High-Dynamic Speed Control for High-Speed Turbo Molecular Pump Under Impact Load Disturbances

The magnetic levitation turbo molecular pump (MLTMP) is driven by a high-speed permanent magnet synchronous motor to create a high-vacuum environment for semiconductor manufacturing and other industries. When gas pressure changes abruptly, the MLTMP experiences significant speed fluctuations, which would degrade industrial productivity and even destabilize sensorless control. To address this issue, this paper proposes a high-dynamic speed control strategy for high-speed MLTMP under impact load disturbances. First, a speed control method incorporating a load torque observer and an active disturbance rejection controller (ADRC) is designed. Second, based on the adaptive law developed in conjunction with the ADRC process, an adaptive incremental compensation method is proposed. Finally, the proposed high-dynamic control strategy is performed on a high-speed MLTMP. Experimental results show that the strategy achieves significant attenuation of speed fluctuations, shortens the speed convergence time, and improves the dynamic performance of the MLTMP. Note to Practitioners—The motivation of this study is to address the speed control problem of high-speed magnetic levitation turbo molecular pumps (MLTMP) used in semiconductor manufacturing. During processes such as chip bonding, sudden pressure surges in vacuum chambers lead to rapid increases in MLTMP load torque, degrading pumping speed and compression ratio, and destabilizing sensorless control systems. Conventional speed control methods often rely on an accurate mathematical model of the motor, focus on steady-state performance, and have limited dynamic response capability. The proposed high-dynamic speed control strategy combines a Luenberger load torque observer with an active disturbance rejection controller (ADRC), along with an adaptive incremental compensation (AIC) method. It effectively suppresses speed fluctuations, shortens convergence time, and lowers instantaneous motor power to avoid saturation risks. This strategy replaces traditional PI control and can be implemented in real time via digital signal processor, making it suitable for high-speed, complex load scenarios. This strategy is validated on an experimental platform. Future extensions may include dynamic control of other high-speed rotating equipment such as turbo compressors and precision centrifuges.