Influence of the Placement Accuracy of the Brushless DC Motor Hall Sensor on Inverter Transistor Losses
Round 1
Reviewer 1 Report
In the introduction, references are given to works [2-6], which consider the effect of Hall sensors misalignment on the characteristics of the system and methods for overcoming undesirable consequences. However, of these six works, the works [3,5,6] are closest to the topic of research, of which one work belongs to the authors of the peer-reviewed article. Thus, it can be argued that the review of the results obtained by other authors related to the research topic is not representative enough.
Perhaps the lack of links is due to the novelty of the topic. However, in 2021, several papers were published that deal with the same object and the same subject (BLDC Motor, Hall sensor installation error and methods for compensating for this error). Here is a list of works that are not mentioned in the peer-reviewed article.
- Xuliang, J. Xiaoming and Z. Yan, "Compensation method for commutation torque ripple reduction of BLDC motor with misaligned hall sensors," IECON 2017 - 43rd Annual Conference of the IEEE Industrial Electronics Society, 2017, pp. 1862-1867
- Wang, Z. Q. Zhu, H. Bin and L. M. Gong, "A Commutation Optimization Strategy for High Speed Brushless DC Drives with Inaccurate Rotor Position Signals," 2021 Sixteenth International Conference on Ecological Vehicles and Renewable Energies (EVER), 2021, pp. 1-9
- -H. Lim, S. M. Park, B. J. Hyon, J. S. Park, J. -H. Kim and J. -H. Choi, "Study on PMSM Rotor Position Compensation Method for Hall-Effect Sensor Installation Error Using WBG- Based Inverter," 2021 24th International Conference on Electrical Machines and Systems (ICEMS), 2021, pp. 1844-1848
- Liu, Y. Liu and H. Zhang, "A Data-Based HALL Sensor Position Deviation Compensation Method," 2021 IEEE 4th International Electrical and Energy Conference (CIEEC), 2021, pp. 1-6
It should be noted that in addition to references to works devoted to research on the subject of the reviewed article, it is necessary to consider the main ideas proposed in these works. It is also necessary to show the differences between the approach of the article under review and the approaches considered in the works of other authors. In the introduction and further, in the following paragraphs, there is no consideration of compensation methods for inaccurate installation of Hall sensors, proposed by other authors.
Perhaps there is a stylistic and semantic inaccuracy in the sentence on lines 44-47. The introductory word "Unfortunately" apparently refers to a decrease in the accuracy of calculating the angular position due to the shift in the initial angle of the exhibition of the Hall sensors and the deviation of their relative position from 120 electrical degrees. But the regret expressed in the statement refers to the fact that the published papers discuss this problem. In fact, the identification of an existing problem in publications is a good fact.
In line 65, there is a repetition of the collocation "point out to", which was used earlier in line 44. This figure of speech is still alive in the reader's memory when it occurs in the text a second time, which can cause stylistic discomfort.
There is a rather vague explanation of fairly simple things. The decrease in EMF, which is called voltage in line 80, is caused by the deviation of the magnetic angle from 90 electrical degrees. It makes sense to explain this non-optimality using Faraday's law of electromagnetic induction, and explaining the lag of the EMF induced in the stator from the permanent magnet flux by a phase lead of 90 electrical degrees, given by the differentiation of the flux, and the resulting lag of the EMF from the flux due to the minus sign in the EMF formula. The delay in commutation to the supply voltage windings leads to a decrease in torque, but does not lead to a decrease in the EMF induced in the connected windings. After all, the EMF of rotation is also induced in windings that are not connected to a DC voltage source.
Line 84 says "for one chase", which must have meant "for one phase".
The term "BEMF voltage" used in line 91 is questionable. After all, back EMF is the voltage induced in the winding. By writing "BEMF voltage" the authors get a repetition of the form "voltage of voltage", which does not make sense.
The term "BEMF voltage factor" is used in this work. More commonly used in relation to the Ke parameter are the terms "Voltage constant" or "BEMF Constant".
The term "semiconductor keys" (line 102) should be replaced by the common phrase "power switches".
In Western literature, a commonly used synonym for 120-degree commutation (line 107) is the term "Six Step Commutation".
Given the multiple deviations from common terminology in the article, it is worth checking the correct use of the term "120 degree unipolar method" for the method described in lines 107-110. The cited reference [22] recommends a work written by non-native English speakers.
