Review Reports
- Gerlando Frequente*,
- Massimo Caruso and
- Giuseppe Schettino
- et al.
Reviewer 1: Anonymous Reviewer 2: Anonymous Reviewer 3: Anonymous Reviewer 4: Anonymous
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors- The paper proposes a hybrid modulation strategy for an asymmetric cascaded multilevel inverter, but the description of the modulation strategy (Section 3) is not clear, and there is a lack of theoretical analysis on how it reduces switching losses and improves waveform quality.
- Table 2 presents the parameters of the PED board. However, the circuit parameters of the three-phase 7-level ACHBMLI hardware-in-the-loop platform are unclear. Where is the three-phase 7-level ACHBMLI in Figure 6?
- There are many typos and formatting errors. For instance, Table 4 on line 287 should be Table 3. The content from lines 255 to 267 is the same as that from lines 315 to 327.
- Why are the presentation forms of the simulation results in Figure 5 and the experimental results in Figure 7 the same? The experimental results should present the oscilloscope results.
Author Response
Reviewer 1
- The paper proposes a hybrid modulation strategy for an asymmetric cascaded multilevel inverter, but the description of the modulation strategy (Section 3) is not clear, and there is a lack of theoretical analysis on how it reduces switching losses and improves waveform quality.
Dear Reviewer,
from the theoretical point of view, the voltage harmonics are generated from both H-Bridge in cascaded connection. Nevertheless, the main contribution stems from the H-Bridge with twice the VDC due to its higher-amplitude voltage harmonics. Indeed, the hybrid modulation is based on the modification of the gate signals of the H-Bridge with twice VDC.
The PD-PWM technique entails higher switching losses due to its elevated switching frequency, while SMS, operating at the fundamental frequency, achieves considerably lower losses. The proposed hybrid modulation technique combines the advantages of both methods while alleviating their drawbacks. As illustrated in Fig. 2, the HBb control signals are generated as in SMS at the fundamental frequency, whereas the HBa control signals are derived using a high-frequency triangular carrier, as in MC-PWM, thereby influencing both switching behavior and harmonic performance.
Regarding conversion efficiency, the proposed hybrid modulation allows for increasing the efficiency because the HBb (H-Bridge supplied with 2Vdc) works at the fundamental switching frequency.
Thus, to address your comment, we have added a theoretical analysis in Section 3 (lines 221-246).
2. Table 2 presents the parameters of the PED board. However, the circuit parameters of the three-phase 7-level ACHBMLI hardware-in-the-loop platform are unclear. Where is the three-phase 7-level ACHBMLI in Figure 6?
-Dear reviewer,
The parameters of both the passive load and the ACHBMLI are detailed in Table 1 and have also been taken into account for the HIL implementation. Moreover, in figure 7, the 7-level ACHBMLI is implemented on a hardware-in-the-loop (HIL) platform, whereas the characteristics of the HIL404 are described in Table 3. For clarity, we have modified figure 7 and added a sentence on lines (343-345).
Additionally, the parameters of the passive load and the ACHBMLI are the same as those in the previous Section 4, as described in Table 1.
3. There are many typos and formatting errors. For instance, Table 4 on line 287 should be Table 3. The content from lines 255 to 267 is the same as that from lines 315 to 327.
-Dear reviewer,
We have corrected the error related the Table 3 and we have deleted the repetition of lines 315-327. Furthermore, we provided a general check of the document to avoid possible formatting errors.
4. Why are the presentation forms of the simulation results in Figure 5 and the experimental results in Figure 7 the same? The experimental results should present the oscilloscope results.
-Dear reviewer,
Our purpose was to adopt the same presentation format because the HIL system enables us to store the experimental data, which can then be used to generate figures with greater clarity and detail. This approach also ensures consistency between the simulation and experimental results. However, we agree with your comment and we modified the presentation forms of the experimental results.
Author Response File:
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors- The introduction seems to be incomplete. It is not clear what the originality of this paper is.
- THD equation in [7] is incorrect.
- In line 203, it is stated that “In this approach, the two H-bridges operate at different switching frequencies, allowing for an optimized balance between efficiency and harmonic performance”. However, it is not clear how the transitions between the switching strategies are implemented. Are there any adopted criteria for that?
- The results only show the steady state efficiency. Since the approach here focuses on combining available PWM strategies, the transient effect of switching from one PWM method to another on harmonics should be investigated.
- The performance of the combined switching strategy should be investigated by varying the switching frequency.
