Hyperynx 9.0 will give consistant results whatever using Icx pro engine or Eldo Engine.
Then when this HyperLynx 9.0 will be released?
We need the Eldo simulator to model connectors and other S-parameters stuff, but the results now seem to be not reliable...
If the model has too much non switching time for the stimulus period (the model is overclocked), you can see different results, as explained in this technote:
If it is not overclocked, and you look at the pulses in the standard (not eye diagram) operation you may see clues as to the differences.
Since there are many variables in any simulation, it is hard to know exactly what is going on based on the eye diagrams alone. If you open a service request the issue can be investigated in detail.
I hope this helps.
Thanks a lot - this was the case!! The model was overclocked.
But is it possible to make this simulation successful anyway? The model is Xilinx's Spartan-6 LVDS, which we want to use for 1Gbps transmission. The model has rising and falling waveforms developed for as long as 4ns ! Would it be a correct practise to edit the waveforms and shorten them?