openEMS simulation of the Loop Feed

[Yoshiyuki Takeyasu]

Hello,

I'm new to this list. I'd like to share my openEMS simulation of the Loop Feed. The point is that the radiation pattern is sensitive to the depth of the choke ring. Phase center of the feed is calculated also. Please refer to my article at:
http://www.terra.dti.ne.jp/~takeyasu/Loop_Feed_3_En.pdf

Hope this may help.

Yoshi Takeyasu
JA6XKQ

[NISHANTH D]

I saw your simulation results, and they look very close to what I’m trying to achieve. I’ve been working on a similar setup, but my results don’t match what I expect, and I suspect I might have some modeling errors.

If you have a moment, could you please take a quick look at my simulation setup and results to see if you notice any mistakes? I can share the model details, geometry, and plots for reference.

For the dimensions, I've followed Alex Petit's modified loop feed (OMMA6) design, and I suspect the error could be due to a poor electrical connection between the loop and the coax feed. However, I couldn't make a perfect connection. I am new to this software (Altair Feko )

Thanks a lot for your time and for sharing your results—it’s been really helpful to learn from them.

[Alex P]

Development and Implementation of the OM6AA design from all Copper to a simplified Aluminum Cake Pan fabrication.

Regards,

Alex Pettit

HLine3D

https://github.com/AP-HLine-3D/HLine3D

[b alex pettit jr]

" For the dimensions, I've followed Alex Pettit's modified loop feed (OMMA6) design, 

and I suspect the error could be due to a poor electrical connection between the loop and the coax feed. However, I couldn't make a perfect connection. "

==========================================================================================

Most of the aluminum cake pans use a nearly invisible  Hard Coat which is non conductive ..

Removing it is not trivial as you can't really See the difference between It and the Bare Aluminum,

ALL must be removed around the nut area or whatever connects the loop feed extender to the cake pan.

Small residual ridges allow the nut to float on the non conductive surface .. and make no or intermittent continuity ...( From Experience   )

Somewhat like This :

Easy Solution : purchase a Non Coated Cake Pan !

Alex Pettit

[NISHANTH D]

Thanks, Alex. I do have another problem making the actual feed itself. I have an N-female to N-female connector, which is only 34 mm long, and the center contact sits too shallow(~4mm) . Suppose one end is connected with the loop and the other, which is to be connected to another N male connector (N female--N male to SMA male --LNA) for which the stick-out should be 8mm. From all this, if I manage to construct the feed, I get only ~20mm loop height from pan(~inc. pan thickness). I've made a bad drawing myself *thought of connecting the outer ring of N male to the pan. Was this workable, or any other way to get around this problem?

[Yoshiyuki Takeyasu (JA6XKQ)]

Hi,

I have no experience using Feko. Since I cannot decipher the details of the model from the attached documents, the following are potential problem areas I've inferred from my experience with openEMS.

When connecting model components, you should overlap them rather than just making them touch.

You need to pay close attention to the overlap between the insulator and the conductor where the coaxial transmission line passes through the choke ring.

Is the cell meshing resolution sufficient to account for the loop element diameter, the inner conductor diameter and outer conductor thickness of the coaxial transmission line, and the thickness of the choke ring?

To check these points one by one, I recommend building the model sequentially instead of simulating the entire Loop Feed at once. Start with a model of just the loop element and the point source excitation, then add the coaxial transmission line, and finally, the choke ring.

Furthermore, before simulating the 1420 MHz model, I recommend calculating the OM6AA 1296 MHz model. As my simulation showed, the characteristics change with each dimension, so you should first verify the model's validity. The OM6AA simulation is ideal as a benchmark for this purpose.

Regards.

Yoshi Takeyasu

JA6XKQ

2025年8月14日木曜日 21:01:31 UTC+9 b alex pettit jr:

[b alex pettit jr]

How about This ?

[b alex pettit jr]

Before you model the very fine details,  I recommend you  Build Something  and  Validate Your Model 

Best Regards,

Alex Pettit

=================================================================================================

[b alex pettit jr]

  FWHM  my model  68 Dg   vs  your model  71.4 Dg

[NISHANTH D]

  Thank you for your response, and I apologize for the delay in my reply.  @JA6XKQ, thanks for your advice. Firstly, I've modeled the coaxial transmission line (for details: I've created a cylinder of 1e-3m and another cylinder of radius 6e-3m, whose inner region is dielectric and the outer face is PEC). Also I've checked the validity of this model which is good . Therefore No problems with this .However the coax transmission to loop is where making me confused . In the previous model that I've shared the loop edge touches the edge of the inner conductor of coax ,I suspect that due to this poor electrical contact(the edges does not meet properly) .So I may not have got the desired result. In the next step I've followed your advise, to overlap conductors.However feko raises an error due to invalid  mesh because during overlap the segment length to the radius becomes 10>radius/segment length (for valid model the ratio should be >10). I really struck with this problem, can you give your model images at the transition region .

[b alex pettit jr]

It was suggested that a 1/4 WL Sleeve Balun was required for a Loop Feed ..

I built this and found No Functional Difference ..

>>>>  Your Comments, Please <<<<

Regards,

Alex Pettit

=======================================================

[Yoshiyuki Takeyasu (JA6XKQ)]

For a given Ground Plane / Choke Ring, the Loop Element is in a symmetrical arrangement, which requires balanced feeding. Therefore, using a balun is ideal.

You wrote, "I built this and found No Functional Difference." Which characteristics did you evaluate to reach this conclusion? The property most affected by the presence or absence of a balun is the boresight axis symmetry of the radiation pattern.

Because your NEC2 model does not include a coaxial line, it simulates balanced feeding. As a result, the current distribution of the loop element and choke ring is axially symmetric, and your simulation results show an axially symmetric radiation pattern.

On the other hand, my openEMS simulation model includes a coaxial line, which results in unbalanced feeding. As a result, the radiation pattern is asymmetrical.

It is difficult to evaluate this asymmetry of the radiation pattern in a real-world operational setup. Similarly, it's unlikely that you'd be able to find (or feel) a performance difference between using a balun and not using one in terms of antenna gain or C/N (Carrier-to-Noise Ratio) of the received signal. However, the return loss characteristics (bandwidth) may vary due to the frequency characteristics of the balun.

However, if the application is for radio astronomy, which requires a good XPD (Cross Polarization Discrimination Ratio), the value of using a balun would be significant.

I prefer to "Keep it simple."

Regards.

Yoshi Takeyasu

JA6XKQ

2025年8月19日火曜日 13:41:24 UTC+9 b alex pettit jr:

[Yoshiyuki Takeyasu (JA6XKQ)]

to All,

I am concerned that discussing simulation techniques using specific simulation software is not in line with the purpose of this mailing list. Therefore, I would like to discuss simulation techniques through personal email exchanges. What do you think? I will continue to post about applications to radio astronomy.

Regards.

Yoshi Takeyasu

JA6XKQ

2025年8月19日火曜日 22:50:36 UTC+9 Yoshiyuki Takeyasu (JA6XKQ):

[b alex pettit jr]

Hello Yoshi,

Thank You for the reply ...

The only evaluation I performed was to use a loop with/without a Balun as the feed element for a disk Yagi  and acquire a drift scan over several hours

A different  number of samples was taken with each test. ( due to rain )

The noise VS pk HLine amplitudes are extremely close

I anticipated a significant difference ...

Thank You for the reply,

Alex Pettit

=====================================================