Stats

  • Locked Locked
  • Replies 10
  • Subscribers 126
  • Views 27621
  • 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

How to transfer spice model to spectre model

archive

I got following spice model for external MOSFET from the vandor, how can I import this into spectre to get the right simulation results.

I added "simulator lang=spice" line at the top, but it doesn't give me the right result.

MODEL REMOVED BY MODERATOR BECAUSE VENDOR MODEL REQUEST PAGE MAKES IT CLEAR THAT THE MODEL MAY NOT BE DISTRIBUTED.

 

  • We probably need more information to give a good answer...

    Did you get any error messages from spectre? 

    How do you know the result isn't correct?

     

    One thing... Did you encase your model file in:

     

    simulator lang=spice

     <spice model file goes in here>

    simulator lang=spectre

     

    And include the model via a spectre include file.

     include "./<path_to>/FDD5612.scs"

     

    As an aside, if you want to include a spice subcircuit into a schematic, see How to Simulate a Subcircuit (Netlist) With Spectre in ADE

    /blogs/rf/archive/2009/01/07/tip-of-the-week-how-to-simulate-a-subcircuit-netlist-with-spectre-in-ade.aspx

     

     best regards,

    Tawna

     

  • Hi, Tawna,

    Thanks for your quick response. I followed your instruction here as bellow:

    Edit the model file to be encased by

    simulator lang=spice

     <spice model file goes in here>

    simulator lang=spectre

    Then, I go to “setup”è”simulation files”, and put the file in to “include path”.

    When I start the simulation, it give me the following error

    “Error found by spectre during circuit read-in: “input.scs” 26:M0 is an instance of an undefined model FDD5612.

    If I put the file in “Model Library”, it can recognize the model, but with simple boost circuit it doesn’t give me the expected result.

    It gives me lots of warning, and terminated prematurely due to error.

     

    Thanks,

    Wenbo

  •  Putting it in the "Include Path" field is bound not to work, since that is a list of directories that it looks for relative files specified in the model files, stimulus file or definition files lists.

    It really would help to see the warnings and errors you are getting - without that it's really hard to debug. I did see that there are some extremely small resistors (1u) and some of the sheet resistances in the MOSFET models are not good for spectre.

    Overall, knowing the errors and warnings is much more likely to allow us to debug it.

    Just had a thought - are you actually allowed to post the model on a public site? Most likely the model is owned by the vendor and you are breaking any license/NDA agreement you have by posting it here. Better would be to post a link to the vendor's site if it is available for public download.

    Regards,

    Andrew.

  • Andrew.

    Thanks for let me know about that license thing, I removed the file from the site.

    Please see attached output file, it has all the warning and error message.

    Thanks,
    Wenbo

  •  Wenbo,

    There was no attachment. To add attachments, you need to do it via the web site - you can't (I believe) just add them to an email reply to the forum.

    Regards,

    Andrew.

  • Andrew,

    Here is the file.

     

    Thanks,

     

    Wenbo

    spectre.zip

  • Hi Wenbo,

    A couple of things. You could try setting the global option minr to something smaller - say 0.1m . That should stop it filtering some of the small parasitic resistors inside the models.

    However, I believe some of the problem may be caused by the floating nodes in the model - it talks about some of the nodes being floating. I can't check now because of having deleted the model from your original post in order not to violate Fairchild Semiconductor's conditions of use.

    So, what I would suggest is:

    1. Get permission from Fairchild Semiconductor to send the model to Cadence Customer Support (or ask them to send it)
    2. Get the model and testcase showing the problem to Cadence Customer Support - http://support.cadence.com

    Then there's a chance of being able to debug it.

    The model was actually written for PSPICE, and as such takes advantage of some of the assumptions in PSPICE (which is more geared up for larger voltage off-chip circuits, whereas spectre is more geared up for typical IC problems). I see that Fairchild also provide BSIM3 based models for many of their transistors - but it appears not for this particular device.

    Best Regards,

    Andrew.

  • Hello Andrew,

     excuse me if I bother you with a problem that was presented and apparently solved many times, but Spectre can't read the spice model of the component BF862. I followed in detail the post 

    How to Simulate a Subcircuit (Netlist) With Spectre in ADE

    /blogs/rf/archive/2009/01/07/tip-of-the-week-how-to-simulate-a-subcircuit-netlist-with-spectre-in-ade.aspx 

     but when the simulation starts, this error message appears:

     Error (SFE-874) : ".../spice_BF862.prm" Unexpected end of line. Expected equals sign, numeric value or string value.

     

     The model of the device can be downloaded freely at

    http://www.nxp.com/products/rf/transistors/mosfet/jfets/n_channel_junction_field_effect_transistors_for_general_rf_applications/BF862.html 

     

    This is the file I used:

     

    [CONTENT REMOVED BY MODERATOR, AS THIS IS NXP's IP]

     

     

    Any suggestion would be really appreciated.

     

    Best regards

    Steve 

  • Steve,

    I removed the content of the model file that you posted above; it's NXP's model, so it should not be re-published on this forum.

    The fundamental problem is that it isn't really in SPICE syntax. First of all, there's no .SUBCKT header - I was expecting that there would be something like:

    .SUBCKT JBF862 1 2 3

    at the top. Secondly, there's a bunch of L, R and C components, but no instantiation of the model - it just defines the model. So no transistor instance. And finally it doesn't have a "." before the last line. I'm not that familiar with PSPICE (this may be a PSPICE model), despite it being a Cadence product, and so I tried using the new "pspice_include" statement that is in MMSIM13.1 to see if that helps. It didn't, because even then I would expect it to have a .SUBCKT statement. I suspect this is a file that can be read into Orcad Capture and then used within that environment (again, not my product area).

    So the best bet would probably be to contact customer support so we can explore this together in more detail; you might also need to contact the model vendor for some support.

    Kind Regards,

    Andrew.

  • Andrew,

     

    thank you for the fast response. Excuse me again: I will contact the customer support as soon as possible. 

     

    Kind regards,

     

    Steve 

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.