Consider a 4Q motor control with a H bridge driver. The bridge is being driven in PWM by means of a current control loop inside a speed control loop. The motor is required to ramp up, maintain a velocity, and then ramp down (braking).
There is a lot of conventional wisdom out there which states that whilst braking you are using the motor as a generator (regenerative) and therefore a load resistor must be switched into the DC link to dissipate the kinetic energy from the load. Attempting to brake the motor simply by taking the PWM to the opposite extreme (presumably respecting the max current of the motor) will cause the energy to be dissipated in the armature resistance.
An alternative view is the physics one. If we apply a force to a load it starts moving, and will carry on moving until we apply a force in the opposite direction. If this problem were in space, we would fire a rocket to accelerate, and then turn the ship round to fire in the other direction to ramp down again (hey, who remembers Defender :-) . Now, in our 4Q bridge all we are doing is forcing a current to flow against the generated emf, that is we are creating a force in the opposite direction. So where does the kinetic energy go?
And which is the correct way of looking at the problem?