This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMK04832-SP: IBIS model not performing as expected

Part Number: LMK04832-SP

Hello,

I am using the LMK04832-SP to perform some simulations and I am receiving unexpected results using the CLKoutx, LVDS_CO model to a LVDS receiver. The waveform do not appear to show the expected waveform.

  • Do you have a schematic of the network to which the LMK04832-SP LVDS output is connected? At a bare minimum we need to know what network is connected to the LMK04832-SP, including some of the details of the receiver (i.e. does it have internal termination, its own common mode, etc), and the location of the voltage probes which produce the included graph.

    If there are any DC blocking capacitors between your source and load, have you ensured the simulation runs long enough to allow any average DC bias across the DC blocking caps to settle? IBIS is, in my experience, very bad at identifying the correct initial DC offsets for series capacitors.

  • There is only 2" transmission lines between the LMK and the LVDS receiver. The LVDS receiver has on internal termination (100 Ω). 

  • If you replace the receiver with a 100Ω ideal resistor, does the simulation still behave strangely?

  • This is what I receive with a 100Ω ideal resistor.

  • I tried setting this up on my machine (I used a microstrip instead of a stripline but I think it's inconsequential, the transmission line is close enough) and I see the following:

    I also tried with 200MHz and saw comparable performance (fast edges, slight ringing), nothing like what you're seeing.

    I have the LMK04832-SP ibis model configured so that one pin in CLKout0, other pin is CLKout0*, both pins using the LVDS_CO model.

    I do recall that we found an earlier issue in the ibis model for this device and uploaded a new version to the web at some point. I wonder if you have an outdated model? Can you try redownloading your model from ti.com and trying with the new copy of the model?

  • My apologies, it looks like CLKouto* was set as output instead of output inverted. Typically for me, setting one pin to a specific output sets the opposite.  

    Results are coming out as expected, Thank you Derek!

  • Great, I'll mark thread as closed.