It is also worth replacing the term "branch" in the phrase "transistor uppers of each branch of the inverter" with the commonly used "leg".
It would be correct to use the definite article "the" before the word "Figure" in line 126.
When describing the mechanical part of the stand, the word "clutch" is used (line 130). Perhaps a better term is "coupling". It would be nice to reveal the type of coupling, for example "elastic cam backlash-free coupling ".
In the text, in line 139, the value of the number of pairs of poles p=4 is given, and from table 2 it follows that p=2. Apparently, when writing the article, the authors confused the number of pairs of poles with the number of poles. In this case, it is also necessary to correct the positional accuracy value by writing 0.5 instead of 0.25 (line 140).
From the sentence occupying lines 165-168, it follows that the article takes into account only static (conducting) losses in transistors, and switching losses are not taken into account. This approach should be supported by the assumptions used and reference to published work.
No explanation is given for the relatively small change in losses in the transistors of the phase W compared to the losses in the transistors of the phases U and V with a change in the initial alignment angle of the Hall sensor B.
Pages 5-8 use the designation HB and page 9 use the designation H2 for the same hall effect sensor.
The histograms in Figure 7 contain switching losses. It is not clear how their values were obtained, because from the above description of the experiment it follows that the switching losses were not measured. If the measurement of switching losses was carried out, then there is doubt about the correctness of the evaluation of these losses in the presence of unequal frequency parameters of current probes and differential probes. The difference in frequency parameters leads to a difference in signal delays.
To correctly calculate switching losses in the inverter based on the results of measurements of the instantaneous values of phase currents and drains-sources voltages, it is necessary to use methods to compensate for the inequality of time delays introduced by different frequency parameters of the measuring probes. The difference between the frequency parameters of current probes and Voltage differential probes can be especially significant.
This article does not say anything about methods and compensation for the influence of probes or about acceptable simplifying assumptions. From this point of view, the results of measurements of losses in the inverter cannot be considered correct until the method for measuring the product of instantaneous values u(t)×i(t) is described or the possibility of neglecting the influence of probes is proved..
The conclusion does not contain findings for the effect of the displacement of the position of the Hall sensor on power losses.
Author Response
"
In the introduction, references are given to works [2-6], which consider the effect of Hall sensors misalignment on the characteristics of the system and methods for overcoming undesirable consequences. However, of these six works, the works [3,5,6] are closest to the topic of research, of which one work belongs to the authors of the peer-reviewed article. Thus, it can be argued that the review of the results obtained by other authors related to the research topic is not representative enough.
Perhaps the lack of links is due to the novelty of the topic. However, in 2021, several papers were published that deal with the same object and the same subject (BLDC Motor, Hall sensor installation error and methods for compensating for this error). Here is a list of works that are not mentioned in the peer-reviewed article.
- Xuliang, J. Xiaoming and Z. Yan, "Compensation method for commutation torque ripple reduction of BLDC motor with misaligned hall sensors," IECON 2017 - 43rd Annual Conference of the IEEE Industrial Electronics Society, 2017, pp. 1862-1867
- Wang, Z. Q. Zhu, H. Bin and L. M. Gong, "A Commutation Optimization Strategy for High Speed Brushless DC Drives with Inaccurate Rotor Position Signals," 2021 Sixteenth International Conference on Ecological Vehicles and Renewable Energies (EVER), 2021, pp. 1-9
- -H. Lim, S. M. Park, B. J. Hyon, J. S. Park, J. -H. Kim and J. -H. Choi, "Study on PMSM Rotor Position Compensation Method for Hall-Effect Sensor Installation Error Using WBG- Based Inverter," 2021 24th International Conference on Electrical Machines and Systems (ICEMS), 2021, pp. 1844-1848
- Liu, Y. Liu and H. Zhang, "A Data-Based HALL Sensor Position Deviation Compensation Method," 2021 IEEE 4th International Electrical and Energy Conference (CIEEC), 2021, pp. 1-6
It should be noted that in addition to references to works devoted to research on the subject of the reviewed article, it is necessary to consider the main ideas proposed in these works. It is also necessary to show the differences between the approach of the article under review and the approaches considered in the works of other authors. In the introduction and further, in the following paragraphs, there is no consideration of compensation methods for inaccurate installation of Hall sensors, proposed by other authors.
"
The manuscript has been revised according to the following detailed notes.
