help,mutual coupling about large array!!!!!

[gw_uestc]

hello,every one

how i evaluate the mutual coupling about large array,

and how can i simulate the array in HFSS or IE3D, if i think about the coupling

between array cells?

best regards!

 

[jian]

Hi, gw_uestc:

For IE3D, you can use the AIMS-II and GEMS-F solvers to solve large antenna arrays. AIMS-II is good for multiple port structures while GEMS-F can save more memory and time for single port structure if it is setup propperly. Both AIMS II and GEMS-F are iterative matrix solvers and you may need some fine tuning in the setting even there are default values. On the IE3D 11.5, you can use the IE3D FASTA scheme which is more general and it is good for multiple port structures. Regards.

 

[Lokter_Whatsen]

Dear gw_uestc,

In HFSS, you can simulate both a finite or infinite array. For finite arrays, you can set the HFSS to find the Array Factor, whereby it will use the results to find the desired finite array results. For an infinite array, you will have to set the Master-Slave boundaries on the unit cell itself. Unfortunately, I think other than plane 2D fields of the supposedly element in an array environment, you can't see any 3D radiation patterns. That is version 9.2, i don't know if version 10 has been improved. Do refer to your user manual.

Hope this helps,
Whatsen

 

[gw_uestc]

thanks a lot for both of your help !

and if i want to realize a square microstrip arrays with 64 cells,and what is the feed network ,i use chebyshev current distribution.

best regards!

 

[jian]

Hi, gw_uestc:

I think you may need to synthesis the feeding network. You can use some formula to design the chebychev current distribution for the feed network without considering the coupling between the patches and the feed network. However, if the coupling between the patches is strong, you may get significant more energy on the patch elements on the boundary of the array. If the coupling between the patches and the feed network is not very strong, you can consider the "separation" approach discussed in the Chapter 13 of IE3D user's manual. For the "sepration" approach, you simulate the patches without the feed network so that you can include all the coupling between the patches. Then, you simulate the feed network with all the coupling between the feed network included. Then, you can cascade them together to form the complete array. In this way, you are able to monitor the true power going into each element. Based upon it, you can adjust your feed network to achieve better agreement to the chebychev current distribution. Certainly, such an approach may still neglect the coupling between the "patches" and the "feed network". To include everything, you have to simulate it as a whole and you may not be able to find how good the current distribution matching chebychev curve. In such a case, you can only see how good the radiation pattern is and you need to adjust the dimensions by monitoring the change in the radiation pattern.

Best regards,