[julie_m]

The simplest kind of synchronous motor has a stator coil with two pole pieces, and an armature consisting of a simple permanent magnet with a North pole on one side of the spindle and a South pole the other side. When power is applied to the coil, one pole piece becomes a North pole -- so attracting the South pole of the armature and repelling the North pole -- and the other becomes a South pole, so attracting the North pole of the armature and repelling the South pole. The armature rotates accordingly so the opposite poles line up. Then the supply reverses, and so do the poles of the stator. The armature has enough inertia to carry on in the direction it is already rotating, so it carries on rotating in the same direction. But there is no saying which way it will go when it gets that first initial kick from rest.

This also means that if something stops the armature from going round, it will automatically reverse itself to clear the obstruction -- a property exploited in old lighting effects, which used various combinations of moving lamps, filters and mirrors to scatter beams of light. In some applications, such as revolving display stands and microwave oven turntables, it does not even matter which way it is turning. (In a shop window, it can even add a bit of variety if the figures revolve in different directions on different days. Or even used by spies to pass on yes/no or binary-coded messages to one another; with three revolving figures in a shop window, one could signal one of eight possible rendezvous points to a confederate .....)

So in a clock which you want always to go forwards, you can force the direction of rotation if you have some kind of pawl mechanism which slips in one direction, but catches in the opposite direction. (Or you could have a manual starter which gives the motor an initial kick before power is applied, so it continues in that direction. Or, if you are really cheap, you simply rely on the clock hands being turned forwards to force the motor around in that direction; the odds are in your favour that after a power failure, the clock will need to be advanced.)

It sounds as though your clock does have an anti-reverse pawl, but it has become a bit gummed up with dried lubricant or something, or perhaps a weakened or broken spring; which is causing it to take some while before it catches and causes the motor to reverse.

Another way to make a permanent magnet synchronous motor always go in the same direction is to have two windings, with one fed via a capacitor so as to introduce a phase shift, and the pole pieces spaced at 90° intervals. Now the pattern at the first pole piece goes off - weak N - strong N - weak N - off - weak S - strong S - weak S - off, while at the second goes strong N - weak N - off - weak S - strong S - weak S - off - weak N. The third pole piece is of S whenever the first is N and vice versa, and of the same strength; likewise the fourth and second pole pieces. This pattern of two interacting fields will always attract and repels the armature in the same direction, because the weak states still cause the armature to favour a particular position. If the capacitor is wired permanently between the two windings, we can use a single pole changeover switch to select forward or reverse by app.ying power directly to one or the other end of the capacitor. This method also makes the motor more powerful, since there are two coils helping to push and pull the armature around; but it is considered somewhat overkill for a clock.