"
Perhaps there is a stylistic and semantic inaccuracy in the sentence on lines 44-47. The introductory word "Unfortunately" apparently refers to a decrease in the accuracy of calculating the angular position due to the shift in the initial angle of the exhibition of the Hall sensors and the deviation of their relative position from 120 electrical degrees. But the regret expressed in the statement refers to the fact that the published papers discuss this problem. In fact, the identification of an existing problem in publications is a good fact.
In line 65, there is a repetition of the collocation "point out to", which was used earlier in line 44. This figure of speech is still alive in the reader's memory when it occurs in the text a second time, which can cause stylistic discomfort.
"
The manuscript has been revised according to the following detailed notes.
"There is a rather vague explanation of fairly simple things. The decrease in EMF, which is called voltage in line 80, is caused by the deviation of the magnetic angle from 90 electrical degrees. It makes sense to explain this non-optimality using Faraday's law of electromagnetic induction, and explaining the lag of the EMF induced in the stator from the permanent magnet flux by a phase lead of 90 electrical degrees, given by the differentiation of the flux, and the resulting lag of the EMF from the flux due to the minus sign in the EMF formula. The delay in commutation to the supply voltage windings leads to a decrease in torque, but does not lead to a decrease in the EMF induced in the connected windings. After all, the EMF of rotation is also induced in windings that are not connected to a DC voltage source. "
The manuscript has been revised according to the following detailed notes.
"
Line 84 says "for one chase", which must have meant "for one phase".
The term "BEMF voltage" used in line 91 is questionable. After all, back EMF is the voltage induced in the winding. By writing "BEMF voltage" the authors get a repetition of the form "voltage of voltage", which does not make sense.
The term "BEMF voltage factor" is used in this work. More commonly used in relation to the Ke parameter are the terms "Voltage constant" or "BEMF Constant".
The term "semiconductor keys" (line 102) should be replaced by the common phrase "power switches".
In Western literature, a commonly used synonym for 120-degree commutation (line 107) is the term "Six Step Commutation".
Given the multiple deviations from common terminology in the article, it is worth checking the correct use of the term "120 degree unipolar method" for the method described in lines 107-110. The cited reference [22] recommends a work written by non-native English speakers.
It is also worth replacing the term "branch" in the phrase "transistor uppers of each branch of the inverter" with the commonly used "leg".
It would be correct to use the definite article "the" before the word "Figure" in line 126.
When describing the mechanical part of the stand, the word "clutch" is used (line 130). Perhaps a better term is "coupling". It would be nice to reveal the type of coupling, for example "elastic cam backlash-free coupling ".
"
Corrections have been made in the text of the article.
"In the text, in line 139, the value of the number of pairs of poles p=4 is given, and from table 2 it follows that p=2. Apparently, when writing the article, the authors confused the number of pairs of poles with the number of poles. In this case, it is also necessary to correct the positional accuracy value by writing 0.5 instead of 0.25 (line 140)."
The typical number of pole pairs for a BLDC motor is 4, however, this study used a motor with 2 pole pairs - hence some inconsistency. This has been corrected
"
From the sentence occupying lines 165-168, it follows that the article takes into account only static (conducting) losses in transistors, and switching losses are not taken into account. This approach should be supported by the assumptions used and reference to published work.
The histograms in Figure 7 contain switching losses. It is not clear how their values were obtained, because from the above description of the experiment it follows that the switching losses were not measured. If the measurement of switching losses was carried out, then there is doubt about the correctness of the evaluation of these losses in the presence of unequal frequency parameters of current probes and differential probes. The difference in frequency parameters leads to a difference in signal delays.
To correctly calculate switching losses in the inverter based on the results of measurements of the instantaneous values of phase currents and drains-sources voltages, it is necessary to use methods to compensate for the inequality of time delays introduced by different frequency parameters of the measuring probes. The difference between the frequency parameters of current probes and Voltage differential probes can be especially significant.
This article does not say anything about methods and compensation for the influence of probes or about acceptable simplifying assumptions. From this point of view, the results of measurements of losses in the inverter cannot be considered correct until the method for measuring the product of instantaneous values u(t)×i(t) is described or the possibility of neglecting the influence of probes is proved.."
Specialized software "Tex DPO Power" was used in the tests, which on the basis of the current and voltage waveforms calculates conducting and switching losses. As for the compensation of the offset between the current and voltage signals - the procedures for its correction are now included in the text of the article and result directly from the procedure developed by the manufacturer of the measuring equipment - Tektronix company.
