Search

The switching speed of modern power semiconductors in traction converters increases significantly due to the availability of silicon carbide (SiC) semiconductor materials. This results in significantly lower converter losses, but also in a much higher stress on the insulation system of the motor windings. In this publication, therefore, a method is presented, which reduces the dv/dt rate of the converter output voltage to values of today's IGBTs, while maintaining the switching speed of SiC semiconductors and avoiding inherent additional losses. For this, the half-bridge of each converter phase leg is divided into two parallel SiC branches whose output voltages are added by means of an interphase inductor. By a slightly delayed control of the two branches and an additional resonance capacitor at the output, a motor voltage with cosine edges with motor-compatible dv/dt rates can be achieved using the inductor's leakage as resonance inductor theoretically avoiding any resistive damping. [ABSTRACT FROM AUTHOR]

In railway transportation systems the traction motor (and his complex electrical winding insulation system) is one of the most important and critical parts in the drive/propulsion chain. In particular, as the inverter-fed machines are operating with highest power density under suboptimal environmental condition. In addition also the high dv/dt voltage rate causes stress for the motor windings and leads to insulation degradation. Continuous monitoring of the machine windings is thus necessary to detect weak insulation in an early stage. This prevents breakdown, enables strategic- preventive Condition Based Maintenance. The proposed online insulation status monitoring technique is aiming on detection of insulation deterioration far prior an insulation fault actually occurs. It is based on evaluating the motor transient current's response on voltage pulses excited by standard inverter switching. Based on the direct correlation between insulation system degradation and visible changes of the insulation capacitances) an insulation state indicator (ISI) can be calculated. Comparison and smart analyses of these ISIindicator obtained for healthy machine to later measurement values allows an assessment and benchmark of the insulation condition. The applicability and performance of the method is verified by experimental results. The presented investigation and results were cross checked with literature and theory. [ABSTRACT FROM AUTHOR]