What model to take to simulate the probe feed using MoM?

[cybcad]

Question on MoM

i analyze the microstrip patch fix antenna using MoM, but i don't know what model should i take to simulate the probe feed? can someone give me a help?



thanks a lot!

 

[rautio]

Re: Question on MoM

If you are using SonnetLite, just put in a via port. To put in a via port, just put in a via to ground, then go to the ground level and put a port number on it.

 

[cybcad]

Re: Question on MoM

hi rautio
i am not using Sonnetlite, i am going to program by myself using Green's function in conjunction with MoM,but i don't know how to model the coaxial probe feed, could you give some advice?

 

[adel_48]

Question on MoM

Hi Cybcad,
Actually I have the same problem, I am simulating a fully metallic antenna with probe feed using a my own MoM code. For the excitation the easiest method is a monopole with delta gap voltage model. However, till now I am having a problem to model the feed accurately since the input impedance is strongly dependent on the feed model, the method I know, would be to incorporate FEM with the MoM by using what is called FE-BI technique, this would allow for the use of Waveports of FEM, while keeping the solution domain reasonably small. But I think it would make the code quite complex.

Best Regards,
Adel

 

[kleinesmurf69]

Re: Question on MoM

hello adel_48,

do you happen to know if the feed model as described in 'antenna and em modeling with matlab, sergey n. makarov" is accurate? i am thinking of using this book as a start to develop some MoM code in c++ using the matlab scripts accompanied with the book. if there is a better way of starting to learn programming numerical electromagnetics please let me know.

best regards.

 

[rautio]

Re: Question on MoM

If you are using Galerkin approach, you have to calculate the voltage induced on the via (and all other subsections). Any subsection that is not a port, set the voltage to zero. Any subsection that is a port (like the via), make the voltage = Vi, where Vi is the voltage on the ith port. You are then left with a Y or Z parameter matrix. Convert it to S-parameters.

All ports of this type include fringing fields around the port. Typically a couple tenths of a pF in capacitance. In most cases, this capacitacne must be removed. There several de-embedding algrorithms to do this, if needed.

The problem with wave ports is that if the medium is inhomogeneous (more than one dielctric), Zo is not unique.

 

[adel_48]

Question on MoM

Hi cybcad and rautio,

for cybcad,
Actually I knew that delta gap modeling is the worst modeling technique, but also the simplest. I compared the results with FDTD and it yielded a very good results, for lower frequencies <10GHz. The book of Makarov is a really good book, it has helped me a lot in making my own code. However, you will have to read some other books on MoM since the theory is not fully described in the book.

for rautio,
thank you very much for this information, but do you have any books or papers that describe the method or gives some more details? I would really appreciate that, as I am in a real need of modeling a coaxial feed.

BR
Adel_48

 

[cybcad]

Re: Question on MoM

hi adel_48
i have some papers on using MoM analyzing probe-fed microstrip antenna,the author taked the attachment model to present the probe currents.But the author didn't introduce which set of green's functions he taked,so i don't know how to embed the model in my code.
by the way,where can i find the book "antenna and em modeling with matlab, sergey n. makarov"?
i have uploaded two papers to the column IEEE Papers, Study Papers,you can download them from ,i hope these papers will be some help to you.

 

[rautio]

Re: Question on MoM

Hi Adel_48 -- The standard text is "Field Computation by Moment Methods" by Roger Harrington. Prof. Harrington taught me MoM in the early 1980's and was my PhD advisor. The book is very intense in linear algebra and assumes one or two semesters of EM theory.

Sonnet uses a rectangular via that is uniformly filled with current. If the horizontal area of the via is 1/2 the area of a rooftop basis function, it attaches to the rooftop basis function no problem. The via must go on one side or the other of the peak of the rooftop function.

The Green's function for shielded environment vias is most easily solved as a Sturm-Louville system as described by Balanis in his text. Shielded environment can be used for antennas as described in the SonnetLite manual. Shielded environment has the advantage that the Green's function is very easily calculated to full numerical precision, in sharp contrast to the unshielded Green's function.

 

[a_aziz]

Question on MoM

Hi.

could give me about your method of feeding.

 

[rautio]

Re: Question on MoM

Go to www.sonnetsoftware.com, download SonnetLite, install, and unlock. Should take about 15 - 20 minutes total. You probably ought to take another 45-90 minutes and go through the Sonnet tutorial, bring up SonnetLite, click on Help->Tutorial.

When you are finshed with that, bring up SonnetLite, click on Help-->Manuals-->Sonnet User's Guide. Look at chapter 19 on vias and chapter 21 on antennas. You have to do a little work to do antennas in Sonnet, but if you proceed as described in the manual, it should work out just fine.

You can feed the antenna from below with a via (to model a coax center conductor), or from the side with microstrip. When you want to use a via port to excite the antenna, just click on the port tool in the tool box, then click on the bottom end of the via. To use a microstrip feed, just run the microstrip feed line out to the edge of the substrate, click on the port tool, then click on the microstrip line where it touches the edge of the substrate.

I think you will be amazed at how sincerely simple it is.

 

[cybcad]

Re: Question on MoM

Hi a_aziz
i feed the microstrip antenna by coxial probe, i don't know how to model the feeding in MoM, and i want to use MoM to compute the input impedance of microstrip antenna.



THX

cybcad