Dear all users and forum administrators,
I am a new IC6.1 user and I use to work on IC5.1.41 for a long time. I feel so many unknown things happened to IC6.1 and some functions are missing in the ADE.
In the new ADE 6.1.3, I could not find a way to do Corner or Monte Carlo simulation with OA design kits. I wonder if it is the problem of tool or design kits and would like to ask for help to experienced users here.
How could you run corner simulation in the new Cadence? Any hints please?
In IC 6.1, Monte Carlo analysis has been moved to ADE XL. So you need to create a ADE XL view. Here is a brief set of instructions on how to do this for your existing ADE state setup. For for more complete set of instructions you can refer to the "ADE XL Users Guide".
From your Virtuoso schematic editor window,: select "Launch->ADE XL". Once the ADE XL window opens you can go to the "Data View" assistant panel and select "Click to Add Test" under "Tests".You will be asked to specify a the design to be used. Afterwards, the "Test Editor" window will appear. The Test Editor window is essentially the same as the ADE window that you are familiar with. From the Test Editor you can load your saved ADE states by selecting "Session->Load State". You can now close the Test Editor window and return to the main ADE XL window. You are now ready to simulate.
By default, ADE XL will be configured to run single point simulations. To enable Monte Carlo analysis, you will need to change the run mode: Look for the field in the top right corner of the ADE XL "Single Run, Sweeps and Corners". Change this field to "Monte Carlo Sampling". Select the "Simulation OPtions" button beside this field to adjust the number of points, etc. Finally,, select the run button to lbegin the Monte Carlo.
Thank you all, I found the option and simulation runs. It works fine.
In reply to Liangr:
LiangrThank you all, I found the option and simulation runs. It works fine.
In reply to MohE:
Just because you're a student doesn't mean that you don't have access to the documentation. It's included with the software. Use the Help menu at the top right of pretty much any Virtuoso window (e.g. the CIW, or the Schematic Window, or the ADE XL window) and do Help->Virtuoso Documentation - it will bring up the help system (cdnshelp). In ADE XL can also do Help->Contents and it will take you to the contents of the ADE XL manual.
Alternatively run "cdnshelp" from the UNIX command line.
Alternatively go to your installation of the IC tools and under the <ICinstDir>/doc directory there are a load of directories containing html and PDF documentation. For example <ICinstDir>/doc/adexl/adexl.pdf
In reply to Andrew Beckett:
In reply to dineshp:
I don't believe this is possible currently. Please contact customer support and request this (there's an enhancement CCR 984209 asking for this - if you request this a duplicate can be filed).
I am a beginer of the the simulation of monte carlo in ADE XL, I run the simulation as told, but it informs "Monte Carlo run stopped because no statistical data generated for the test", when I consult the guide book and search in the web, but I can not find the reason.Can you help me? I appreciate for any help.
In reply to dovestreet:
Please read the forum guidelines. Appending to an old thread from 2009 doesn't make a great deal of sense and is confusing.
You might want to take a look at this post.
Since you've not given much information here, and are using an unsupported version of the tools, is rather hard to know how to help you.
Thanks for your quick answer.I appreciate it. After reading the user guide, in the chapter of monte carlo analysis, I found the sentences of the description about stastistics for example (below), which are included in the model library. and I checked my model library (file.SCS) and I can not find these sentence and that means the model enclude the statistics variation. So I ask if these sentenses about stastistics is essential for the monte carlo simulation.
Thanks a lot
vary dist =
Yes, they are essential.
With the statistics block, spectre would be unable to produce any statistical variation of any parameters, and hence there would be no point running monte carlo - it needs to know the distribution of the parameters being varied so that the variation can be meaningful.