A 3.2 GHz Injection-Locked Ring Oscillator-Based Phase-Locked-Loop for Clock Recovery

Round 1

Reviewer 1 Report

The proposed study describes a 3.2 GHz injection-locked ring oscillator-based phase locked loop that targets clock recovery for space use. Because there is no inductor in the architecture, the combination of the injection-locking technique and the usage of ring oscillators results in exceptionally low surface area and outstanding jitter characteristics. The use of an injection-locked ring oscillator in place of an LC oscillator, charge pump, and frequency divider in the PLL circuit allowed for a considerable reduction in chip size, which is the study's key contribution. The language used in the article is appropriate for engineers. There are several criticisms of the literature review, the presentation of the article, and the conclusions, though.

1.  Not all significant studies on injection-locking in ring oscillators are included in the literature review. For example, the well-known study "A Study of Injection Locking in Ring Oscillators" (Behzad Mesgarzadeh and Atila Alvandpour, 2005 IEEE International Symposium on Circuits and Systems, DOI: 10.1109/ISCAS.2005.1465873) is not included.

2.       In Table 2, unfortunately, the authors did not compare the proposed PLL with the best of the state-of-the-art by the widely used criterion of minimum figure of merit (FoM). For example, FoM = -244 dB in the study:

S. Ek, T. Pahlsson, C. Elgaard, "A 28-nm FD-SOI 115-fs Jitter PLL-Based LO System for 24–30-GHz Sliding-IF 5G Transceivers", IEEE Journal of Solid-State Circuits (V. 53, Is. 7, July 2018).

Readers would be interested in such a comparison.

3.       In Eq. 1 and in the text (lines 87 and 89), different symbols are used to designate the phase difference between the oscillator and the injected signal.

4.       In Eq. (3), Ii is the injected current, but in the text (lines 79 and 104), Iinj is the injected signal. Is there a difference between Ii and Iinj? Please clarify or correct.

5.       Line 104. The quantity Iosc is not defined.

6.       Equations (4) and (5) are missing. After Eq. (3) the next is Eq. (6).

7.       Equation (6) is well-known. Therefore, there should be a reference.

8.       Quality of Fig. 10 should be increased. On the vertical and horizontal axes, the numbers are poorly distinguishable.

9.        In the line 258, is written “Recall from 5.1, …”. What is “5.1”? Please clarify or correct.

10.   Line 370. The statement “As the comparison to the state-of-the-art showed, our work offers the lowest surface area occupied with a surface of 0.1mm²” is not correct, because as shown in Table 2, the achieved area is 0.09mm² in the references [16] and [19]. The normalized area is indeed the lowest, but not the actual area.

11.   Conclusions are weak and should be improved. Moreover, conclusions should not repeat the abstract.

 

Author Response

Hello,

Please see the attachment.

Thank you in advance,

Best Regards,

Author Response File: Author Response.pdf

Reviewer 2 Report

1. The critical literature review has been very poorly developed. It contains only 20 items on the subject. Needs to be improved. The oldest position in the literature is item no.: "R. Adler, « A study of locking phenomena in oscillators », Proc. IEEE, vol. 61, no 10, p. 1380‑1385, 1973, doi:  10.1109/PROC.1973.9292." From the point of view of the topic under discussion, it is acceptable. Please considerably expand the critical review of the literature with other items, but also issues related to the discussed topic, e.g. the influence of temperature on the generator's relative or absolute stability, how the value of the supply voltage affects the technical parameters of the generator, linear and non-linear corrections, etc.

2. Charts no. 10 a and b are of poor quality, data from the OY and OX axes cannot be read, they require improvement.

3. The conclusions drawn up in the article need improvement - they are too general. In total, that's 5 sentences for an article with a total of 13 pages.

4. Please assign a, b, c to Fig. 13, 14, respectively.

5. Whether PLL parameters were measured separately, if so, please provide the values of the holding and gripping ranges in the article.

Author Response

Hello,

Please see the attachment.

Thank you in advance.

Best Regards,

Author Response File: Author Response.pdf