Talk:Time-interleaved ADC

Review

The article is very well written and easy to understand, however just a suggestion to add more references from the literature and further study links can improve the quality of the article.

Thanks Muhammadrehan645 (talk) 12:54, 12 July 2024 (UTC)[reply]

Thank you a lot for your review and your suggestions! Paneappenasfornato (talk) 14:11, 12 July 2024 (UTC)[reply]

Use of Wikipedia page as reference

Reference no. 3 is a Wikipedia article. As per Wikipedia policy you can't cite Wikipedia articles. In this case, the link earlier in the line is good enough; an academic source on the topic (I am sure there are many) would be better as a reference.

Leaving that aside, very clearly written and easy to understand! Shivaprsd (talk) 13:18, 12 July 2024 (UTC)[reply]

Thank you a lot for your review and for the insight on the wikipedia reference, I have just changed it. Thank you again! Paneappenasfornato (talk) 14:11, 12 July 2024 (UTC)[reply]

Review n. 3

The article is well written and interesting. Maybe you could try adding more references, since there are some quite long sections without any reference. I fixed the equations in your article following the Wikipedia:Manual of Style/Mathematics. The main problem was that tall fractions should not be inserted as inline text, they don't look good as they change the distance between different text lines. Also, math subscripts should be in roman and not italic font if they don't refer to a variable as for ISO-80000 norm. Other than that, congratulations for this new and interesting article! AcBCDE (talk) 16:10, 12 July 2024 (UTC)[reply]

Thank you for your positive words and your valuable work on the article! Paneappenasfornato (talk) 07:25, 15 July 2024 (UTC)[reply]

Observations and suggestions for improvements

The following observations and suggestions for improvements were collected, following an expert review of the article within the Science, Technology, Society and Wikipedia course at the Politecnico di Milano, in July 2024.

In the section “Working principle, there is a mistake when referring to the input signal Vin being at frequency fclk/2, fclk being the reciprocal of Ts (the overall sampling period), thus 1/(2Ts). First of all, the signal depicted in figure is at frequency 1/(8Ts). Besides, if the signa was at frequency 1/(2Ts) it would not satisfy the Nyquist’s theorem.

Aside from the mismatch between channels, there are other 2 mechanisms that degrade the resolution of the TI converter with respect to the one of the sub ADCs. One is the interaction between the slices through the common reference voltages. This typically requires the implementation of powerful reference buffers inside each slice to mitigate the interaction. The second mechanism is due to the interaction through the inputs of the slices because of the reactive impedance at the input of the chip (e.g., inductance due to bonding wires). This interaction can be greatly alleviated by using an integrated input buffer, which drastically reduces the impedance at the input of the sub ADCs.

I found the section “Typical applications” overly long, being as long as the core of the topic. Aside from this, when discussing the Delta Sigma Modulator, it’s not correct that the bandwidth of the converter is established by the OSR, the loop order and the number of bit of the quantizers: the bandwidth is given by the signal, and it fixed by the decimator filter that follows the modulator. OSR, loop order and number of bits give the resolution of the converter, establishing the noise shaping property. Finally, the figure related to the “scheme of heterodyne receiver” does not correspond to the description in the paragraph. “RF amplifier” must be substituted by “LNA”, “IF Amplifier & Filter” with “IF Filter & Amplifier”, and the last two blocks (Audio Amplifier + Loudspeaker) must be an ADC.

--Aandurro (talk) 15:24, 29 August 2024 (UTC)[reply]

I found and corrected one spot where

T_{\mathrm {s} }

should have been

T_{\mathrm {c} }

. I renamed

T_{\mathrm {c} }

to

T_{\mathrm {clk} }

to be consistent with

f_{\mathrm {clk} }

. There are two clock cycles on each channel for each cycle of the input signal so I think the expressions are correct.

Do you have any suggested citations for information on performance degredation discussed in your second paragraph?

I haven't tried to address your 3rd paragraph. Best for me if you left your comments in small bites. ~Kvng (talk) 17:26, 3 September 2024 (UTC)[reply]