cst deembed s parameter

hi tungsin

i use the smith paper i should examine the robust method recommendation!

ok man
keep connect, i'm interest in this

i write a matlab code for extraction the permitivity and permeability based on smith paper. i upload it, if you write the matlab code of rebust method i hope to upload that that we both investigate the codes and solve their problems.
in this code first i enter the hfss output for s-parameters in : S11 and S21 in dB, phase in Rad
regards

sorry because of late replying, i'm on vacation.

i upload C-code which is based on Robust method, you can refer and let me know if you have any comment.

hi tungsin
i have to go to a trip for 14 days. after that i will run your code and tell you what i will see!
thanks you very much

In the retrieve method, I think there is a very improtant problem is the constituent of the structure.

There is a rather new but complicated retrieval technique in:

IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 56, NO. 6, JUNE 2008
"Determining the Effective Constitutive Parameters of Finite Periodic Structures: Photonic Crystals and Metamaterials"
http://www.ieeexplore.ieee.org/xpls/...nt=32&index=15

As far as I can see, the method can be applied to 3D metamaterials with SRRs and wires. I think it is a more robust method, but costly, you need to implement a GPOF code or something, (and of course you need more and bigger simulations)...

That's an interesting paper,

but I'm not very good in programming this kind of code.
I will be very interesting if someone can.

regards

piboo

As mentioned above, the Sij( phase) from CST shoulb be deembed in order to get correct results . Could anyone give more detail how to deembed Sij from CST? Thanks very much.

hi Zottiri Kusuta

please upload this paper.
regards

Hi all,
this topic is very useful for me
now I m doing this work (Retrieving the refractive index) follow smith's paper
SAJ25, I want to ask you for decribtion in your matlab code :
for gh=1:length(n)-1
n(gh)=n(gh)*sign(imag(alpha(gh)));
end
Thanks you very much

hi sangwoo
please check my matlab code and if it has any bug tell me as soon as possible.
regards

how about this matter?

i still meet a barrier in convergence of multi-layers case

dear all, any idea for this point?

Thanks for your discussion!
I am very interest with this field.

Group,

After inspecting the codes uploaded to this discussion, neither seem to implement the algorithms correctly. An important thing to consider here is the "branching." In the Smith code for example there is an inverse cosine function. There are an infinite number of answers here because

cos(a + 2*pi*m) = cos(a) m=...,-2,-1,0,1,2,...

The uploaded code assumes m=0, which is not a correct assumption.

In the "Robust" code, the branching problem is handled through continuity of the material functions. This feature is not present in the uploaded code and again the m=0 assumption was made. The algorithm also defines a procedure for determining the correct thickness of the slab that does not seem to be part of the uploaded code.

These codes are a great start, but they are incomplete and will likely only give correct results for thin slabs and at frequencies away from resonance.

Unfortunately, I have not written these algorithms myself, but I may have a need for them in a few months. If I do, and I get them working correctly :D, I will certainly share with the group. I actually hope somebody will make the modifications and share their code before then.

Thanks!
-Tip

Based on the idea in Part C of paper "Robust method to retrieve the constitutive effective parameters of metamaterials", i've got stable and continuous effective impedance just as shown in Fig.3. The structure i simulated is the microstrip with csrr etched in ground plane and gap in the strip---resonant type composite right/left-handed transmission line unit cell. Furthermore, from equation (6) and (4), the imaginary part of the refractive index n can be determined. Without considering the branching choice problem first, there exists data filp in the real part of n, espcially in the transmission zero (with small s21) and transmisson pole (with small s11). Up to now, i have not seen any paper has mentioned or studied such a phenomenon.Anyone has any idea?

hi Jian-Qiang Gong
i didn't see any paper for multi resonance structures! if you see that tell us here

Hi, SAJ25
What do you mean by multi resonance structures? Don't you mean the SRR/Wire structure studied most belongs to the single resonance structure? So why the retrieval method in the paper "Robust method" can only deal with the single resonance structure rather than the multi resonance structure?

i mean for example a S structure with multi resonance characteristics! how we can apply this method to such structure!?