- The lines in 315-327 are the copy of the lines in 255-267.
Author Response
Reviewer 2
- The introduction seems to be incomplete. It is not clear what the originality of this paper is.
-Dear reviewer,
we have rewritten the introduction to emphasize both the objective and the originality of the paper (lines 104–116)
2. THD equation in [7] is incorrect.
-Dear reviewer,
We have corrected equation (7).
3. In line 203, it is stated that “In this approach, the two H-bridges operate at different switching frequencies, allowing for an optimized balance between efficiency and harmonic performance”. However, it is not clear how the transitions between the switching strategies are implemented. Are there any adopted criteria for that?
-Dear reviewer,
Our study focuses on a comparison between a novel modulation technique and the two most popular modulation techniques for the AMLI. Importantly, the transition from one technique to another is not performed continuously. For each test, the microcontroller is restarted and the subsequent technique is loaded. To clarify this approach, we have added some sentences (lines 251-252 and lines 335–337).
4. The results only show the steady state efficiency. Since the approach here focuses on combining available PWM strategies, the transient effect of switching from one PWM method to another on harmonics should be investigated.
Dear reviewer,
As mentioned in the previous comment, in both simulation and real-time validation, no direct transition between different modulation techniques is implemented. For each test, the microcontroller is restarted and the subsequent technique is then loaded. We have clarified this aspect in the manuscript (lines 335–337).
5. The performance of the combined switching strategy should be investigated by varying the switching frequency.
Dear reviewer,
to highlight the influence of the carrier frequency on the efficiency of the ACHBMLI, real-time tests were conducted with carrier frequencies of 5, 10, and 15 kHz for both PD-PWM and HBR. This study does not include the SMS technique, since it inherently operates at a constant switching frequency. As shown in Figure 12, for the same switching frequency and modulation index, the HBR technique consistently achieves higher efficiency compared to PD-PWM. In general, it can be observed that efficiency decreases as the switching frequency increases, although this effect is more pronounced for PD-PWM. Specifically, at a switching frequency of 15 kHz, the efficiency of PD-PWM drops below 20%, whereas in the worst-case scenario for HBR, it remains above 30%. We have described this in lines 416-424.
6. The lines in 315-327 are the copy of the lines in 255-267.
-Dear reviewer,
We have deleted the repetition of lines 315-327.
Author Response File:
Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe concept of asymmetric multilevel converters was investigated about 20 years ago:
- Song-Manguelle, S. Mariethoz, M. Veenstra, A. Rufer, A Generalized design principle of a uniform step asymmetrical multilevel converter for high power conversion, EPE-2001, European Conference on Power Electronics and Applications, 27-29, Aug. 2001, Graz, Austria.
- Song-Manguelle, A. Rufer, Asymmetrical Multilevel Converter for large induction machines drives, EDPE- 2001, International conf. on Electrical Drives and Power Electronics, pp. 101-107, The High Tatras, Slovakia, Oct. 2001.
The paper is limited to a system with two cells per phase only, and limited to [1, 2] p.u DC-link voltages. Therefore, it is not well extended to other system configurations nor different voltage ratios between cells.
The author should extend their analysis to [1, 3], i.e. 1Vdc and 3Vdc for two cells, where the number of levels is Na = 3^n. This should also be tested with their Electtronica hardware. Normally for a high-level publication, the analysis should cover n-cells with many other combinations of dc-link voltages.
- Line 175, page 6: “Modulation Technique Implemented” can be changed to “Proposed Modulation Technique”
- Line 180 – 202, page 6:
- Please simplify these sentences and focus only on the description of the proposed modulation strategy.
- Please add references for these sentences
- What stator currents?
- What problematic harmonics? Please explain
- Please change “working principle” to “operating principle”
- Figure 2, page 6: The blue background is not useful. It can be removed.
Simulation Results
- Line 223 – 226: What version of MATLAB has been used for simulations?
- For these results, what are the switching and fundamental frequencies?
- Figure 4 – Figure 5:
- What voltages (line-to neutral or line-to-line) are shown in these figures?
- Please add title and x-label on each subplot
- Figure 4(d-f): reduce the frequence to show the families of harmonics produced by each modulation strategy
- Please add the THD on each subplot
Description of Bench equipment
- Line 268, page 8: “Description of the Bench Equipment” can be changed to “Description of the Testbench Equipment”
- Figure 6, page 9:
- Label the individual components shown in the figure
Real-Time Validation:
- Figure 7- Figure 8:
- What voltages (line-to neutral or line-to-line) are shown in these figures?
