Signal Integrity Improvements

[grant.dasher]

Hello,

I have been using Fusion for a number of years but just started using it to work on high speed PCBs.  I haven't seen a recent post about the current state of issues in the last few years, so I am starting a new one with some suggestions given the current state (Oct 2024) of the product.

1) The new feature to color netclasses in layout is awesome.  It would be even better if it could be extended to color signals in schematic view as well (unless I'm missing something already?).

2) The inability to specify distance when laying out differential pairs without using design rule constraints is strange.  It should be possible to set width and spacing at routing time.  Ideally including the ability to calculate this automatically using microstrip/stripline/etc approximations based on the current layer and stackup information (but just specifying the width would be good enough).

3) The meander tool is unintuitive and often requires ignoring violations to work effectively in complicated situations.   At least better documenting that by ignoring violations you can handle more complex scenarios "by eye" would have saved me a lot of time trolling through the forums.

4) The  signal integrity extension is very helpful, but in ability to run "analyze" on multiple signals in batch is very annoying.  When routing, for example, an SRAM chip with a 16-bit wide address bus and a 16-bit wide data bus it would be very useful to batch run analyze of all those nets.

5) There should be a way to use the batch result from #4 and compare trace delay based on the simulation results within a signal group.  For example, being able to quickly look at the distribution of trace delays across all signals in a data bus and using that information instead of length/in addition to length in design rules checks would be very helpful as a sanity check.

6) With these these features, Fusion is starting to become close to useful for medium complexity RF layouts too.  I am hoping in the future Autodesk and Ansys figure out a way to expand their partnership into simulating S parameters of filter structures and other things that are often beyond the price point of non corporate customers or require complicated import/export workflows.  The fusion pricing model is much more reasonable for occasional use of these features.

[jorge_garcia]

Hi @grant.dasher,

I hope you're doing well. Thank you for taking the time to write these out. You'll soon see some major improvements on 2 and 3, so stay tuned. In regards to 6 we continue to deepen our partnership with Ansys so more features will be coming to the SI extension.

I'll share your comments with the development team so that they can be scheduled. 

Let me know if you have any other questions or concerns.

Best Regards,