Rotor and stator flux orientations are now standard concepts in vector and direct torque control of ac drives. The salient-pole rotor machines, where magnetic saturation plays a key role, still pose notable problems in flux, rotor position and speed estimations for motion-sensorless control, especially in the low-speed range (below 30 rpm in general), leading to numerous dedicated state observers. This letter introduces a rather novel (or generalization) concept- <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">active</i> <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">flux</i> or <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">torque-producing</i> <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">flux</i> -and its utilization in all ac drives by employing a unified state observer for motion-sensorless control in a wide speed range. The active-flux concept turns all salient-pole traveling field machines into nonsalient-pole ones. The active-flux vector is aligned to the rotor <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">d</i> axis for all synchronous machines and to the rotor-flux vector axis for induction machines. This way, the rotor position and speed observer seems more amenable to a wide speed range, with smaller dynamic errors. This observer, based on the active-flux concept, is pretty much the same for all ac drives. An example of implementation for an interior permanent-magnet synchronous motor with weak permanent magnets and large magnetic saliency that compares very favorably with respect to most signal injection methods, down to 1 rpm and up to 4000 rpm, is provided through digital simulations. Experiments are under way.