Cadence Monte Carlo---How to save sizing

[mordak]

Hi,

I am running Monte Carlo (MC) simulations for a big circuit in cadence and saw that mismatches would significantly affect the performance of the circuit. Now I want to figure out which part is more sensitive and redesign.

In ADEXl you cannot have direct plot to look at the voltage of each node (as far as I know), so I want to run a simulation in ADEL using sizing which was used in each MC run.

By the way, I am aware that you could include/exclude components in an MC run. However, I was wondering if I can have access to sizing of my components for each run of MC. For example, say a resistor has a value of 1 Kohm. I want to know what the value is for each run of the MC so I can apply those values and have a single run in ADEL using those values.

Thanks

 

[BigBoss]

This is a Pareto/Sensitivity Simulation problem.So the question : Which elements are dominant on the circuit performance.
If you search a bit in Cadence help, you'll get the right information about that..

 

[Dominik Przyborowski]

In ADEXl you cannot have direct plot to look at the voltage of each node (as far as I know)

You are able to print all DC voltage for each MC run and transient waveforms also.

What kind of circuit You have? High mismatches are very often caused by unbalanced amplifiers (i.e. when You are driving nfet by OTA with pfet current mirror in load)

 

[mordak]

This is a Pareto/Sensitivity Simulation problem.So the question : Which elements are dominant on the circuit performance.
If you search a bit in Cadence help, you'll get the right information about that..

You are able to print all DC voltage for each MC run and transient waveforms also.

What kind of circuit You have? High mismatches are very often caused by unbalanced amplifiers (i.e. when You are driving nfet by OTA with pfet current mirror in load)

Thanks guys for replying! I kinda have some idea which part of the circuit has more sensitivity to the mismatch. It is a big circuit I am running, contains DAC, amplifiers, ....
Maybe it is better to rephrase my question: Is there any way to print size of all components for each run of MC? I mean if nominal value of a resistor is 1 Kohm, I want to know what the value is in each MC run.

 

[Dominik Przyborowski]

If You open result browser You are able to plot values in dcOPInfo

 

[mordak]

If You open result browser You are able to plot values in dcOPInfo

Tnx Dominik. I am only running tran so I assume it would tranop in the result browser. For passive components I could see the sizing after mismatch was applied to each component, however, for transistors there was only operation points and parasitic caps ... and there was no sizing. By the way that would be tedious to go through the result browser to figure out sizes of all components, is there any way to print them out in a file?