What happens when OU is reach?

[Honk]

So I believe that the torque is constant regardless of speed/rpm (speed of torque is the same as speed of light), but the useful torque which we can convert in to energy are changing in a logarithmic scale.
Vidar

You might be right in this, except the scale of the torque.
The torque of an electric motor is linear, not logarithmic. http://www.gizmology.net/motors.htm
You can see the typical torque curve at the attached picture.

Regarding the original topic "What happens when OU is reached?", I believe I pretty much know how it will turn out.
1) No magnet motor will ever run on it's own without adding control energy.
    If it does run, the torque will be extremely low due to the balancing effect of the magnets that limits the accessible torque.

2) Having great stall torque without increasing stall currents is the key to sucess. Input current must stay unaffected by stall.
    And great stall torque allows for high rpm:s at no load.

3) The peak point of possible Overunity is reached when the motor is loaded down to half it's free spinning rpm.
    The control energy of the motor should not be affected by the load or speed. This is very important.
    E.g, you measure 50ft-lbs of torque at stall, and the free rpm reach 3000 at no load. Let's say the control energy is 500 watts.
    At 3000 rpm free spinning you don't have any Overunity. The motor consumes 500 watts. COP = Zero
    When you attach a generator to the shaft and load it down to 1500 rpm you will have access to half the stall torque at 25ft-lbs.
    According to this formula HP = (TQ * 2 * PI * RPM) / 33000 it's easy to calculate the motor output.
    HP = (25 * 2 * 3,14 * 1500) / 33000 = 7,14Hp = 5,32KW. The motor will have a COP of 10.6.

4) The motor will not be Overunity when free spinning or being loaded down below a certain point.
    But it will be greatly Overunity within certain working points of the RPM range. And peak COP will occure at half the free spinning rpm.
    Just make sure that you have great stall torque when designing the motor and keep the control energy to minimum (to pass any sticky spots).

I believe this is why inventers or the experimental "try out" guys miss the goal.
They just look for free rotation but they should in fact look for overunity at load within the usuable RPM range.

[eavogels]

In most of my experiments I build in some kind of emergency stop for example an easy to remove stator magnet or an ordinary break system.
No need to tell that I never used that, but when I now and then make a test without an emergency stop, I stand close to the door, with safety glasses on.
Eric