Revised Permanent Magnet Model
Fig. 1 depicts an edge-view of an uncharged, non-spinning, disc-shaped permanent ceramic magnet. The axis is subjected to a torque that points toward positive z. We shall suppose that the magnetís response is precisely that of an equal-mass, unmagnetized disc. In brief, the magnet theoretically has no intrinsic angular momentum, and consequently there is no precession.
A Disc Subjected to a Torque
In previous articles such a magnet has been modeled as an array of uncharged microscopic current loops. There it was usually suggested that a spinning ring of one-sign charge (typically positive) is superimposed on a non-spinning ring of opposite-sign charge. In view of the theoretical non-precessing behavior of the parent magnet suggested above, however, there seems reason to reassess this model. For a spinning ring of one sign of charge should have electromagnetic (and perhaps mechanical) angular momentum and should precess when subjected to an external torque. Collectively the myriad microscopic current loops, comprising the permanent magnet, might have a non-trivial angular momentum.
A suggested solution to this conundrum is that each microscopic (uncharged) current loop be modeled as superimposed, counter-rotating rings of positive and negative charge. The net angular momentum of each loop would then be zero.
It is noteworthy that each such loop would still become electrically polarized when it translates in its plane. Thus the myriad electric dipole fields would still sum to have radial electric field components when the overall magnet spins. And of course when the entire magnet is spun it will have mechanical angular momentum, quite as an unmagnetized counterpart will.