Stats

  • Locked Locked
  • Replies 4
  • Subscribers 126
  • Views 16492
  • 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

Using multiple dynamic parameters when one is a variable

DouglasAH

Greetings,

I'm running a simulation that requires me to raise the temperature midway through the simulation. It is a precision circuit and for some reason the numerical noise gets quite bad when raising the temperature. Decreasing reltol from 10e-6 to 1e-6 solves the accuracy issue but of course slows down the cold part of my simulation considerably. I therefore thought I could use dynamic paramters for reltol as well as for the temperature

The simulation has to be run at several different maximum temperatures which is why I'd like to use a variable for the temperature. The GUI does not allow me to set multiple dynamic parameters however and setting up a paramset does not handle variables very well. Below is the paramset I'm trying and the netlister gives me an error:

"ERROR (SFE-874): "/home/dah/Desktop/paramset.scs" 3: 18: Unexpected open parenthesis "(". Cannot run the simulation because of syntax error. Correct the error and rerun the simulation

myset paramset {
time reltol temp
0 0.00001 VAR("minT")
150u 0.000001 VAR("maxT")
}

Obviously it's reacting to the parenthesis in the paramset. Is there a way around this limitation?

Don't know if it's relevant or not but it is an AMS simulation using the UNL netlister. Also, I'm using version IC6.1.7-64b.500.20

Best regards
Douglas

  • Dimitra Papazoglou

    Hi Douglas,

     I see that variables are not yet supported in the dynamic paramset. I'd advise you to submit a case and request a duplicate CCR of 1937419.

    On the other hand I see that VAR() syntax is supported in param vectors but this means you can change only 1 param (temp) dynamically (not both reltol and temp). So the below works:

    Also, I'd recommend you to use the errpreset parameter instead of reltol to control the accuracy. The errpreset values are liberal (least accurate bust fastest) , moderate, conservative (most accurate).

    Regards,
    Dimitra

  • DouglasAH
    in reply to Dimitra Papazoglou

    Hi Dimitra, thank you for the reply.

    I will submit a support case and hopefully this capability will be there in the future.

    Best regards
    Douglas

  • Andrew Beckett
    in reply to DouglasAH

    Hi Douglas,

    BTW, it wouldn't have been VAR("minT") in the Spectre file anyway - that's ADE syntax not Spectre syntax. It would have been just minT or maxT in the paramset definition. However, currently spectre does not support parameter references within the paramset definition - that's mentioned in the CCR that Dimitra mentioned. So it would need a fix to spectre, plus the transient analysis form (which does support paramset nowadays) being updated to allow parameter names to be mentioned on the form (that might well need use of VAR("minT") - which would get translated during netlisting to the native spectre syntax).

    Regards,

    Andrew.

  • DouglasAH
    in reply to Andrew Beckett

    Hi Andrew,

    Thanks for confirming the issue.

    In case it's of any interest I managed to work around the problem by using two different tests in ADE Assembler, each with its own parameter sets. It's workable for just two values of maxT but would be too clunky to extend to more.

    Best regards
    Douglas

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.