hfss input current port

hi,
for finding impedance we need input current and voltage values, but how we can measure input current and voltage values???
that's my question too.

Actually you have two structures there. One is the antenna itself and
second is the transmission line. The transmission line is a parallel wires
line type. I modeled the transmission line once and I got anywhere from
70 ohms to 150 ohms depending of the separation and dielectric
constant(I used a PCB as medium). So, I think the best you can do is model
the transmission line and the antenna separated. The antenna feeding can be
done directly to the base of it using lumped port. And once you got the two
models agree on the impedance put all together and re-simulate.

Can anyone upload tutorial or training stuff for p.a.t.c.h. antenna design by HFSS?

Thanks in advance.

Lupin

Patch antenna tutorial attached.

For the antenna impedance problem, you can find the input reflection co-efficient and then knowing the length of the tramnsmission lines, you can find Zin.

-svarun

Marshal:

There are probably a couple of answers to this, depending on what kind of data you're looking for.

If you just want a "quick and somewhat dirty" calculation for the Zin of the antenna elements, you can probably do this with a lumped port between the antenna elements, neglecting the twin lead fed entirely. This gives you a characterization for the antenna elements that neglects possible coupling to the twin lead feed. It probably runs faster and requires a smaller simulation volume, though.

If you want to include coupling between the twin lead feed and the antenna, but still want input feed characteristics for the antenna alone, you should use a wave port (which might be a little tricky to set up if your antenna has a lower physical ground plane), and set a de-embedding reference plane to the point where the twin-lead connects to the antenna. This gives you a Zin for the antenna that also takes into account the feed-to-antenna coupling, which might be significant.

Finally, if you want the effect of both the antenna and a prescribed feed length, I would suggest a wave port located at the lower end of your workspace so that you also include a proper amount of finite twin-lead feed. I don't know if you're working with a lower ground plane or not. If you are, use the design distance. If you don't use a lower ground plane (i.e., this is in free space) then you better model a sufficient distance to the lower absorbing boundary according to the vendor specifications for such surfaces. You don't want absorbing boundaries eating up field lines that are supposed to close in upon themselves in the model.

You could use a lumped port at the bottom between the leads, but this would add an additional discontinuity to your results that might add unacceptable error to yoru results at high frequencies. The wave port will start with the understanding that you already have a TEM mode propagating up your twin-lead line.

Make sure that you make your wave port large enough to capture all the feed lines. Be sure to check out the field lines in the port to make sure that you don't have too many of them terminating on the sides of your port surface; that would mean you need to make the port area larger.

Hope this helps.

--Max

hi maxwellian;

thank you for your good reply, could you please upload a picture or draw model about using lumped port in this structure , I realy have not used lumped port until now and don't know any thing about it and hence can't understand well how should i use it in my structure.
thanks a lot.

Hi marshal:

I am not certain how this is accomplished in HFSS, but here is how I see it done in most EM simulators that provide lumped ports.

Usually, you can select two vertices on two different, but nearby metal bodies. A discrete port can usually be placed on an "ideal wire" that you create between these two points. Depending on the simulator, you can usually define that a discrete port is to be placed somewhere between the two vertices. Some simulators require you to either create one wire with a point in the middle; others require two wires that are connected together. In the middle of the wire, you can define a discrete port.

Be sure to define the port to have a self-impedance of 50 ohms (or whatever impedance system you want to use). It should monitor both current and voltage. Usually there are settings to control the discrete ports that you can set when you create it. If they can record either voltage or current, the other can be inferred through knowing the impedance of the discrete port. Once they have impedance and at least voltage, the S-paramters can be determined easily.

Look into the software documentation under ports (there should at least be a PDF of the users' guide or something) that will show you the process for defining discrete ports. I'm sorry I can't be more help on how to do this specifically with your software.

Best regards,

--Max

Hi Marshal

In the end of your project do you need to connect the element to 50 or 75 ohm?

if its 75 do like you have been told if its 50 you have to add balun to transform
the element impedance to 50 ohm. usually its a coax and this input easy to
simulate.

PL

hi plasma;

thank you for your guidance. i think you mean i can use a coaxe line (which is simpler to define port) and convert it to a balanced line with a balun . but
i want to know which is the simplest and most accerate method
i think the lumped port feeding is better as maxwellian said
but i don't know how should i define impedance line and calibration line and how should i choose the value of impedance .
regards.

Hi Marshal

look at https://www.edaboard.com/ftopic135962.html

and look at the pdf.