- Please add title and x-label on each subplot
- Figure 4(d-f): reduce the frequence to show the families of harmonics produced by each modulation strategy
- Please add the THD on each subplot
- Figure 9 - 10:
- Please define the terms “IBR”, and “BUS”
- Please add title on each subplot
Limitation of the proposed approach:
- The authors should provide a clear mathematical description of the proposed modulation strategy.
- The proposed modulation technique is specific for 7-level inverter. A generalized formulation of this method is necessary.
- The harmonic characteristics of the output voltage generated by approach is not analyzed.
Author Response
Comments and Suggestions for Authors
The concept of asymmetric multilevel converters was investigated about 20 years ago:
- Song-Manguelle, S. Mariethoz, M. Veenstra, A. Rufer, A Generalized design principle of a uniform step asymmetrical multilevel converter for high power conversion, EPE-2001, European Conference on Power Electronics and Applications, 27-29, Aug. 2001, Graz, Austria.
- Song-Manguelle, A. Rufer, Asymmetrical Multilevel Converter for large induction machines drives, EDPE- 2001, International conf. on Electrical Drives and Power Electronics, pp. 101-107, The High Tatras, Slovakia, Oct. 2001.
The paper is limited to a system with two cells per phase only, and limited to [1, 2] p.u DC-link voltages. Therefore, it is not well extended to other system configurations nor different voltage ratios between cells.
Dear Reviewer,
Thank you for your comment. As noted in the Introduction (lines 104–110), the primary objective of this work is to propose a novel and efficient hybrid modulation technique for asymmetric multilevel inverters. Therefore, this study aims to present the proposed technique and compare it with the most commonly used modulation strategies for this specific configuration, in which the DC-link voltages range between [1, 2] p.u.
.
The author should extend their analysis to [1, 3], i.e. 1Vdc and 3Vdc for two cells, where the number of levels is Na = 3^n. This should also be tested with their Electtronica hardware. Normally for a high-level publication, the analysis should cover n-cells with many other combinations of dc-link voltages.
Dear Reviewer,
the purpose of the upcoming work will be precisely to adapt the proposed modulation technique to this configuration as well.
- Line 175, page 6: “Modulation Technique Implemented” can be changed to “Proposed Modulation Technique”
Dear reviewer, we have changed the title of the paragraph as you suggested
- Line 180 – 202, page 6:
- Please simplify these sentences and focus only on the description of the proposed modulation strategy.
- Please add references for these sentences
- What stator currents?
-Dear reviewer,
In this case, an application example in an electric drive system was presented to demonstrate the impact of MC-PWM on the current, with additional clarifying remarks added to enhance comprehension (lines 188-189). Furthermore, Reference 28 has been included.
- What problematic harmonics? Please explain
Dear reviewer,
The aim is to emphasize that, although the overall THD remains relatively low when applying SMS, certain modulation indices can introduce problematic harmonics into the system. These harmonics may lead to increased torque ripple and degradation of the current waveform quality, potentially affecting the overall performance and efficiency of the drive. Therefore, we have added a sentence briefly explaining this (lines 198–201).
- Please change “working principle” to “operating principle”
Dear reviewer, we changed the sentence.
- Figure 2, page 6: The blue background is not useful. It can be removed.
Dear reviewer,
we changed the Figure.
Simulation Results
- Line 223 – 226: What version of MATLAB has been used for simulations?
Dear reviewer,
All simulation was conducted through MATLAB R2024b. We have added this information in lines 263-264.
- For these results, what are the switching and fundamental frequencies?
Dear reviewer,
The switching frequency is equal to 10 kHz, whereas the fundamental frequency is equal to 50 Hz. We have added this information and described the relationships between the modulation and carrier signals, the fundamental and switching frequencies, as well as how the latter vary depending on the chosen modulation strategy (lines 272-275).
- Figure 4 – Figure 5:
- What voltages (line-to neutral or line-to-line) are shown in these figures?
- Please add title and x-label on each subplot
- Figure 4(d-f): reduce the frequence to show the families of harmonics produced by each modulation strategy
- Please add the THD on each subplot
Dear reviewer,
We have specified in line 277 that the phase voltages are shown in Figure 4. Moreover, we have incorporated all your suggestions.
Description of Bench equipment
- Line 268, page 8: “Description of the Bench Equipment” can be changed to “Description of the Testbench Equipment”
Dear reviewer,
we have changed the title of the paragraph as you suggested.
