Issue with Ansys2025 R1: different results in eigenmode and EPR analysis tutorial 4.02

[GuidoMenichetti]

Information

• Qiskit Metal version: 0.1.5 0.1.5
• Python version: 3.11
• Operating system: Windows

What is the current behavior?

We are encountering an issue while running the Qiskit Metal tutorial 4.0.2: Eigenmode and EPR analysis. Specifically, the results we obtain do not match those reported in the tutorial.

The following assignment in the code does not seem to update correctly:

eig_qb.sim.setup.vars.Lj = '11 nH'

When we check directly in Ansys Electronics Desktop (version 2025 R1), the value of Lj remains at '10 nH', the original default.

As a result, the 6th step value of eig_qb.sim.convergence_f is 6.20 rather than 6.25 as shown in the tutorial.

We suspect the issue might be related to the interface between Qiskit Metal and Ansys. We used the recommended Qiskit environment for the installation.

Steps to Reproduce:

Follow tutorial 4.0.2 as provided.

Attempt to set eig_qb.sim.setup.vars.Lj = '11 nH'.

Check the value of Lj inside Ansys — it remains '10 nH'.

Observe the different output in eig_qb.sim.convergence_f.

Expected Behavior:
The updated value for Lj should be correctly passed to Ansys, reflecting '11 nH', and the output should match the tutorial's results.

Could this be a compatibility issue with Ansys 2025 R1, or is there a recommended workaround?

Thanks in advance for your support!

Steps to reproduce the problem

What is the expected behavior?

Suggested solutions

[abhishekchak52]

This issue persists on the latest version (0.5.2.post4 and higher). Tested using Ansys 2025R2 on Linux (with ongoing win32com -> pyaedt translation).

Needs investigation.

EDIT: @GuidoMenichetti, could you check if this persists on your system?

[zlatko-minev]

I guess we can check the tutorials, but I wonder if some default options changed somewhere for sizesOn Dec 13, 2025, at 8:53 AM, Abhishek Chakraborty ***@***.***> wrote:abhishekchak52 left a comment (qiskit-community/qiskit-metal#1019) This issue persists on the latest version (0.5.2.post4 and higher). Tested using Ansys 2025R2 on Linux (with ongoing win32com -> pyaedt translation). image.png (view on web) image.png (view on web) Needs investigation. —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[abhishekchak52]

I think I know what's going on.

Lj is set as a project-level variable inside Ansys by EPRAnalysis, but the rendered junction port (which is processed inside QHFSSRenderer) only uses values from the QGeometry junction table (see here). So the junction inductance takes the value of "hfss_inductance" in that table (which is usually set by the qubit QComponent).

I don't think any of the EPRAnalysis setup variables override this value. Please correct me if I'm wrong @zlatko-minev