Browsing by Author "Wang, Samuel"

Now showing 1 - 2 of 2

A 12-Step Sensorless Drive for Brushless DC Motors Based on Back-EMF Differences
(2015-06-01)
Wang, Samuel
;
Lee, An-Chen
;
機械工程學系
;
Department of Mechanical Engineering
This paper presents a new 12-step sensorless drive for brushless dc motor based on back-electromotive force (back-EMF) differences, which are estimated from the disturbance observer (DOB) structure. Availability of rotor position information with a resolution of 30 degrees and commutation without phase shift are the advantages of using the back-EMF differences. Unlike the conventional methods that obtain part of the back-EMF signal from the manipulation of three phase voltages, the proposed DOB structure can access the entire back-EMF profile. By designing a proper low-pass filter in the DOB structure, the back-EMF difference can be well detected with lower sensor noise. Compensation of delay commutation introduced by the filter is also discussed. Experiment results showed that the proposed sensorless drive is capable of running the motor in lower speed, and performs better under different speed and load torque as compared with the conventional method.
A Current Index Approach to Compensate Commutation Phase Error for Sensorless Brushless DC Motors With Nonideal Back EMF
(2016-06-01)
Lee, An-Chen
;
Wang, Samuel
;
Fan, Chia-Juei
;
機械工程學系
;
Department of Mechanical Engineering
Commutation phase errors during implementation are inevitable in the sensorless drive of the brushless dc (BLDC) motor. This paper analyzed the current waveform during the conducting phase under commutation phase error and nonideal back electromotive force (EMF). The relation between the elapsed time of the commutation phase error with the current waveform, the supplied voltage, and the nonideal back-EMF is derived. Based on this relationship, a reliable optimization method for commutation phase errors compensation is presented. Such method is to minimize the elapsed time of the commutation phase error with robustness properties with respect to parameter variations. In such a way, the peak values of the current at the front and back of the current waveform at each conducting period become more identical, which leads to minimum torque ripple. A digital signal processor-based sensorless BLDC motor controller to implement the proposed method is developed to verify its effectiveness experimentally. Experiment results showed that the proposed sensorless drive is capable of running in different speed and load torque with good performance.