- Figure 6, page 9:
- Label the individual components shown in the figure
Dear reviewer,
We have included a figure that illustrates the individual components.
Real-Time Validation:
- Figure 7- Figure 8:
- What voltages (line-to neutral or line-to-line) are shown in these figures?
- Please add title and x-label on each subplot
- Figure 4(d-f): reduce the frequence to show the families of harmonics produced by each modulation strategy
- Please add the THD on each subplot
Dear reviewer,
We have specified that Figure 8 represents the phase voltage. In addition, we have implemented your suggestions.
- Figure 9 - 10:
- Please define the terms “IBR”, and “BUS”
- Please add title on each subplot
Dear reviewer,
we have revised the manuscript according to your suggestions.
Limitation of the proposed approach:
1. The authors should provide a clear mathematical description of the proposed modulation strategy.
Dear Reviewer,
We have added a clearer mathematical description, highlighting the advantages of the proposed technique in comparison with the PWM and Staircase techniques (lines 221-246).
2. The proposed modulation technique is specific for 7-level inverter. A generalized formulation of this method is necessary.
Dear Reviewer,
Indeed, the proposed modulation technique has been developed and demonstrated specifically for a 7-level inverter in this work. However, the underlying principles of the method could be generalized to multi-level inverters with a different number of levels. Extending the formulation to a generalized N-level inverter is a valuable direction for future work and is currently under investigation.
3. The harmonic characteristics of the output voltage generated by approach is not analyzed.
Dear reviewer, this aspect is of fundamental importance for validating the proposed technique and has therefore been thoroughly analyzed and discussed through simulations in Figures 4 and 5, as well as in real-time validation shown in Figures 8 and 9.
Author Response File:
Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsAssessment of a Hybrid Modulation Strategy for Asymmetrical Cascaded Multilevel Inverters under Comparative Analysis
In this paper, the authors present a hybrid modulation technique for Asymmetrical Cascaded H-Bridge Multilevel Inverters (ACHBMLIs). The proposed method optimizes the switching scheme by operating the high-voltage H-Bridge at the fundamental frequency, thereby significantly reducing switching losses while maintaining low harmonic distortion levels comparable to traditional Pulse Width Modulation (PWM). The idea of this paper can gain more attention if it were modified for multilevel inverters with DC-link capacitor balancing. Furthermore, the presentation of this manuscript needs to be improved. I have listed several comments as follows:
1) The grammar needs to be corrected in some places.
2) The state of the art should be enhanced by introducing recent works.
3) Figure 1 should be redrawn.
4) The proposed method should be tested under critical conditions ( step change in the Load).
5) The proposed method should be tested in the case of a closed-loop grid-connected system.
6) Can this method be efficiently designed for other multilevel inverters with DC link capacitor balancing?
7) The HIL system should be discussed clearly. The block diagram of the HIL system needs to be addressed.
Author Response
Reviewer 4
Assessment of a Hybrid Modulation Strategy for Asymmetrical Cascaded Multilevel Inverters under Comparative Analysis
In this paper, the authors present a hybrid modulation technique for Asymmetrical Cascaded H-Bridge Multilevel Inverters (ACHBMLIs). The proposed method optimizes the switching scheme by operating the high-voltage H-Bridge at the fundamental frequency, thereby significantly reducing switching losses while maintaining low harmonic distortion levels comparable to traditional Pulse Width Modulation (PWM). The idea of this paper can gain more attention if it were modified for multilevel inverters with DC-link capacitor balancing. Furthermore, the presentation of this manuscript needs to be improved. I have listed several comments as follows:
1) The grammar needs to be corrected in some places.
Dear reviewer,
Thank you for your observation. We have carefully revised the manuscript and rewritten several parts of the text to correct grammatical issues and improve clarity and readability.
2) The state of the art should be enhanced by introducing recent works.
Dear reviewer,
we have updated the bibliography with the following references:
[2] T. A. Taha et al., “Recent Advancements in Multilevel Inverters: Topologies, Modulation Techniques, and Emerging Applications,” Jul. 01, 2025, Multidisciplinary Digital Publishing Institute (MDPI). doi: 10.3390/sym17071010.
[7] N. Karania, M. A. Alali, S. Di Gennaro, and J. P. Barbot, “Advanced High Switching-Frequency Cascaded H-Bridge Multilevel Inverter Based Shunt Active Filter for PV Generation: A Case Study,” IEEE Open Journal of Industry Applications, vol. 6, pp. 262–280, 2025, doi: 10.1109/OJIA.2025.3563851.
