Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Problem with Monte-Carlo Analysis

Status
Not open for further replies.

Ravinder487

Full Member level 3
Joined
Jul 9, 2010
Messages
169
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,296
Location
Bangalore, India
Activity points
2,469
I did Monte-carlo analysis on Sense amplifier to find offset due to Vth mismatch.I've set options in Monte-carlo as shown

but after the completion of simulation I'm getting only one plot.
I was trying to plot Vth for each iteration during Mote-carlo analysis.Why I'm getting only one plot.Can someone figure out the mistake I was making!!!.Please help me its urgent
 

there are many to choose from monte carlo.
No.
There are only two categories. One is final trial's result of MC and the other is all trial's results of MC.
 

Attachments

  • aho.png
    aho.png
    111.6 KB · Views: 340
With the same settings I get all the plots I need. (ver 6.1.3)

93_1304847010.png


10_1304847132.png
 

Ability of Post Processing of Monte Carlo in Cadence ADE is largely dependent on version of ADE which you use.

So post processing using Result Browser directly is almost not dependent on version of ADE which you use.
 
Last edited:
when I'm plotting scalar data I'm getting histogram but when I've selected wave in place of scalar I'm getting only one curve,I don't know how to resolve this.

My work is to reduce offset of comparator by back-gate biasing.To simulate this I need to to do transient analysis and find Offset of comparator at specific time(at 1nS).So I need values of Vth at 1ns during each iteration and find maximum variation in Vth.

I hope I'd given enough information on the work I was intended to do!!
 

Thanks Pancho for your continues support.Right now I'm out of work and sorry I can't post screen shots right,but I will definitely post them tomorrow.Definitely I'm in need of your help.Thank you once again!!
 

This is screenshot of netlist and also attached netlist file below
Use "mc1_tran-montecarlo" in PSF not "tran-tran".

I don't have any idea on "infotimes"
Without using "infotimes", how can you get "vth" at specific time point ?
https://www.edaboard.com/threads/202730/

I can't find out any statement for this in your netlist.

Don't you have any statement for this in the followings ?
include "/home/ravi/Desktop/SCS/basic.scs"
include "./_graphical_stimuli.scs"
 
Last edited:
Thanks Pancho, how can I insert those statements into my netlist,If possible post some tutorials or screen shots.
 

(1) Are there "MN0.vth" in "mc1_tran-montecarlo" ?

(2) You plotted "MN0.vth" from "tran-tran" in first post. However I can't find out any statement for getting "vth" in your netlist.
How can you get "vth" ?

Don't you have any statement for this in the followings ?
include "/home/ravi/Desktop/SCS/basic.scs"
include "./_graphical_stimuli.scs"
 
Last edited:
1)I've plotted NM0's Vth from "mc1_tran-montecarlo" and I got 100 plots of Vth this time.
2)Using calculator I've opened results browser and then from "tran-tran" I've plotted Vth,that's the reason you couldn't see it i my netlist.
Don't you have any statement for this in the followings ?

No I don't have any statements as such.
 
Last edited:

2)Using calculator I've opened results browser and then from "tran-tran" I've plotted Vth,
that's the reason you couldn't see it i my netlist.
....................................................
No I don't have any statements as such.
That is very strange.

You have only following statement in your netlist.
saveOptions options save=allpub

You don't have any statement for saving "vth" in youe netlist.

Enter followings in Unix Command Prompt and read messages.
spectre -h save
spectre -h tran

1)I've plotted NM0's Vth from "mc1_tran-montecarlo" and I got 100 plots of Vth this time.
So, do you mean your problem can be resolved ?
 
Last edited:
My problem of plotting Vth curves for each iteration has got solved but I need to compare Vth values during each iteration at 1ns.please help me to do this!!!
 

I need to compare Vth values during each iteration at 1ns.
please help me to do this!!!
Surely read results of "spectre -h tran".
Use "infotimes".

Or use value() function for data of mc1_tran-montecarlo".
 
Hi all, I'm getting following error when I've chosen 'Process' in 'Analysis Variation'
mc1: Attempt to run Monte Carlo analysis with process variations, but no process variations were specified in statistics block. I don't get this error when I change it('Analysis Variation') to 'Mismatch' what could be possible mistake
 
Last edited:

Do you use Cadence gpdk90 ?
yes

statistics {
mismatch {
vary pvt_mc dist=gauss std=1/1
vary pu0_mc dist=gauss std=1/1
vary pltw_mc dist=gauss std=1/1
}
}
endsection stat_mis
yeah I've the same statements in my gpdk090_mos.scs.
Does this mean that there is no process variation?
I'm doing Monte-Carlo simulation on differential amplifier with CMFB. I want to know the variation in output common mode voltage due to PVT, is it sufficient to only opt for 'Process' in 'process Variation'.
As a general question when to chose different options?
 
Last edited:

I'm getting histogram plot if I choose 'Process' in 'process variation' drop menu. Does this suffice my requirement(plotting variation in output CM voltage )
I've three options(1.Process 2.Mismatch 3.Process&Mismatch) available in 'process variation' drop menu, I know their meaning but I don't know their application( as where to choose these different option)
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top