformer.