Learning Forum

[dharmatejap]

Hi,

How to set up waveport correctly for differential CPW? (Terminal solution)

I have total four conductors with outer two as GND and inner two as signal traces for differential signalling.

I selected two GND as reference when setting waveport. In the port set up dialog box, there is a Note about connecting the multiple reference conductors. I am not able to understand how to "connect multiple reference conductors in the plane of port".

I tried CPW example set up to my differential CPW setup, but I am getting warning messages that more conductors than terminals can affect the results.

Please help me resolve these issues.

[AncS]

Hello Did you try the example file from the 'Open Example' section?
Do not worry about warnings they won't affect the simulation results. Warnings are different from errors.
Please provide some screenshots of model setup using 'Image' option to understand your problem in better way.

[dharmatejap]

Attached image of my setup.
What is the correct way to set up wave port with this?
I am not sure how to follow the Note mentioned in port set-up box. "Multiple reference conductors.... connected in the plane of port"
Please help.
Thanks

[tyler.reid]

Make sure the outer edge of your waveport makes contact with the CPW grounds. For GSSG like you have, the ports should be reduced to contact the two outer grounds.

[dharmatejap]

I did that but it didn't help in removing the warning messages.

[AndyJP]

ok. i missed what you mean at first

[Praneeth]

May I know why you are using two separate traces for ground.
Please make sure that the both ground conductors are touching each other.
All the very best.

[dharmatejap]

My set up is GSSG.
I do want to connect two grounds but in wave port scenario how do I do that?
Is it possible with any boundary conditions on the face of waveports?
I tried connecting a metal across but that doesn't seem to work.

[AndyJP]

Probably, he wants o see how out of phase excitation works on "coplanar". I guess, it is a bad idea and can not be called a coplanar, rather just two slotlines (fed by a hybrid maybe?). So first, you need a model of a hybrid coupler.
That is not a coplanar line, first of all. But anyway, it is not my project, and everyone are free to try and see.
// in coplanar, the mode itself is driven by the central conductor. So it is not just two slots, but a hot coax wire with the jacket unfolded into two infinite planes connected together somewhere beyond physically, or virtually, it does not matter. therefore, in the wave port, you define just one integration line starting from only one hot strip. The only possible wave solution will automatically produce the opposite field vector in the paired slot. Or it may produce a microstrip mode in the solved model of the ground backed cpw, if matching is bad.

[Praneeth]

Please share the snapshot of how you have connected the ground traces as mentioned in this post.
Best regards.

[tyler.reid]

The outer edge of the port is treated as conductive for setting up the fields, so if the port touches both then the two outer metal strips would be at the same potential. In my experience a grounded coplanar waveguide structure with ground layer under the traces is more common, in which case there should be a metal ground layer underneath. The two ground lines are then connected to the ground plane with vias along their length. If there's nothing connecting them, then there is nothing to keep them at the same voltage.