Stats

  • Locked Locked
  • Replies 5
  • Subscribers 126
  • Views 5981
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Difference between PM frequency from stb and result fromcalculato？

zuiying

Hello，I am using stb to simulate my OTA. The schemetic is as follows.

The setting in stb and simulation results are as follws.

I want to know unity gain bandwidth. I direct plot the PM frequency and it tells me 6.899MHz.

But when I use calculator to calculate the the zero crossing point of the gain curve，it tells me 7.512MHz.

This two values is not very close. Which one is correct or more accurate? Thank you!

Parents

  • My virtuoso version and spectre version are as follows.

  • in reply to zuiying

    Both versions are very old, but I think the problem here is how the interpolation is done is different between how it's done in the simulator and how it's done using the method you chose. There's probably a high-degree of potential for interpolation error because you don't have many points in the waveform (40 points on a log axis between 100k and 10G. Maybe try simulating with (say) 40 points per decade and seeing the difference.

    I believe (from a discussion I had and a change request I had a few years ago, CCR 1033825) that the stb analysis does a more realistic interpolation than doing it in post-processing, and particularly bad was the phaseMargin function which was doing linear interpolation on the magnitude (rather than the log gain and log frequency). However, I've not checked detail exactly what stb analysis reports itself (note that you can get that number directly from the simulator - it's in the output database). With more points I'd expect them to be similar and converge.

    If I had the time I'd do more investigation... (sorry!)

    Andrew

Reply
  • in reply to zuiying

    Both versions are very old, but I think the problem here is how the interpolation is done is different between how it's done in the simulator and how it's done using the method you chose. There's probably a high-degree of potential for interpolation error because you don't have many points in the waveform (40 points on a log axis between 100k and 10G. Maybe try simulating with (say) 40 points per decade and seeing the difference.

    I believe (from a discussion I had and a change request I had a few years ago, CCR 1033825) that the stb analysis does a more realistic interpolation than doing it in post-processing, and particularly bad was the phaseMargin function which was doing linear interpolation on the magnitude (rather than the log gain and log frequency). However, I've not checked detail exactly what stb analysis reports itself (note that you can get that number directly from the simulator - it's in the output database). With more points I'd expect them to be similar and converge.

    If I had the time I'd do more investigation... (sorry!)

    Andrew

Children

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.