[10] C. Nevoloso et al., “On the Inadequacy of IEC 60034-2-3 and IEC 60034-30-2 Standards for Power Losses, Efficiency and Energy Class Evaluation in PWM Multilevel Inverter-Driven PMSM,” IEEE Open Journal of the Industrial Electronics Society, vol. 6, pp. 962–981, 2025, doi: 10.1109/OJIES.2025.3574857.
[24] S. Lemssaddak, A. Ait Elmahjoub, M. Tabaa, A. El-Alami, and M. Zegrari, “A Novel Multilevel Inverter Topology Generating a 19-Level Output Regulated by the PD-PWM Method,” Energies (Basel), vol. 18, no. 13, Jul. 2025, doi: 10.3390/en18133227.
[32] M. Khalid, N. Raj, and S. Member, “Switched-DC Inverter Based Variable Switching Frequency DPWM Technique for Spreading the Harmonic Energy With Reduced Current THD,” IEEE Trans Ind Appl, vol. 61, no. 5, doi: 10.1109/PES.
3) Figure 1 should be redrawn.
Dear Reviewer,
we have redrawn Figure 1.
4) The proposed method should be tested under critical conditions (step change in the Load).
Dear Reviewer,
we have integrated this study into the revised and provided additional comments discussing the behaviour of the proposed method under a step change in the load (see lines 376-381 and Figure 10).
5) The proposed method should be tested in the case of a closed-loop grid-connected system.
-Dear reviewer,
The proposed modulation technique is certainly applicable to closed-loop controllers in both grid-connected and electrical drive applications. Indeed, our future work will focus on applying this modulation technique to an IPMSM drive.
6) Can this method be efficiently designed for other multilevel inverters with DC link capacitor balancing?
Dear reviewer,
Although this topic is of fundamental importance for this configuration, it is not among the objectives of the present article.
7) The HIL system should be discussed clearly. The block diagram of the HIL system needs to be addressed.
Dear reviewer,
We have revised the section describing the operation of the HIL404 (lines 326–332) and included Figure 6, which illustrates the 7-level, 3-phase ACHBMLI implemented within the HIL404. Additional details have also been added to Figure 7 to clarify the components implemented in both the HIL404 and the microcontroller.
Author Response File:
Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for Authors- The theoretical analysis of switching losses has been added from lines 221 to 246. However, since the three-phase 7-level ACHBMLI is simulated on the HIL404 platform, how were the efficiency comparisons in Figures 12 and 13 obtained?
- The Experimental results should be measured using an oscilloscope. And, the standardization and rigor of experimental results require improvement. For example: What are the horizontal and vertical axes in Figures 8(a), (b), and (c)? What are the THD values in Figures 8(d), (e), and (f)?
Author Response
Reviewer 1
1. The theoretical analysis of switching losses has been added from lines 221 to 246. However, since the three-phase 7-level ACHBMLI is simulated on the HIL404 platform, how were the efficiency comparisons in Figures 12 and 13 obtained?
Dear Reviewer, the Typhoon HIL allows the importation of electronic components' datasheet parameters to emulate their behaviour accurately. Since our objective is to compare performance across different modulation techniques, we have therefore chosen to use the components already included by default. This functionality makes it possible to estimate the converter losses, enabling the determination of the ACHBMLI efficiency and the analysis of its variation under different modulation techniques. To clarify this process, we have added a sentence in lines 361-365. Furthermore, a new Figure 13 with a brief description of the switching and conduction losses as a function of the modulation index has been added (lines 448-457).
As shown in Figure 7, several modifications have been made to the test bench, including the use of the HIL606 and the addition of an oscilloscope. This updated setup is presented for clarity; however, it should be noted that the results remain unchanged whether the HIL404 or HIL606 platform is used.
2. The Experimental results should be measured using an oscilloscope. And, the standardization and rigor of experimental results require improvement. For example: What are the horizontal and vertical axes in Figures 8(a), (b), and (c)? What are the THD values in Figures 8(d), (e), and (f)?
Dear Reviewer,
We have replaced Figures 8 and 9 with those obtained using the oscilloscope shown in Figure 7. Furthermore, to enhance clarity and completeness, we have added a sentence indicating the THD values of the output voltage (lines 392-393).