"The conclusion does not contain findings for the effect of the displacement of the position of the Hall sensor on power losses."
Corrections have been made in the text of the article.
Reviewer 2 Report
In general, for the rotor position sensing of BLDC motors, it is based on magnetic elements and Hall sensors. The purpose of this article was to investigate the effect of BLDC motor quality on the correct operation of a control semiconductor system. And the effect of rotor position observation system inaccuracy in BLDC motor on current amplitude. Finally, the effect of sensor misalignment on the power dissipation of individual inverter transistors was investigated through experiments such as thermal imaging cameras.
1. Reference 11 is not in the text. Please correct the sentence appropriately.
2. At the end of the introduction, please describe the method proposed in this study. In particular, I would like you to mention the differentiation of the method proposed by the authors compared to existing studies. Without originality, it feels in the form of a simple technical report.
3. Incorrect placement of the sensor increases losses and asymmetrically increases inverter transistor current. Due to the unbalanced load, some transistors heat up much more than others. This is a very natural fact. The authors noted that a good solution might be to use an advanced software method of correcting sensor placement errors. This method corrects the asymmetry of the signal controlling the switching of the inverter transistor, and it would be better to deal with this as a future research topic.
Author Response
- Reference 11 is not in the text. Please correct the sentence appropriately
It has been corrected
2. At the end of the introduction, please describe the method proposed in this study. In particular, I would like you to mention the differentiation of the method proposed by the authors compared to existing studies. Without originality, it feels in the form of a simple technical report.
This paper points out the strong dependence of the current flowing through individual transistors on the accuracy of Hall sensor placement. This dependence makes it imperative that the performance quality of the rotor position observation element be considered in the design of drive system inverters, especially where compensation algorithms cannot be used due to control system limitations.
3. Incorrect placement of the sensor increases losses and asymmetrically increases inverter transistor current. Due to the unbalanced load, some transistors heat up much more than others. This is a very natural fact. The authors noted that a good solution might be to use an advanced software method of correcting sensor placement errors. This method corrects the asymmetry of the signal controlling the switching of the inverter transistor, and it would be better to deal with this as a future research topic.
These algorithms were and are the subject of the authors' research and are cited in the references. Unfortunately, some simple drive systems cannot perform the calculations necessary to compensate for the inaccuracy of the Hall sensors and it is to these structures that this research was dedicated.
Round 2
Reviewer 1 Report
In the introduction, references are added to the works of third-party authors [3,4], [7,8], [10,11], which consider the effect of inaccurate installation of Hall sensors on the operation of electrical machines.
However, the introduction does not provide a sufficient description of third-party methods and does not show the shortcomings of their methods. Consideration of the shortcomings of third-party works would make it possible to consider the issue of the influence of inaccurate installation of Hall sensors on losses in the inverter as topical. Now we can not make a conclusion about whether the subject of research is relevant or not.
The introduction should contain arguments in favor of the relevance of the issue under consideration. This argument must be conclusive. To do this, it is necessary to evaluate third-party work on this topic. Have the influence of inaccurate installation of the Hall sensors on the losses in the inverter been taken into account in these works? If not taken into account, then you need to write about it directly in the introduction. If the influence of losses is taken into account, then it is necessary to show how the known methods of accounting for losses differ from the method proposed in the reviewed article.
The revised version of the article still does not provide evidence of the relevance of the proposed topic.
The review (1st version) included the following remark: «It should be noted that in addition to references to works devoted to research on the subject of the article under review, it is necessary to consider the main ideas proposed in these works. It is also necessary to show the differences between the approach of the article under review and the approaches considered in the works of other authors. In the introduction and further, in the following paragraphs, there is no consideration of compensation methods for inaccurate installation of Hall sensors, proposed by other authors».
In the corrected version of the article, this remark was not taken into account. The differences between the approaches in the reviewed article and the approaches considered in the works of other authors are not shown [3,4,7,8,10,11].
All other comments of the reviewer, which were given in the review, were taken into account by the authors. And the new version of the article looks almost ready for publication. Correcting the introduction in line with the comment above will give the green light for the publication of the article.
Author Response
The introduction has been rewritten with reference to suggestions.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Thank you for your reply
Author Response
Thank you for taking the time to review our article.
Round 3
Reviewer 1 Report
All comments of the reviewer were taken into account by the authors when correcting the article. The latest version of the article can be published.
