I am trying to simulate my extracted views but am currently having some difficulty creating the actual netlists.
I have a testbench, which instances the device under test. The testbench is a schematic view, whilst the device under test has both schematic and extracted views.
I have added the extracted view onto the switch list using 'envSetVal'. Then I have the following calls in my script:
- design( myLib myCell extracted)
However, this fails as the netlister says the testbench does not have an extracted view. I was under the impression that if there was not an extracted view available as with the testbench, then the next view on the switch list would be used for netlisting (in this case the schematic view). Therefore, this should be ok??
Is this the correct method to create extracted netlists, or am I missing something obvious? Do I have to define exactly what cells use what view?
(I should also point out that all my schematic netlisting sims etc work fine. It is just creating the extracted netlists that I am having difficulty with)
Thank you very much in advance. Any help would be greatly appreciated!
Are to trying to netlist the testbench schematic or the DUT schematic? From your description it is the DUT schematic that contains the extracted view.
If you want to create a netlist for the DUT schematic ensure that myLib/myCell in design( myLib myCell extracted) matches the DUT libName/cellName. In this case it is not necessary to set the switch list:
design( myDUTLib myDUTCell "extracted" )
If you want to create a netlist for testbench schematic which will include the extracted view for the DUT, then the design() call should be as follows:
design( myTESTBENCHLib myTESTBENCHCell "schematic" )
In reply to Jim McMahon:
Thanks kindly for your reply.
I managed to fix the problem and it was a rather silly mistake that I had made. My method was correct, however, in my haste to edit the switch list within my .cdsenv, I had actually editted the hspice switch list rather than the spectre switch list. Therefore, when running spectre, the extracted views were never included in the simulations.
Thanks again and I hope I can be of assitance in the future.
In reply to coco009:
Sounds like you haven't looked into the 'Hierarchical Configuration' approach to managing which representations are in use at netlist/simulation time. It'll make life a lot easier as you add more variants to the mix, e.g. 'sch_20090223' instead of 'schematic', 'av_RC_70C' instead of 'av_extracted', etc.
See the Cadence Hierarchy Editor User Guide for the details and/or chat with your Cadence Field contacts for quick ramp-up.
In reply to aplumb:
Anything that saves time is definately worth a look. I'll investigate this further when I get a chance.