The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> Monte Carlo Simulation Error
https://designers-guide.org/forum/YaBB.pl?num=1203976519

Message started by Upemba on Feb 25^th, 2008, 1:55pm

Title: Monte Carlo Simulation Error
Post by Upemba on Feb 25^th, 2008, 1:55pm

This is my first time to run Monte Carlo with Cadence tool, using Spectre for simulator and I am just trying to run a simple DC simulation and I am getting the following error:

Error found by spectre during Monte Carlo analysis 'mc1'.
mc1: Attempt to run Monte Carlo analysis with mismatch variations, but no mismatch variations were specified in statistics block.

Analysis 'mc1' terminated prematurely due to error.

Can someone helps me with the error?

Title: Re: Monte Carlo Simulation Error
Post by sheldon on Feb 26^th, 2008, 3:18am

Upemba,

Title: Re: Monte Carlo Simulation Error
Post by sheldon on Feb 26^th, 2008, 3:22am

Upemba,

There is an application note in the Modeling section that describes the
required setup for Monte Carlo simulation.

http://www.designers-guide.org/Modeling/montecarlo.pdf

When you perform Monte Carlo simulations, there are process variations, inter-die
variations, and mismatch, intra-die variations. You need to describe the mismatch
statistics in order to perform mismatch analysis. You also need to use in-line subckts,
see the paper for more details.

Best Regards,

Sheldon

Title: Re: Monte Carlo Simulation Error
Post by Upemba on Feb 26^th, 2008, 9:01am

Sheldon,

Thanks for taking the time to respond to my question. That was very insightful especially the pdf document that you pointed to.

From your answer, that makes me realize that the models that I am using do not have the statistics blocks required to run a Monte Carlo simulation where the mismatch and/or process variations are specified.

Also looking at the pdf document, I saw something that I would like to bring to your attention.
Equation 10 on page 6 of that document , I thought that epsilon should not include [1+ ] just based on expressions 8 and 9 deductions.
Basically, I am saying that

epsilon should be equal to [(XRPafac/sqrt (AB)) + XRSPbfac]x(XRSP - 1)

Please let me know in case I am missing something.