Ansoft - Differential feed

[gunajraman]

ansoft differential

Hi

I am feeding a patch dipole. I need to create a Differential feed.

How do I do that in Ansoft (My model has to use Coax feed)

Thanks.

 

[Marshal]

definition of a differential feed

I think we have the same problem.
I need to simulate such stracture too, and asked it in :

but I haven't gotten my answer yet. if you could get your answer please help me.

 

[svarun]

Can you look at some type of a Balun to achieve this ?

-svarun

 

[jallem]

In HFSS there is not concept of "ground". In other words there is not
plane with a "zero" potential. That is because HFSS works with the solutions
of the Maxwell equation and not a circuit equivalent.
That said, it means that a lumped port will be floating. So when you
feed two surface separated with a port the feeding is already differential.
You can try using two port in series connected to each other forming a floating
or virtual zero at the join. Also you feed both with V/2. But this process
is unnecessary. But if the structure has a common plane then I guess you can
use it as the reference, but be careful, if you put something to simulate a virtual
ground that in real life is not there then you may get unuseful results.

 

[jzhao2006]

In HFSS there is not concept of "ground". In other words there is not
plane with a "zero" potential. That is because HFSS works with the solutions
of the Maxwell equation and not a circuit equivalent.
That said, it means that a lumped port will be floating. So when you
feed two surface separated with a port the feeding is already differential.
You can try using two port in series connected to each other forming a floating
or virtual zero at the join. Also you feed both with V/2. But this process
is unnecessary. But if the structure has a common plane then I guess you can
use it as the reference, but be careful, if you put something to simulate a virtual
ground that in real life is not there then you may get unuseful results.

Let me try to continue the discussion and add what my understanding, meanwhile, I also have a question.

1) a voltage potential is the integration of E field from point A to point B. Under static field condition, the integration value has nothing to do with the integral path. However, at high frequency or dynamic field, this is not valid. With different integration path, we will get different value. We can use any value as the voltage, but we then need to stick with that path.
2) But people may still want to use the voltage definition at high frequency but with a very close distance between point A and point B. It is very similar to the measurement in the lab, we always use the probe with very close distance between the two terminals.
3) Now consider a transmission line, we can only measure the voltage between the two conductors, but never measure the voltage along one conductor.

Now the question,

When we define a port, we assume the current flow into one terminal will flow out from the other one, but under certain situations, that is not correct. Also how to excite high order modes?

 

[svarun]

a voltage potential is the integration of E field from point A to point B. Under static field condition, the integration value has nothing to do with the integral path. However, at high frequency or dynamic field, this is not valid. With different integration path, we will get different value. We can use any value as the voltage, but we then need to stick with that path.

Even at high frequencies, if the mode is a TEM mode, the voltage is still independent of the path. This is because the fields satisfy static-like equations.

-svarun

 

[gunajraman]

Exciting Higher order Modes can be specified when we assign the waveport.
And the cutoff frequencies for higher modes can be computed analytically.