Montecarlo Mismatch Parameter list

[cmos.Analogvala]

How to see values of various parameters (which are varies in statastic
block) of various devices taken in each run of "mismatch only"
montecarlo simulations ?

Marshnil

 

[Riad KACED]

Hi Marshnil,

What you are looking for is not available from ADE in IC5141. But if
you don't fear the command lines and have access to the latest
versions of MMSIM6.2, then it is possible. You just need to add the
following 3 lines into the mc statement of your input.scs (or the file
invoked from your runSimulation script):

savemismatchparams=yes
mismatchparamfile="mismatchParamFile"
mismatchscalarfile="mismatchDataFile"

If your Spectre version handles the above statement, then it would
create those files and the necessary data in it, i.e the value of each
parameter involved in mismatch variation, for each instance and for
each run.

For example, I'm simulation couple of transistors, whose statistical
section in the Spectre model file shows:

mismatch {
vary vthnmis dist=gauss std=1/1
vary toxnmis dist=gauss std=1/1
}

The "mismatch only" MC analysis damps the following results into
"mismatchParamFile" and "mismatchDataFile":

*** mismatchParamFile
vthnmis toxpmis
I1
I1.M2
I1.M1

*** mismatchDataFile
#1
-2.4191 -0.309083
0.215916 0.416531
0.275842 1.56505
#2
....

For more information, look at Cadence's SourceLink Solution #11418896
"how to get param files for monte carlo mismatch variations". You
could look at the following post as well:
http://www.designers-guide.org/Forum/YaBB.pl?num=1203667793

I have tried to find some information in the Spectre docs but nothing
returned, even in my MMSIM 7.1.0.037.isr2. One might ask Cadence to
get the docs updated.

Hope this help,
Regards,
Riad.

 

On Sunday, April 26, 2009 at 5:50:57 PM UTC+4:30, Riad KACED wrote:

Hi Marshnil,

What you are looking for is not available from ADE in IC5141. But if
you don't fear the command lines and have access to the latest
versions of MMSIM6.2, then it is possible. You just need to add the
following 3 lines into the mc statement of your input.scs (or the file
invoked from your runSimulation script):

savemismatchparams=yes
mismatchparamfile="mismatchParamFile"
mismatchscalarfile="mismatchDataFile"

If your Spectre version handles the above statement, then it would
create those files and the necessary data in it, i.e the value of each
parameter involved in mismatch variation, for each instance and for
each run.

For example, I'm simulation couple of transistors, whose statistical
section in the Spectre model file shows:

mismatch {
vary vthnmis dist=gauss std=1/1
vary toxnmis dist=gauss std=1/1
}

The "mismatch only" MC analysis damps the following results into
"mismatchParamFile" and "mismatchDataFile":

*** mismatchParamFile
vthnmis toxpmis
I1
I1.M2
I1.M1

*** mismatchDataFile
#1
-2.4191 -0.309083
0.215916 0.416531
0.275842 1.56505
#2
...

For more information, look at Cadence's SourceLink Solution #11418896
"how to get param files for monte carlo mismatch variations". You
could look at the following post as well:
http://www.designers-guide.org/Forum/YaBB.pl?num=1203667793

I have tried to find some information in the Spectre docs but nothing
returned, even in my MMSIM 7.1.0.037.isr2. One might ask Cadence to
get the docs updated.

Hope this help,
Regards,
Riad.

Hi Riad
I use same spectre model file. and I want to know how i can get the distribution percent in this model?