Power Efficiency Characterization with Various Gate Oxide Thicknesses in Class DE Amplifiers for HIFU Applications

Round 1

Reviewer 1 Report

It is important to increase the efficiency of ultrasound transducers in HIFU applications, particularly in catheter-based, interstitial or endoluminal devices, this study compares the effect of gate oxide thickness and the power efficiency. While it is shown that an efficiency of 94% can be achieved, the output power of 1.8 W seems to be too low for HIFU applications. It would be interesting to see how much the transducers will be efficient at higher powers suitable for ablating the tumor. The study may incorporate further simulations to include these results.
Some other comments are below:

 

P1 L29: What is meant by ‘driver’,  please explain.

P2 L 47: …. Therefore a phased array is used to not only steer the beam but to avoid heating sensitive structures.

P5 L 168: Did you mean P_conduction?

Author Response

P1 L29: What is meant by ‘driver’,  please explain.

• We added the following sentence in P1.

“HIFU driver is a self-contained power supply which regulates the power required for transducer or array of transducer.”

 

P2 L 47: …. Therefore a phased array is used to not only steer the beam but to avoid heating sensitive structures.

• We added the following sentence in P2.

“Therefore, a phased array is used to not only steer the beam but to avoid heating sensitive structures.”

 

P5 L 168: Did you mean P_conduction?

• We changed the wrong words to Pconduction in P5.

“Finally, P_conduction and P_S were verified. P_conduction can be determined using the following equation.”

Reviewer 2 Report

Choi et al. designed the CMOS type device and studied the effect of the thickness of the oxide layer.  A few minor revisions are listed below;

1)    The cited references are small in number; cited some more recent work

2)    The author should mention the figures correctly, for example, figures 2 (a) and (b) in the text  in case of more than one figures

3)    The authors should add a comparative table or figure with previously reported value to enhance the worth of their work

4)    The authors just mentioned thick, mid and thin oxide layers what’s about the thickness?

Author Response

1)    The cited references are small in number; cited some more recent work.

• We added some useful and recent references.

 

2)    The author should mention the figures correctly, for example, figures 2 (a) and (b) in the text  in case of more than one figures

• We added the following sentence in P4.

“Fig. 2(a) and Fig. 2(b) show a low-side gate driver in thick- and mid-gate oxide, respectively, and have a cross-coupled structure. Fig. 2(c) shows a low-side gate driver in thin-gate oxide and has an inverter structure.”

 

3)    The authors should add a comparative table or figure with previously reported value to enhance the worth of their work

• We added the comparative result in Table I as Ref. [28] in P7.

 

4)    The authors just mentioned thick, mid and thin oxide layers what’s about the thickness?

• We cannot give the exact oxide thickness because of the NDA. Approximate thickness was written in the paper.

“The thickness of the thick oxide was approximately 50 nm, and that of the thin oxide was approximately 10 nm.”

Reviewer 3 Report

This work is attractive to the authors. The gate oxide thickness plays an important role in FET and other applications. These experiments are systemically tested. But there are several problems that need to be improved.

1. Figure 2 and 3 is unclear to express your idea. They must be clear and easily understood.

2. Figure 4 and 5 can be merged. In present way, the figure looks chaos.

5. The literatures are relativley few. The author need to add more relative literatures since 2020.

Author Response

1. Figure 2 and 3 is unclear to express your idea. They must be clear and easily understood.

• We added the following sentence in P4.

“Fig. 2(a) and Fig. 2(b) show a low-side gate driver in thick- and mid-gate oxide, respectively, and have a cross-coupled structure. Fig. 2(c) shows a low-side gate driver in thin-gate oxide and has an inverter structure.”

 

2. Figure 4 and 5 can be merged. In present way, the figure looks chaos.

• Figure 4 and 5 are different figures depending on the gate oxide thickness, so it is better to keep them as they are.

 

5. The literatures are relativley few. The author need to add more relative literatures since 2020.

• We added some useful and recent references.

Reviewer 4 Report

In this paper the authors show how to obtain high power efficiency, while maintaining high output power levels by using a thin gate oxide thickness of 10 nm. The application of the CMOS process is of significance for the ablation of cancer cells based on high-intensity focused ultrasound devices. However, due to the fact that the paper is based in simulations, more new aspects should be adressed:

1) The description of the ultrasound transducer that can be used for this application is missing. I just saw the schematic in Figure 1 with a ideal and conventional equivalent circuit of a PZT transducer (Butterworth-Van Dyke model). Which is the effect of transducers that have multiple resonances or even losses? How does the power efficiency changes? (see article "Analysis and equivalent circuit for accurate wideband calculations of the impedance for a piezoelectric transducer having loss" by M.J.Hagmann, AIP Advances, 7th August 2019).

2) Which is also the effect of increasing the VDD (DC Voltage supply) for example up to 50 V in Figure 1? Does the power efficiency change? It would be also very interesting to now if the class DE power amplifiers can be also combined with other ultrasound technologies, such us CMUTs and PMUTs.

