how to do deembedding in hfss?

hfss discontinuity de-embed

8O who can tell me if i'm right
1.simulate the structure with transmission lines added to both ports
2.use "into object" direction,define the "length" same with the transmission lines at both ports.
3.caculate
i wonder if what i have got is the result after deembedding?

deembedding ports

Yes,you got the deembedded result.
Deembedded only useful when you want to verify measurement result with your simulation result,because you must align the two reference plane.

deembed hfss

hi pigkiller and anyone who can advice

your step 3, "calculate" refers to...?

i cannot find any calculate button...

right now i am learning how to perform de-embedding on my structure, but unsure abt the subsequent steps after i enter my vector length for de-embed.

looking forward to your reply.

thanks !

deembedding in hfss

you can simply define deembedding in the port preferences.

Warning: the deembedding hfss uses is simply a subtraction of a line. In mathematical terms only a rotation of the reflection coefficient! (in a lossless case). Do not deembed over discontinuities.

D.

de embedding hfss

"you can simply define deembedding in the port preferences."

The port preferences you are refering to here, are they found in the 'Source/Boundary Manager' or the 'Post Process'?

de embedding hfss model

I believe the de-embedding option is in the "post process" tap under "waveport".

de-embedding hfss

I have seen an interesting test to determine the quality of an EM software's de-embedding.

Create a through transmission line; it doesn't matter if it is microstrip, stripline, or whatever, and the impedance of the line doesn't matter either. If fact, I think it's better if it's something non-standard, (not 50 ohms).

Set the de-embedding reference plane in from each port so that the reference planes will touch in the middle, effectively de-embedding out the entire problem.

If the EM solver is perfect and has excellent calibration, then |S11| is zero (minus infinity dB) and the through phase (Ang{S21}) is exactly zero.

The |S11| answers are quite telling. Try it any EM solver you like. The |S11| data will tell you something about the effective dynamic range of the simulator.

And the best thing about this test is that it doesn't depend on measurements--you know the right answer in advance.

--Max