Author Response File:
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors1) The introduction still lacks a clear explanation of what this paper contributes beyond the existing literature. There is no critical discussion of the currently available hybrid switching techniques, which are the main focus of this work. References [28] and [29] are cited only for general background information. However, there are several up-to-date papers on hybrid switching—such as the one below—that should be referenced and discussed to highlight the research gap more effectively: https://doi.org/10.3390/en18082053.
2) The THD equation in [7] remains incorrect. This gives a poor impression, especially considering that THD is one of the main focuses of the paper.
3) In Figures 11 and 12, both the current THD and the efficiency of the SMS are shown as the highest, which appears inconsistent. A higher current THD generally results in higher copper losses, leading to lower efficiency. Therefore, this result needs clarification or correction.
4) In Figures 12 and 13, the inverter efficiency for HBR and PWM appears quite low. Typically, the inverter efficiency exceed 90% over 1 kHz switching, even with classical transistors, IGBT and MOSFETs. It is therefore recommended to verify the accuracy of the measured data and the efficiency calculation method.
Author Response
Reviewer 2
1) The introduction still lacks a clear explanation of what this paper contributes beyond the existing literature. There is no critical discussion of the currently available hybrid switching techniques, which are the main focus of this work. References [28] and [29] are cited only for general background information. However, there are several up-to-date papers on hybrid switching—such as the one below—that should be referenced and discussed to highlight the research gap more effectively: https://doi.org/10.3390/en18082053.
Dear Reviewer,
We have completely revised the Introduction to clearly describe the contribution of our paper beyond the existing literature, adding 5 new references and discussing about recent papers published in the literature (lines 105-150).
2) The THD equation in [7] remains incorrect. This gives a poor impression, especially considering that THD is one of the main focuses of the paper.
Dear Reviewer,
We appreciate your observation and apologize for the oversight. The THD formula has been corrected in the revised version of the manuscript.
3) In Figures 11 and 12, both the current THD and the efficiency of the SMS are shown as the highest, which appears inconsistent. A higher current THD generally results in higher copper losses, leading to lower efficiency. Therefore, this result needs clarification or correction.
Dear Reviewer,
These values are justified by the fact that the same low-order harmonics present in the voltage spectrum are also found in the current spectrum. This occurs because the load does not filter out low-frequency current harmonics. The cutoff frequency of the RL load (filter) is ; therefore, considering a 50 Hz fundamental frequency, attenuation starts from approximately the 13th harmonic with a slope of −20 dB/decade. Such isofrequency components contribute to the active power, thereby increasing the overall efficiency, as the power generated through the SMS also contributes to it.
In contrast, for the PWM and hybrid techniques, the voltage and current spectra do not share these isofrequency components, since the load filters out high-frequency harmonics. It should also be noted that, within one fundamental period, high-frequency techniques involve a switching rate in the order of several kilohertz, whereas SMS requires fewer than twenty switch transitions. As can be easily inferred, this considerably increases the switching power losses in the PWM and hybrid techniques.
To clarify this aspect, we have added new sentences in the paper (449-458).
4) In Figures 12 and 13, the inverter efficiency for HBR and PWM appears quite low. Typically, the inverter efficiency exceed 90% over 1 kHz switching, even with classical transistors, IGBT and MOSFETs. It is therefore recommended to verify the accuracy of the measured data and the efficiency calculation method.
Dear Reviewer,
It is true that the maximum efficiency is generally achieved at lower switching frequencies; however, efficiency does not depend solely on this parameter. Other factors, such as dead time and the amplitude modulation index, also have a significant impact on the system efficiency, as demonstrated in [1]. To clarify this aspect, we have added new sentences to the paper (lines 466-469) and the following new reference:
[1] G. Schettino, A. O. Di Tommaso, R. Miceli, C. Nevoloso, G. Scaglione and F. Viola, "Dead-Time Impact on the Harmonic Distortion and Conversion Efficiency in a Three-Phase Five-Level Cascaded H-Bridge Inverter: Mathematical Formulation and Experimental Analysis," in IEEE Access, vol. 11, pp. 32399-32426, 2023, doi:
Author Response File:
Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsAll comments have been addressed.
Author Response
Reviewer 1
- The theoretical analysis of switching losses has been added from lines 221 to 246. However, since the three-phase 7-level ACHBMLI is simulated on the HIL404 platform, how were the efficiency comparisons in Figures 12 and 13 obtained?
