Problem with simulation of CPW fed slot antennas in IE3D

[kbmanick]

hi,

i have designed one cpw fed slot antennas using ie3d. if i simulate with

1. 20 cells/wavelength i am getting -12db return loss

2. with 40 cell/ wavelength i am getting -50db. which one is correct?

frequency of operation 5.72 to 5.85 ghz.

 

[jian]

ie3d key

Hi, kbmanick:

Did you apply AEC, if you do not define the edge cells properly, the results may not converge well and you will see big difference with changing meshing settings. If you apply AEC properly, it should be converged very well for most cases. You can send me the .geo file (jian@zeland.com) and I can check it for you. Regards.

 

[kbmanick]

zeland ie3d lic file

dear jian,

i have sent u the mail with geo file. have u got it? i gave extension port.

 

[jian]

Re: clear my ie3d doubt

Hi, Bala:

I sent you the explanation and revised geometry files 5 minutes ago. By the way, I just noticed that the resonace of your structure is not due to the CPW fed slot. It is due to the finite ground. The finite ground's resonance is at about 5.5 GHz and the slot's resonance is at about 8 GHz. When you define so wide negative ports on the finite ground, it will certainly affect the results when you change the meshing. Regards.

 

[kbmanick]

Re: clear my ie3d doubt

RESPECTED SIR,

I AM GETTING SOME CLEAR IDEA TO USE IE3D. YOUR MANUAL BOOK IS SUPERB. WE ARE HAVING 10 VERSION ONLY. TILL NOW I AM NOT ABLE TO OPEN THE FILES U HAVE SENT TO ME( CPW SLOT).


I ALREADY SENTA MAIL TO GET THE EVALUATION KEY. I MENTIONED THE HOST ID ALSO. BUT I DID NOT GET ANY REPLY.

IF U FIND TIME SEND ME THE 10 VERSION SUPPORTED FILES.

WITH THANKS,

KBMANI.

 

[jian]

Re: clear my ie3d doubt

Hi, bala:

I am attaching the zip file containg the V10 files. I also included the magnetic current modeling of your CPW structure with infinite ground plane. You will see the resonace is pushed much higher and it is an indicationg that the resonance of your current antenna is due to the finite ground instead of the slot.

Best regards,

 

[kbmanick]

Re: clear my ie3d doubt

Respected Sir,

i have got the 12 version lic ( temp) key. it is really superb. i checked in 12 and 10 versions. my heartful thanks to u. but if u clear my following doubts, i may not disturb u in future...i mean i can understand very well. one small request. if u send any docs, plz send it to my personal id ( balamani_2001@rediffmail.com).


1. for my structure how u gave the highest frequency as 12 ghz? if i give 6 ghz why some deviations? what is the procedure?

2. how u calculate the automatic edge cell width. u gave .196891 and only for slot structure u gave .09844. how?

3.i could not get clear info from ur last reply. u mentioned that due to finite ground resonant freq is pushed to higher?

4. i gave infinte gnd plane only. but in fabrication we will give finite gnd plane. will it affect more? but in manual they gave one procedure.

z=1 layer given GaAs substrate.

z=0 with as usual data. shall we give any substrate in order to make finite gnd?


plz take ur own time and reply. u might be having some impotant work.



with thanks,

KBMani

 

[jian]

Re: clear my ie3d doubt

Hi, Bala:

1. The Highest Freq (Fmax) = 5.9 and the Cells per Wavelength (Ncells) = 40. For the same meshing density, you can change them to Fmax = 11.8 and Ncells = 20. That is the reason I just changed them to Fmax = 12 and Ncells = 20.

2. For the finite ground plane structure, I used the AEC RATIO = 0.1 and it makes the edge cell width to be 0.1968. In fact for AEC Ratio = 0.1, the center conductor is not meshed with edge cells, you may consider reduce the AEC RATIO or use multi-layer AECs.

For the infinite ground plane structure, I also use AEC RATIO = 0.1. However, the estimation program estimate the waveguide wavelength differently due to the difference in the dielectric configuraiton. The estimation of waveguide wavelength can be very accurate for multi-layer structure. Normally, you do need to change the Ncells and make sure the meshing is fine enough. For your structure, it is meshed quite fine and it should be ok.

3. When I model the slot antenna using infinite ground plane, the resonant frequency is 8 GHz and it is much higher than the 5.5 GHz predicted using the finite ground plane version. In the other word, the slot's resonance is at 8 GHz. The resonance happens in your finite ground structure is not due to the slot, it is due to the finite ground plane size. In some sense, your finite CPW ground is a very important part of your resonator.

4. In your original structure, you define an infinite ground at z = 0. