Stats

  • Locked Locked
  • Replies 4
  • Subscribers 63
  • Views 14771
  • 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

Problems using RFTline Lib

Daron

Does anyone have experience using the RFTline Lib in Cadence IC616?  I am trying to develop a multi-conductor transmission line model to implement in a high speed I/O design.  In the development of my model, I am simulating S-Parameter data across frequency and comparing the results with EM simulations in Sonnet.  So far my results are not agreeing, and I am not sure what I'm doing wrong.  

My goal is to characterize the the match as well as crosstalk coupling between lines.  I have used both mclin and ncline to model multi-conductor microstrip lines.  I have instantiated a stackup and defined the dielectrics and conductors as they will be physically implemented.  

The network I am modeling is illustrated in the figure below, with the exception that I am simulating 3 pairs instead of 2: 

I have set up a 6 conductor nclin and am simulating a 12 port network.  My test bench looks as such: 

The dimensions I defined are: length = 2000 mils, Ws = 2 mils, s = 2 mils, h = 2 mils, t = 0.7 mils, 2 <= Sc <= 10 mils.  These parameters correspond to the cross sectional figure above.  I am sweeping from 100M up to 60G, and varying Sc between 2, 4, 8 and 10 mils. 

The problem I am having is that when I increase Sc, the spacing between conductor pairs, my coupling performance degrades, which is opposite to what I think intuitively should happen.  As this spacing increases (approaching infinity), the performance of the line should resemble that of a single pair; however, my results do not yield this behavior.

Below are my S-parameter results, where S55 is the reverse gain (S11), S65 is the forward gain (S21), and S45 and S85 are the coupling from adjacent lines.  Notice how the line performance significantly degrades as coupled spacing increases:   

These results do not match my EM simulations, which yield an increase in performance as coupling space is increased.  Does anyone know of deficiencies with the RFTline lib or limitations to its use?  I have read through the library documentation, and I cannot find any clues that would point to this behavior.  Any input is greatly appreciated.

Cheers,

Daron 

  • Hi Daron,

    Please file a Case with Customer Support (http://support.cadence.com) .   This will need to be looked at in more detail by an application engineer. 

    Feel free to add me to the cc list on the Case you create.

     

    best regards,

    Tawna

     

  • Hi Daron, I also talked with R&D about this. They'd like to take a look at your testcase. I may have a workaround based on what you are reporting. But I'd need to see a testcase to verify. Again, the best way to do this is submit a case throught http://support.cadence.com and cc me on the Case. Best regards, Tawna

  • Hello


    I am not a user of this feature in Cadence suit, but perhaps I can give you some ideas. I will be following this discussion for sure , as it is interesting to my present design, and perhaps it will spare me the need to toggle between two CAD tools, and export S-parameter files.

    When cross talk is tested, the GROUND plane plays a very important role.There is chance that the EM simulation you did ,  GROUND plane was define differently than in the schematic you show. Your schematic seem to use perfect GROUND in both edges of the multi-line structure.

    On top of it, it is not clear to me if you define cross talk between differential pairs OR single ended traces. It looks like you look at single-ended traces.


    If I am correct in my impression, single-ended traces coupling is depended on how GROUND is defined. In differential pair, it should less matter.

    I suggest you test cross talk with PORTs connected in differential-in differential-out manner. see if the odd dependency on spacing still happen.

    Perhaps this will give some clue , but of course, maybe I am way off.

    Eliav

  • Tawna,  

    Thank you for your quick reply.  I am working on obtaining my Host ID to register with Customer Support.  I will notify you when I have completed this and will cc you on my query.  Your help is much appreciated. 

    Best, 

    Daron

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.