The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> Monte carlo simulations
https://designers-guide.org/forum/YaBB.pl?num=1254200527

Message started by raja.cedt on Sep 28^th, 2009, 10:02pm

Title: Monte carlo simulations
Post by raja.cedt on Sep 28^th, 2009, 10:02pm

hi,
i have the following basic doubts regarding Monte Carlo sims.
1.I heard that some model files contains both mismatch and corner variations, so as soon as you run Monte you will get both corners and Monte results. Is it true?

2.How to decide how many runs we have to do for Monte, what i do is run for 300 and 400, if sigma is converging and max value is equal to U+3*(sigma) then that is right point

3.What are the circuit techniques to handle Monte? Is there any reference available?

Thanks for your time
rajasekhar.

Title: Re: Monte carlo simulations
Post by Vladislav D on Oct 4^th, 2009, 9:54pm

1. Commonly, Corners and Monte Carlo are the different types of simulation. You can do Monte in different corners separately. However, as a rule, Monte is done only in the typical corner.
2. Number of run it is just the matter of accuracy. With 100 runs you accuracy is about 1%.
3.
a. As a rule, Vth mismatches decrease with 1/sqrt(W x L)
b. For the MOS transistors, mismatches are worse in a weak inversion
c. The stress sensitivity of Vbe or Vgs voltages is proportional to their charge carrier mobility, making PMOS or PNP transistors with symmetric topology pproximately 3 times less sensitive to the mechanical stress then NMOS or NPN ones.