3) A description of the CMOS technology used is also missed. For many people is also unknown what is a AMS H35 CMOS process. The abbreviations should be described. Also the difference between thick-oxide, mid-oxide and thin-oxide process should be better clarified, as it is the main point of this paper.

4) Line 105 mentions the following: "Digital logic was used to generate the output of the pulse generator". For ultrasound not only power, but also the pulse signals at the output in time domain are of very high importance. Could you add to the paper the input of the digital logic applied to the thin-gate oxide circuit and the output signal that comes out from the CMOS-PZT combination (in time domain)? In this way the ringing/overshoot effects, rise and fall-time and the noise in comparison to the input digital signal can be analyzed. 

5) The paragraph in page 4 (beginning in line 138) is too long, and difficult to follow. I recommend to make it shorter (i.e. split it in various paragraphs).

6) Figure 3. Which is the difference between picture (b) and (c). Is the voltage drop difference (17 V versus 18.7 V) just due to the mid and thin-gate oxide or to the resistor values? Adding the resistor values and the mid and thin-gate oxide thickness in the picture would help to better follow the descriptions.

7) Figure 5. The frequency of these results has to be included.

Author Response

In this paper the authors show how to obtain high power efficiency, while maintaining high output power levels by using a thin gate oxide thickness of 10 nm. The application of the CMOS process is of significance for the ablation of cancer cells based on high-intensity focused ultrasound devices. However, due to the fact that the paper is based in simulations, more new aspects should be adressed:

1) The description of the ultrasound transducer that can be used for this application is missing. I just saw the schematic in Figure 1 with a ideal and conventional equivalent circuit of a PZT transducer (Butterworth-Van Dyke model). Which is the effect of transducers that have multiple resonances or even losses? How does the power efficiency changes? (see article "Analysis and equivalent circuit for accurate wideband calculations of the impedance for a piezoelectric transducer having loss" by M.J.Hagmann, AIP Advances, 7th August 2019).

• Thanks for the in-depth comments. We followed the reference [V1] for PZT analysis. In this thesis, the multi-resonance frequency was characterized for harmonic analysis. Therefore, we added the following sentence in P3.

“The equivalent circuit is shown only for the fundamental frequency. However, it is al-so possible to analyze the third harmonic component, in which case the transferred power to the load is slightly reduced.”

Reference [V1] : Wai Wong, AN INTEGRATED ULTRASOUND TRANSDUCER, thesis, 2013.

DRIVER FOR HIFU APPLICATIONS

 

2) Which is also the effect of increasing the VDD (DC Voltage supply) for example up to 50 V in Figure 1? Does the power efficiency change? It would be also very interesting to now if the class DE power amplifiers can be also combined with other ultrasound technologies, such us CMUTs and PMUTs.

• The result by increasing Vdd to 50 V is discussed in reference [31]. As a result, it was confirmed that the power efficiency was increased, but the value of P_loss were also increased. Therefore, we added the following sentence in P3.

“When the VDD voltage is increased, the overall efficiency is also increased, but the Ploss value is increased and heat is generated, which needs to be considered.”

And the circuit that drives CMUT and PMUT is being studied in our group. We added the following sentence in P9.

“Also, we plan to apply class DE amplifiers to CMUTs, which are being studied a lot recently.”

 

3) A description of the CMOS technology used is also missed. For many people is also unknown what is a AMS H35 CMOS process. The abbreviations should be described. Also the difference between thick-oxide, mid-oxide and thin-oxide process should be better clarified, as it is the main point of this paper.

• We added the reference following sentence in P2. However, the detailed technology such as gate oxide thickness cannot describe in this journal paper because of the NDA. However, the approximated value for gate oxide thickness is described in P1.

 “AMS H35 CMOS technology means 0.35 um technology node for high voltage among the foundry services provided by AMS.”

 

4) Line 105 mentions the following: "Digital logic was used to generate the output of the pulse generator". For ultrasound not only power, but also the pulse signals at the output in time domain are of very high importance. Could you add to the paper the input of the digital logic applied to the thin-gate oxide circuit and the output signal that comes out from the CMOS-PZT combination (in time domain)? In this way the ringing/overshoot effects, rise and fall-time and the noise in comparison to the input digital signal can be analyzed.

 

• In the actual application, the pulse signals is very important as you mentioned. In reference [31], the detailed schematic is shown. However, the pulse output shows less effect to the operation of overall circuits. Also, the output of CMOS-PZT is shown in Figure 4.

Reference [31]

 

5) The paragraph in page 4 (beginning in line 138) is too long, and difficult to follow. I recommend to make it shorter (i.e. split it in various paragraphs).

• As your comments, we split it in various paragraphs in P5.

