Why Transient Noise (Trannoise) Analysis for A Strong-arm Latch / Comparator?

[@sunkarasaigoutham]

excellent material. such high quality slides and content.

[@circuitimage]

Glad you like them!

[@ryanfu3016]

Sir you just saved my ass again

[@circuitimage]

Hi Ryan, I'm glad this helped you again. :)

[@abhiruplahiri1]

Good stuff CC, thanks for the video.

[@circuitimage]

Hi Abhirup, thank you so much for your feedback, and I'm glad you liked it. 😄

[@ying-rongsu8727]

Hi Dr, I’m curios about how to decide the “Fmax and Fmin” frequency range of the Tran noise simulation? Thank you!

[@circuitimage]

Hi Ying-Rong, nice to meet you. Happy New Year. 😊 Thank you for the great question. For the "Fmax" in the train noise, it's okay since that would be captured properly, but not the "Fmin," which required running a long simulation. To get the MHz range noise, you must run at the microsecond order sim, which is very time-consuming; therefore, I don't suggest running the train noise if you'd like to get the results quickly.

[@ItaiFinfter]

What is the minimum frequency needed to be set in the transient noise? Doesn't the clock act like a chopper for f

[@circuitimage]

Hi Itai, Nice to meet you. 😊 The minimum frequency needed to be set in the transient noise is the sampling rate or sampling frequency, fclk. In my case study, the full-rate clocking is 28GHz for 28Gbps data; therefore, 28GHz is a must to capture the noise at every sampled data point.

[@M.Taa11]

Hi Dr, can you please explain Peak transient noise voltage at regeneration time(nV) and how to measure it in the simulator

[@circuitimage]

Hi Mostafa, thank you so much for the good question. We could measure the Peak transient noise voltage at regeneration time(nV), but I'm not sure if that helps to get the input-referred noise since it's a statistic problem; therefore, we must follow the CDF (in the video) procedure to get the 1-sigma value based on the output wrong decision.😀