| Abstract |
Based on the magnetic saliencies of a PM synchronous motor, it is possible to indirectly detect the position of the rotor by measuring only electric variables. This is the case of the sensorless or self-sensing methods. Even in motors with surface mounted magnets the magnetic saliencies are present due to the saturation that the rotor flux produces in the stator. In order to track these saliencies, an exploring signal is usually injected through the inverter of the drive. By using the inverter, the exploring signal is limited in frequency, limiting in turn the dynamics of the position detection. Furthermore, the exploring signal may detrimentally interact with the controller, it produces audible noise and the sensorless algorithm must be embedded with the controller. A new approach consists in injecting the carrier through a small auxiliary inverter. This method allowed implementing the position detection with total independency of the control algorithm, modulation method and frequency used in the main inverter. This can be an advantage in high power drives where the switching frequency is kept low. The auxiliary inverter must feed only a few watts. However, it is a three-phase one, must be coupled with a three-phase transformer and filter system, and requires to be driven with a rather complex space vector-PWM. The new sensorless method proposed in the present paper also uses an auxiliary inverter, but requires a much simpler hardware system. It uses a one-phase inverter connected to the neutral point of the motor and generates only a square wave voltage with fixed frequency and amplitude. In addition, it will be shown that by injecting the carrier in the neutral point, a higher sensitivity to the magnetic saliency is obtained than by the usual methods. An analysis of the advantages and difficulties of the proposed method, as well as experimental results, are presented. |
