Help with HFSS Lumped Port

stripline hfss

Hi,

I was wondering how does one feed a stripline with a lumped port in HFSS.

For microstrip, you would make a rectangle from the upper conducter line to the the ground and insert an integration line from bottom to top.

How does one do this for stripline when you have two ground planes (top and bottom) and the conductive line in the middle?

hfss stripline

Hi,

i dont actually know how can one simulate a stripline with lumped port.

However, you can define a waveport as defined in the manual or help. define a integration line from lower conductor to the strip. If you have 2 strips define 2 integration lines in the same direction and edit one of the sources to 180 degrees out of phase with respect to other if you are interested in odd mode propagation.

hope this helps

cst accurate for lumped port

you may try to define lumped port as ustrip line, from bottom ground to strip line.
And you need to modify the Excitation--> Edit Impedance Multi from 1 to 0.5.
It should work.

Regards,

hfss lumped port

Excuse me, why You use the HFSS to analyze a microstrip strucrure?! Use the AnSoft Designer and Your project will be to much simple!!!

Look at this: **broken link removed**

-->IMPORTANT<-- For the PASSWORD send me a private message

metamaterial ltcc

mesalem shalom,

In order to simulate a stripline in HFSS draw the lumped port to the closest ground. If the strip is in the middle, choose either grounds. Next, short between the two ground along the way of the strip. You can do this by modeling exactly the vias you have on board or instead use Perfect E "walls" on both sides.
Althrough close to the lumped port the fields will not be accurate, the natural stripline mode will propagate correctly away from the port.

If you want a more accurate simulation near the port use a waveport instead.

paolone1340 - Ansoft Designer can simulate striplines ofcourse. Two remarks about that:
1) You should specify in the excitation if this is a stripline or a grounded microstrip (by checking the use nearest ground as reference checkbox = grounded microstrip).
2) Remeber that Ansoft Designer ports assume only one mode. If the strip is not exactly between the two planes and the planes are not shorted (enough) and you want to measure exactly the mode conversion between the stripline mode and the parallel plate mode (between the two grounds) you will need to use HFSS (waveport with two modes).

I would use MoM for big planar models (in wavelenghts). Ansoft Designer MoM has a special solver for big models, so you can just throw the whole model inside and press the solve button.

Two classic "planar" examples where HFSS beats any MoM simulator are LTCC (where the volume of the model is in the order of its planar surface) and RFIC where the reference ground is far away from the strip, and solving inside the metal is important.

Hope this clarifies things.
Itai

modeling stripline with hfss

Hi Itafrenkel -- Please check your units. Not clear to me how you can compare the volume of an LTCC circuit to the surface area, units are different.

For LTCC, I would recommend both an MoM simultator and a volume mesh simulator. If there is a lot of metal (like many ground shield layers), volume mesh is a good idea. Also, for sensitive circuits that come close to the edge of the circuit, volume mesh is a great way to include the effect of the nearby change of dielelctric from ceramic to air. Results from time domain volume meshers (we represent CST) that I have seen tend to be a lot cleaner than HFSS results, look at the very error sensitive current distributions, but both types should be useful. In general, I would think it is most useful to compare volume meshers vs. MoM, rather than HFSS vs. MoM.

I'm sure HFSS has a lot of success stories with LTCC, but you may not be aware of the numerous success stories for Sonnet in particular, and MoM in general. One other tool I can comment on is Momentum, and they have a number of nice success stories too.

One Sonnet success story is EPCOS (spin off of Siemens) in Munich . They came from a standing start about four years ago to beat two very well established Japanese LTCC vendors for a major sole-source Nokia contract in just 2 years. EPCOS succeeded because they used Sonnet. The two Japanese companies failed because they did not use Sonnet. I can send you a Prismark report on that fairly astounding situation. The quad band cell phone from Motorola that I have in my pocket uses a diplexer that was designed by making extensive use of Sonnet.

I'm sure there are a lot of other success stories from both MoM and volume mesh vendors. Who beats whom is most certainly not a clear answer.

My bottom line advice: Don't play golf with one club and don't do EM analysis with one tool. I think perhaps you might agree.

I am now at MTT-S, will not be able to reply frequently.

hfss draw

I wonder: would it be wise to use MOM for the simulation of metamaterials. You know, there would be split ring resonators and wires.

hfss stripline, port

Hi irfan1 -- You can probably use MoM to do metamatrials, depending on the metamaterial. In most cases, I think a volume mesher would be more appropriate. Main thing to look for, if important to you, is being able to exicte it with an incident plane wave. This is something most volume meshers have.

If I were doing microwave design again (like I did for 8 years before getting into numerical EM software), I would want the following "dream team" to work a wide variety of problems:

Sonnet (of course!)
Momentum (for unshielded problems)
CST (time domain volume mesher)
HFSS (freq. domain volume mesher).

Those four would cover just about everything.

integration line lumped port

HFSS is useful to analyze additional loss due to thin ground planes (MS on silicon). TXLINE does not allow the ground thickness to be specified.

hfss mode conversion

hiall,
im also having confusion regarding the terminal/integration lines for CPW. there are two gnds...do i need to define two terminal lines?

ansoft designer edge port modes

Hi Venkat,

Yes you need to define two integration lines.....one each from ground to the strips...

hope this helps.....

hfss two port excitation stripline

hi,
I defined two lines. Z0 matches between simulated and textbook values. BUT when i see "Port Field display" I cud see 2 modes. what does this mean.? 2 modes are excited?. is that ok. I am very new to RF. Sorry if the ? is too stupid!!!