I simulate a VCO in my design and export results to MATLAB using the cds_srr() function. Now, to account for mismatches, I want to run a Monte Carlo simulation (only 1 run).
1) Is there any change in the way I export the data into MATLAB after a MC run?
2) I do not want separate mismatch statistics. Just normal tran results with mismatches accounted for. Is this the final data that is stored? Or are they separate?
3) I did not select "Save data to allow family plots" thinking I have only 1 MC run and so it is not needed. Do I have to select it even for a single run?
If you are going to want to get the raw data into Matlab, you'll need to save the family data, because otherwise the results get discarded after each monte carlo interation (the scalar results are collected, but not the waveforms). You might find that you end up with a single run of data remaining, but you should not rely on that.
I don't understand your second question.
In reply to Andrew Beckett:
My second question: 2) I do not want separate mismatch statistics. Just normal tran results with mismatches accounted for. Is this the final data that is stored? Or are they separate?
I mean, the simulation first does a nominal sim and then followed by a monte carlo sim. So one would think that the two simulation data are saved separately. What I want from the raw data is to reflect the mismatches from the various devices in my design. So, is the normal transient simulation (without mismatch) saved separately? Or is the normal and Monte Carlo sim done separately by spectre and then combined and saved in a single file as a single data set?
Effectively, after I import the "data" in MATLAB, I should see a drop in the SNR etc of the VCO due to mismatches in devices. So how do I import this "data" with mismatch into MATLAB?
In reply to aliasnikhil:
First of all, are you running spectre standalone, or from within ADE? If so, which version of the IC tools are you using (Help->About will tell you).
The nominal results are separate from the monte carlo iterations, albeit in the same psf directory - but different database names.
If you can tell me how you're running and in which version, I can better give you a example of how to do this.
I have a similar situation. I am running spectre (13.1) from the command line and the various result files from the monte carlo are being put in the *.raw directory. But, viva 6.16 (also standalone) does not see the files. How do I access and plot the monte carlo results?
In reply to SteveL:
First of all, this appears to be nothing to do with cds_srr (which is the Matlab interface), so please do not append to an existing thread (see the forum guidelines, which is the first post in each forum).
Secondly, it's pretty hard to answer this without seeing the analysis statements in your netlist - I don't know how you've set it up. There's no reason why this should not work if things are configured correctly in the netlist.