Dear Reviewer, the Typhoon HIL allows the importation of electronic components' datasheet parameters to emulate their behaviour accurately. Since our objective is to compare performance across different modulation techniques, we have therefore chosen to use the components already included by default. This functionality makes it possible to estimate the converter losses, enabling the determination of the ACHBMLI efficiency and the analysis of its variation under different modulation techniques. To clarify this process, we have added a sentence in lines 361-365. Furthermore, a new Figure 13 with a brief description of the switching and conduction losses as a function of the modulation index has been added (lines 449-458).
As shown in Figure 7, several modifications have been made to the test bench, including the use of the HIL606 and the addition of an oscilloscope. This updated setup is presented for clarity; however, it should be noted that the results remain unchanged whether the HIL404 or HIL606 platform is used.
- The Experimental results should be measured using an oscilloscope. And, the standardization and rigor of experimental results require improvement. For example: What are the horizontal and vertical axes in Figures 8(a), (b), and (c)? What are the THD values in Figures 8(d), (e), and (f)?
Dear Reviewer,
We have replaced Figures 8 and 9 with those obtained using the oscilloscope shown in Figure 7. Furthermore, to enhance clarity and completeness, we have added a sentence indicating the THD values of the output voltage.
Reviewer 2
1) The introduction still lacks a clear explanation of what this paper contributes beyond the existing literature. There is no critical discussion of the currently available hybrid switching techniques, which are the main focus of this work. References [28] and [29] are cited only for general background information. However, there are several up-to-date papers on hybrid switching—such as the one below—that should be referenced and discussed to highlight the research gap more effectively: https://doi.org/10.3390/en18082053.
Dear Reviewer,
We have completely revised the Introduction to clearly describe the contribution of our paper beyond the existing literature, adding 5 new references and discussing about recent papers published in the literature (lines 104-150).
2) The THD equation in [7] remains incorrect. This gives a poor impression, especially considering that THD is one of the main focuses of the paper.
Dear Reviewer,
We appreciate your observation and apologize for the oversight. The THD formula has been corrected in the revised version of the manuscript.
3) In Figures 11 and 12, both the current THD and the efficiency of the SMS are shown as the highest, which appears inconsistent. A higher current THD generally results in higher copper losses, leading to lower efficiency. Therefore, this result needs clarification or correction.
Dear Reviewer,
These values are justified by the fact that the same low-order harmonics present in the voltage spectrum are also found in the current spectrum. This occurs because the load does not filter out low-frequency current harmonics. The cutoff frequency of the RL load (filter) is ; therefore, considering a 50 Hz fundamental frequency, attenuation starts from approximately the 13th harmonic with a slope of −20 dB/decade. Such isofrequency components contribute to the active power, thereby increasing the overall efficiency, as the power generated through the SMS also contributes to it.
In contrast, for the PWM and hybrid techniques, the voltage and current spectra do not share these isofrequency components, since the load filters out high-frequency harmonics. It should also be noted that, within one fundamental period, high-frequency techniques involve a switching rate in the order of several kilohertz, whereas SMS requires fewer than twenty switch transitions. As can be easily inferred, this considerably increases the switching power losses in the PWM and hybrid techniques.
To clarify this aspect, we have added new sentences in the paper (449-458).
4) In Figures 12 and 13, the inverter efficiency for HBR and PWM appears quite low. Typically, the inverter efficiency exceed 90% over 1 kHz switching, even with classical transistors, IGBT and MOSFETs. It is therefore recommended to verify the accuracy of the measured data and the efficiency calculation method.
Dear Reviewer,
It is true that the maximum efficiency is generally achieved at lower switching frequencies; however, efficiency does not depend solely on this parameter. Other factors, such as dead time and the amplitude modulation index, also have a significant impact on the system efficiency, as demonstrated in [1]. To clarify this aspect, we have added new sentences to the paper (lines 467-470) and the following new reference:
[1] G. Schettino, A. O. Di Tommaso, R. Miceli, C. Nevoloso, G. Scaglione and F. Viola, "Dead-Time Impact on the Harmonic Distortion and Conversion Efficiency in a Three-Phase Five-Level Cascaded H-Bridge Inverter: Mathematical Formulation and Experimental Analysis," in IEEE Access, vol. 11, pp. 32399-32426, 2023, doi:
Author Response File:
Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsThe paper is well revised according to my comments
Author Response
Reviewer 1
- The theoretical analysis of switching losses has been added from lines 221 to 246. However, since the three-phase 7-level ACHBMLI is simulated on the HIL404 platform, how were the efficiency comparisons in Figures 12 and 13 obtained?
