Stats

  • Locked Locked
  • Replies 4
  • Subscribers 164
  • Views 14711
  • Members are here 0
More Content
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

Creating ESPICE RLGC model from HSPICE

archive

Hi folks, I have an HSPICE W-element 2-conductor RLGC model that I want to use in SigXP. From my (limited) knowledge of formats, ESPICE uses the following:

R (freq dependent)
L
G (freq dependent)
C

The HSPICE file uses three values of R,L,G and C for each frequency step, an example of R is shown below for 300MHz.

+ 3.000000e+002 1.085640e+001 8.692304e+000 1.085640e+001

I need to map from these HSPICE values to ESPICE, but I am confused because the L values in the HSPICE model are frequency dependent.

Does anyone have any advice on this?

Is it possible in SigXP to just use the native HSPICE data in a .dml model with a "begin literal HSPICE" type statement?

Any help gratefully recieved!


Originally posted in cdnusers.org by chrishalford
  • archive

    Chris,

    Are you going to target the simulation to Tlsim or Hspice. If Hspice, just put it in a black box and you're done.

    If Tlsim, the tool actually does have and handle freq dependent L. Set a CutOff Freq in Preferences, put in a geometric t-line model, run a sim, and then take a look at ntl_rlgc.inc to see the format. Then imitate that with you're Hspice data.

    One of the old uProc Kits had an excel spreadsheet that would expand the Hspice-style RLGC into the tabular Espice, but I'm not sure where to find it these days. But it's not too hard to do the conversion.

    Hope that helps,
    Donald


    Originally posted in cdnusers.org by Donald Telian
  • archive

    Hi Donald, thanks for the info.

    I had already tried to make a model based on the ntl_rlgc.inc syntax, but it didn't work very well (compared to HSPICE) so I assumed that I had mapped the matrixes incorrectly. The HSPICE model is of the form:


    .MODEL rmodl sp N=2 SPACING=NONUNIFORM VALTYPE=REAL
    + INTERPOLATION=LINEAR EXTRAPOLATION=LINEAR
    + DC=( 1.019827e+001 8.692304e+000 1.019827e+001 )
    + DATA=(441,
    + 3.000000e+002 1.019827e+001 8.692304e+000 1.019827e+001
    + 3.982371e+002 1.021310e+001 8.693041e+000 1.021310e+001
    + 5.286426e+002 1.023032e+001 8.693893e+000 1.023032e+001
    +... frequency steps up to 20GHz


    This is for R, the same form then repeats for L,G and C. Each line is a frequency step from 300MHz. I mapped it as follows in the DML model:

    FREQUENCY=300
    RMATRIX
    1.019827e+001 8.692304e+000
    8.692304e+000 1.019827e+001
    LMATRIX
    1.88628E-05 1.49143E-05
    1.49143E-05 1.88628E-05
    GMATRIX
    2.28427E-11 -3.74074E-12
    -3.74074E-12 2.28427E-11
    CMATRIX
    6.81912E-11 -1.13359E-11
    -1.13359E-11 6.81912E-11

    Can you please clarify what the three values are in the HSPICE model following the frquency value? I believe HSPICE only shows the lower-triangular parts of the matrices, since RLGC matrices are symmetric?

    Chris





    Originally posted in cdnusers.org by chrishalford
  • archive

    Chris,

    That looks correct to me, although I can't see your original LGC data from your post. You may want to verify that the values in both tool's matrices are per meter (I think they are).

    Donald


    Originally posted in cdnusers.org by Donald Telian
  • archive

    Hi Donald, good point, I'll check.

    I presume the DML must contain L=[length in metres] in the file?

    Chris


    Originally posted in cdnusers.org by chrishalford
Cadence Guidelines

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.