Stats

  • Locked Locked
  • Replies 5
  • Subscribers 125
  • Views 15390
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Hierarchy delimiter and path suddenly changes in PSF

BradW
Hey all, I'm still stuck running in the 5.1 stream and using WaveView. Recently all my ocean scripts stopped working when trying to plot/access certain device currents. When I looked into it, it appears that the paths have changed in the psf file. For example, I was using i(“/Iload/L0/PLUS”) and now the same data is accessed i(“Iload:L0:1”) I've seen this in the past but I've never figured out what I did or what happened to make the paths changed. Anyone have an idea? Right now I'm using aps as my simulator but I've used spectre and seen the same thing. When I use the "outputs" command it still lists the "/Iload/L0/PLUS" node as being available but when I go use the waveform it can't find it. Thanks for any help

  • Do you really mean WaveView (which is a Synopsys viewer), or Wavescan?

    Anyway, what you're seeing is that you're unable to use "schematic" names (any name starting with a / is treated as a schematic name, and you give each element in the hierarchical name by whatever it's called in the schematic), and are having to use "netlist" names (i.e. the names which actually appear in the netlist - which may get mapped if reserved words etc are used) including the simulator's hierarchy delimiter (which is ".").

    If you can't use the schematic names, it generally means that it can't find the map or amap directories in the netlist directory - these are used to do the schematic to netlist (and vice-versa) mapping. To find this mapping, it looks in the runObjFile in the psf directory, for an entry saying netlistDir. If the runObjFile isn't found, it simply looks for "../netlist" (I believe) from the psf directory - and looks for amap within there.

    So usually this is because the results have been copied elsewhere, or the simulation has been run standalone and the results stored somewhere other than the usual ADE structure, or the amap dir omitted when the netlist directory was copied. Something like that.

    We (as Application Engineers) often see this kind of problem because customers send us their OCEAN script and the netlist file, and nothing else - so we end up having to modify all the signal references to use netlist names to make a testcase work.

    Hope that helps to pin-point the problem.

    Regards,

    Andrew.

  • Thanks for the reply Andrew, Bad slip there, I meant WaveScan. I haven't been moving data around and the amap directory is just where the runObjFile says it should be. Your description sounds just like whats happening so it definitely sounds like a name mapping issue.

  • The only thing I can think of then is that the subckt you've got containing L0 is coming from a model file, rather than a schematic (perhaps something changed?) and so it doesn't have any details about a schematic name for L0 (because it's not in a schematic).

    Probably a wild guess, and unlikely...

    So best thing would be to contact Customer Support, ideally providing the netlist and psf directory - then we can investigate. Might need me to take a look to dig a little, so if you do this, mention to whoever picks up the service request that I'm happy to take a look at the data.

    Regards,

    Andrew.

  • Just for the record I found the issue. Commenting out the following ; (envOption ; 'userCmdLineOption "+rtsf -format psfbin" ; ) fixed the issue with aps.

  • Note that you should not use the userCmdLineOption to change the format. This is because the ADE netlister (and various things) need to know what format is being used, in order to do mapping appropriately.

    In IC5141 this should be set by:

    envSetVal("spectre.envOpts" "simOutputFormat" 'string "psfbin")

    In IC613/IC614 it is:

    envSetVal("spectre.outputs" "simOutputFormat" 'string "psfbin")

    In IC613/4 it is also settable on the Outputs->Save All form (as is the +rtsf flag).

    In IC5141, it's OK to put +rtsf on the userCmdLineOption field, because the netlister doesn't need to know about this.

    Regards,

    Andrew.

     

     

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.