Home > Community > Forums > RF Design > Monte Carlo Problem with Swept PSS and PAC

Email

* Required Fields

Recipients email * (separate multiple addresses with commas)

Your name *

Your email *

Message *

Contact Us

* Required Fields
First Name *

Last Name *

Email *

Company / Institution *

Comments: *

 Monte Carlo Problem with Swept PSS and PAC 

Last post Sat, Dec 28 2013 2:41 AM by Andrew Beckett. 5 replies.
Started by vjain83 24 Dec 2013 01:55 PM. Topic has 5 replies and 5660 views
Page 1 of 1 (6 items)
Sort Posts:
  • Tue, Dec 24 2013 1:55 PM

    • vjain83
    • Not Ranked
    • Joined on Tue, Dec 24 2013
    • San Diego, CA
    • Posts 3
    • Points 60
    Monte Carlo Problem with Swept PSS and PAC Reply

    I am following the example in SpectreRF manual to simulate IP3 using swept PSS and PAC combination. I added  output-referred IP3 as an output, using the Direct Plot form. When I run a nominal simulation using ADE L or XL, the IP3 output expression evaluates without any problem. But when I run a Monte Carlo analysis in ADE XL, the IP3 expression doesn't evaluate, and gives an error:

     "Results pac are not available for ....../psf/mc1_separate/002. Use results() for a list of available results."

     I also have other expressions using pss results. These evaluate correctly when I run the Monte Carlo without a pac analysis. But, when I include the pac analysis, 'some' of the pss-based expressions also fail to evaluate.

    I searched on the Cadence support website etc, but couldn't find a solution that works. 

    • Post Points: 20
  • Tue, Dec 24 2013 11:14 PM

    Re: Monte Carlo Problem with Swept PSS and PAC Reply

    Theee things - which version of the simulator (should appear at the top of the spectre output log), which subversion of the IC tools (Help->About in the CIW), and can you post the expression you're using in the ADE XL Outputs here?

    I'll then check this in the next few days.

    Regards,

    Andrew 

    • Post Points: 20
  • Wed, Dec 25 2013 1:26 AM

    • vjain83
    • Not Ranked
    • Joined on Tue, Dec 24 2013
    • San Diego, CA
    • Posts 3
    • Points 60
    Re: Monte Carlo Problem with Swept PSS and PAC Reply

    Hi Andrew,

    Thanks for the quick reply.

    I am using IC6.1.5-64b.500.17 and spectre subversion 13.1.0.087.isr1.

    The output expression is:

    ipnVRI((v("/net054" ?result "pac") - v("/net027" ?result "pac")) '-2 '0 ?rport resultParam("PORT1:r" ?result "pac") ?epoint -50 ?measure "Output")

    Thanks,

    Vipul 

    • Post Points: 20
  • Sat, Dec 28 2013 12:32 AM

    Re: Monte Carlo Problem with Swept PSS and PAC Reply

    Hi Vipul,

    I'm not sure why that is failing. I tried a very similar expression with my example, using IC615 ISR17 and MMSIM131 ISR2 (so one later than you):

    ipnVRI((v("/outp" ?result "pac") - v("/outm" ?result "pac")) '-2 '0 ?rport resultParam("load:r" ?result "pac") ?epoint -35 ?measure "Output")

    And it worked fine in Monte Carlo...

    Maybe it was a bug in that ISR - I don't have so easy access to that specific ISR at the moment since I'm at home for the Christmas break. Seems a bit unlikely though, but I know there have been some issues with the  result separation stuff for monte carlo, so maybe that's it?

    Can you check in the CIW:

    envGetVal("adexl.monte" "enableMonteCarloSeparateDir")

    and:

    envGetVal("adexl.monte"  "savedatainseparatedir")

    and:

    envGetVal("adexl.monte" "useNewMCPostProcessing")

    to see what they return? I'm expecting all three to return "t".

    Regards,

    Andrew.

    • Post Points: 20
  • Sat, Dec 28 2013 1:12 AM

    • vjain83
    • Not Ranked
    • Joined on Tue, Dec 24 2013
    • San Diego, CA
    • Posts 3
    • Points 60
    Re: Monte Carlo Problem with Swept PSS and PAC Reply

    Hi Andrew,

    Thanks for looking into the issue.

    I checked, and all those envGetVal functions return 't'.

    I also looked in the monte carlo simulation results directory, and noticed that there is no 'pac' directory. The output log file also complains that it couldn't evaluate those expressions because there is no 'pac' result.

    Maybe I can try upgrading to the version that you used?

     Thanks,

    Vipul 

    • Post Points: 20
  • Sat, Dec 28 2013 2:41 AM

    Re: Monte Carlo Problem with Swept PSS and PAC Reply

    Hi Vipul,

    You probably wouldn't see them anyway unless you have the option to save family data (on the Monte Carlo options form) checked, as otherwise the result data gets removed after each iteration.

    You could try turning that on to see if it helps (it shouldn't make any difference though, but at least you'd be able to see if the PAC results are there after simulation).

    Maybe MMSIM131 ISR2 will help; not sure though.

    Regards,

    Andrew 

    • Post Points: 5
Page 1 of 1 (6 items)
Sort Posts:
Started by vjain83 at 24 Dec 2013 01:55 PM. Topic has 5 replies.