Request clarification on signal transmission on coupled paired microstrips

You've misunderstood that. A pair of odd mode signals still propagate together in the same direction. They just have opposite polarities.

'Odd mode' is RF jargon for 'differential'.

Jeroen Belleman

I have a project now where the customer insists on using a shielded twisted pair but the signal will not be well balanced. It's ugly. The even and odd modes both matter.

This is my model for the diff pair:

It takes three Spice txlines to model both modes.

In my case, the lines have to wind up 100 ohm differential, but there can be various common-mode impedances, depending on how the cable is actually constructed. One extreme is two separately shielded 50 ohm coaxes; the other extreme is a tight twisted pair with no shield except the uinverse.

Phil: this will be a bootstrapped photodiode at the end of a shielded twisted pair. I said it was ugly!

Right. Even mode is basically both wires connected together at the ends, working against the shield like a coax. Odd mode is a differential signal applied symmetrically to the wires of the pair.

They can have different impedances and different velocities.

They always do.

What, you design your differential signaling paths without consideration of common mode dynamics?

For shame.

Tim

Yikes. If you need me to yell at your customer, let me know. ;)

Cheers

Phil Hobbs

And they want it to be fast. It is very weird.

I can send you the sim privately, just for fun, if you are interested and will run it well between meals. You are NDA'd with these guys already.