 

6) Figure 3. Which is the difference between picture (b) and (c). Is the voltage drop difference (17 V versus 18.7 V) just due to the mid and 2 thin-gate oxide or to the resistor values? Adding the resistor values and the mid and thin-gate oxide thickness in the picture would help to better follow the descriptions.

• Figure 3(b) and 3(c) have the same structure, indicating that different voltages are generated depending on the oxide thickness. However, R1 and R2 values should be set differently depending on delay, power loss, etc., so they are simply represented as symbols.

 

 

7) Figure 5. The frequency of these results has to be included

• As your comments, we modified Fig. 5.

Author Response File: Author Response.pdf

Reviewer 5 Report

Dear authors, while the paper is nice in shape, there are a few comments and/or suggestions to improve the manuscript. There is some interest in this type of research, but I found this paper only mildly interesting in its present form. Please strongly consider the following suggestions:

1. The significance of the study is not clear to me and there is a serious literature review gap in this paper. I strongly consider that the distinguished authors did not pay attention to the relevant cited references. In this context, please clarify better the advantages of this paper in the introduction section because in the literature a lot of recent papers consider the same proposed approach. I strongly recommend the authors reconsider the related work section for the literature review and discuss the drawbacks of existing works.
2. Please describe the necessity of each reference in the first section, because multiple citations contradict the ethics of journals with international visibility. Please also reconsider [1-4], [5-7], [9-11], and [13-15].
3. Also, the authors must revise the article sections. The research is not conducted correctly. Section two must be separated into, methodology and results.
4. The authors use many abbreviations. I would create an appendix with their list, so it is easier for the reader to find their meaning.
5. The methodology itself is not well described. There are some simple mathematical quantities, but their use in the proposed technique is not precisely defined.
6. The results analysis is not enough. I cannot see deeply analysis related to them and cannot understand the meaning of the results. Please add more analysis. The readers would benefit from a more insightful discussion of the results and a clear statement about the main conclusions drawn from the research carried out.
7. In the conclusion, the section starts with a brief explanation of the paper's goal and explains what the significant findings are and why your paper is really important. Finally, highlight one or two suggestions for future work.

In this form, the paper can be rejected. The authors must revise the whole manuscript.

 

Author Response

1. The significance of the study is not clear to me and there is a serious literature review gap in this paper. I strongly consider that the distinguished authors did not pay attention to the relevant cited references. In this context, please clarify better the advantages of this paper in the introduction section because in the literature a lot of recent papers consider the same proposed approach. I strongly recommend the authors reconsider the related work section for the literature review and discuss the drawbacks of existing works.

 

• We added the following sentence in P2.

“In order to have high power driving and high power efficiency, research is being con-ducted on GaN devices in addition to CMOS devices [18]. In the driving topology, there were efforts to lower the high voltage DC bias [19,20] and to simultaneously control a large array [21,22]. There were also attempts to increase the range of operating frequencies [23] and to try to implement digitally [24,25]. Researches to analyze the problems in the case of actual application [26] are also being made. However, for catheter ablation systems, research to minimize the generation of additional heat by achieving high power efficiency is essential.”

 

 

2. Please describe the necessity of each reference in the first section, because multiple citations contradict the ethics of journals with international visibility. Please also reconsider [1-4], [5-7], [9-11], and [13-15].

 

• We reduced the multiple citations.

 

 

3. Also, the authors must revise the article sections. The research is not conducted correctly. Section two must be separated into, methodology and results.

 

• It seems difficult to separate the methodology and the results, because the results come after the each methodology analysis.

 

 

4. The authors use many abbreviations. I would create an appendix with their list, so it is easier for the reader to find their meaning.

• We don't use abbreviations a lot, so the appendix is ​​unlikely to be necessary.

 

5. The methodology itself is not well described. There are some simple mathematical quantities, but their use in the proposed technique is not precisely defined.

 

• In Eq. 2 and 3, the derivation process is explained in Ref. [29], and it defines all operations of class DE amplifier. Furthermore, we added the following sentence and Equation in P3.

“Class DE amplifier is a method that actively utilizes the resonance frequency of the load.”

“φ=π(1-2D)        (4)”

 

6. The results analysis is not enough. I cannot see deeply analysis related to them and cannot understand the meaning of the results. Please add more analysis. The readers would benefit from a more insightful discussion of the results and a clear statement about the main conclusions drawn from the research carried out.

 

• We added the following sentence in P2.

“Through this study, the optimal structure of the gate driver circuit according to the thickness of the gate oxide was confirmed, and the effect on the power efficiency was analyzed.”

 

 

7. In the conclusion, the section starts with a brief explanation of the paper's goal and explains what the significant findings are and why your paper is really important. Finally, highlight one or two suggestions for future work.

• We added the following sentence in P8.

“This structure will be suitable for catheter ablation type, and it is planned to perform power efficiency increase analysis in array type in the future.”

 

Round 2

Reviewer 5 Report

The authors revise the paper. I think they superficially respond but it is all right for the article management.