Super-Frequency Sampling for Thermal Transient Analysis
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors use a novel super-frequency sampling technique to improve the thermographic measurement of thermal transients.
They sample a thermal transient repeatedly with a periodic but non-uniform measurement that effectively increases sampling density and reconstructs parts of the transient that would otherwise be lost at the standard sampling rate.
Thermal transients often occur on sub-ms time scales, but their infrared-camera sampling rate was limited to 1 kHz.
Their technique increased the effective sampling rate by an order of magnitude, to 10 kHz, resulting in an increased temporal resolution of 100 μs.
This technique has applications in power electronics and other small-scale electronic components.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsIn the second sentence of the Introduction, references to specific scientific publications should be added to the mentioned applications of thermography.
"These include the investigation of buildings [x], bio-medical questions [y], and electronic components [z]".
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for Authors
The paper describes the method for obtaining high-frequency temperature response by combining high-frequency excitation with low-frequency sampling.
The idea of method seems to be valid, however its implementation does not seem convincing enough.
First, it is not clear, why the excitation and sampling can not be synchronized exactly, using the camera synchronization signals, which will exclude an additional step of adjusting the period with the data itself. Such step necessarily reduces the exactitude of reconstruction. Additionally, it is desirable to compare the obtained difference in frequencies with a real one to estimate the exactitude of reconstruction method.
Second point, which seems to me more important, is lack of discussion about influence of camera integration time on the obtained results. The temporal resolution can not be much better, than camera integration time. If the period of excitation modulation approaches the integration time of camera, the obtained results will show temperature changes, which are smaller, than real ones.
Thus, it seems to me that the methodology is flawed, and additional experiments are needed.
Comments on the Quality of English LanguageI did not detect issues with English, though sometimes the sense is not very clear.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsIn the manuscript “Super-Frequency Sampling for Thermal Transient Analysis” submitted by Simon H. Anke, Nils J. Ziegeler, Peter W. Nolte, and Stefan Schweizer, the authors propose a periodic non-uniform sampling technique for thermographic transient measurements. The new technique increases the temporal resolution by one order of magnitude for the analysis of thermal behavior.
The authors first introduce the background of super-frequency sampling and the principles of periodic non-uniform sampling. An algorithm of true period estimation is proposed. The authors apply the algorithm to thermal transient and impedance measurements. In the experiments, the authors present an increase in temporal resolution by comparing it with the traditional method. The authors also analyze the cumulative distribution of the reconstructed sampling period.
Overall, this is a manuscript of good quality. The structure of the manuscript is clear with logical reasoning. The data is well presented. It is not a huge breakthrough in the field, but I find this topic interesting, and it is worth a detailed investigation. This manuscript can be published with the consideration of the following remarks:
(1) In section 3, the authors use a speed-up factor of 10 in the experiment. Why do the authors choose this value to present? How does the algorithm perform with a larger speed-up factor? I understand that in section 4.2, the authors warn of the possible failure of the algorithm at a high speed-up factor. However, the detailed description of the results with different speed-up factors will help people better understand the algorithm.
(2) In sections 2.2 and 4.2, the authors propose the potential uncertainty in the sampling period. How does the noise in the signal affect the algorithm? Compared with the traditional method, how resistant is this algorithm to the noise? Will the change in pulse-to-pulse shape affect the algorithm?
(3) In Figure 8, for the x-label of the left figure, should it be tau_true?
(4) In section 4.3, the authors apply the algorithm to different subareas of the diode. Please consider including the explanation of unevenly distributed reconstructed points for the small area. What caused this huge distribution difference between 1x1 and 11x11?
(5) The authors summarize the results in the conclusion part. It would be very helpful if the authors could discuss more about the limitations and potential uncertainties of the algorithm. This will help people understand the application scenario of this method and the potential future directions.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 5 Report
Comments and Suggestions for AuthorsIn this paper, the authors study the periodic non-uniform sampling technique for 4 measuring thermographic transients, which increases the effective sampling rate by one order of 5 magnitude to 10 kHz resulting in a temporal resolution of 100 µs. The result in this work is right, I think the present work can be published.
Comments on the Quality of English LanguageMinor editing of English language required.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