Dear Reviewer, the Typhoon HIL allows the importation of electronic components' datasheet parameters to emulate their behaviour accurately. Since our objective is to compare performance across different modulation techniques, we have therefore chosen to use the components already included by default. This functionality makes it possible to estimate the converter losses, enabling the determination of the ACHBMLI efficiency and the analysis of its variation under different modulation techniques. To clarify this process, we have added a sentence in lines 361-365. Furthermore, a new Figure 13 with a brief description of the switching and conduction losses as a function of the modulation index has been added (lines 449-458).
As shown in Figure 7, several modifications have been made to the test bench, including the use of the HIL606 and the addition of an oscilloscope. This updated setup is presented for clarity; however, it should be noted that the results remain unchanged whether the HIL404 or HIL606 platform is used.
- The Experimental results should be measured using an oscilloscope. And, the standardization and rigor of experimental results require improvement. For example: What are the horizontal and vertical axes in Figures 8(a), (b), and (c)? What are the THD values in Figures 8(d), (e), and (f)?
Dear Reviewer,
We have replaced Figures 8 and 9 with those obtained using the oscilloscope shown in Figure 7. Furthermore, to enhance clarity and completeness, we have added a sentence indicating the THD values of the output voltage.
Reviewer 2
1) The introduction still lacks a clear explanation of what this paper contributes beyond the existing literature. There is no critical discussion of the currently available hybrid switching techniques, which are the main focus of this work. References [28] and [29] are cited only for general background information. However, there are several up-to-date papers on hybrid switching—such as the one below—that should be referenced and discussed to highlight the research gap more effectively: https://doi.org/10.3390/en18082053.
Dear Reviewer,
We have completely revised the Introduction to clearly describe the contribution of our paper beyond the existing literature, adding 5 new references and discussing about recent papers published in the literature (lines 104-150).
2) The THD equation in [7] remains incorrect. This gives a poor impression, especially considering that THD is one of the main focuses of the paper.
Dear Reviewer,
We appreciate your observation and apologize for the oversight. The THD formula has been corrected in the revised version of the manuscript.
3) In Figures 11 and 12, both the current THD and the efficiency of the SMS are shown as the highest, which appears inconsistent. A higher current THD generally results in higher copper losses, leading to lower efficiency. Therefore, this result needs clarification or correction.
Dear Reviewer,
These values are justified by the fact that the same low-order harmonics present in the voltage spectrum are also found in the current spectrum. This occurs because the load does not filter out low-frequency current harmonics. The cutoff frequency of the RL load (filter) is ; therefore, considering a 50 Hz fundamental frequency, attenuation starts from approximately the 13th harmonic with a slope of −20 dB/decade. Such isofrequency components contribute to the active power, thereby increasing the overall efficiency, as the power generated through the SMS also contributes to it.
In contrast, for the PWM and hybrid techniques, the voltage and current spectra do not share these isofrequency components, since the load filters out high-frequency harmonics. It should also be noted that, within one fundamental period, high-frequency techniques involve a switching rate in the order of several kilohertz, whereas SMS requires fewer than twenty switch transitions. As can be easily inferred, this considerably increases the switching power losses in the PWM and hybrid techniques.
To clarify this aspect, we have added new sentences in the paper (449-458).
4) In Figures 12 and 13, the inverter efficiency for HBR and PWM appears quite low. Typically, the inverter efficiency exceed 90% over 1 kHz switching, even with classical transistors, IGBT and MOSFETs. It is therefore recommended to verify the accuracy of the measured data and the efficiency calculation method.
Dear Reviewer,
It is true that the maximum efficiency is generally achieved at lower switching frequencies; however, efficiency does not depend solely on this parameter. Other factors, such as dead time and the amplitude modulation index, also have a significant impact on the system efficiency, as demonstrated in [1]. To clarify this aspect, we have added new sentences to the paper (lines 467-470) and the following new reference:
[1] G. Schettino, A. O. Di Tommaso, R. Miceli, C. Nevoloso, G. Scaglione and F. Viola, "Dead-Time Impact on the Harmonic Distortion and Conversion Efficiency in a Three-Phase Five-Level Cascaded H-Bridge Inverter: Mathematical Formulation and Experimental Analysis," in IEEE Access, vol. 11, pp. 32399-32426, 2023, doi:
Author Response File:
Author Response.pdf
Round 3
Reviewer 2 Report
Comments and Suggestions for AuthorsNo comments.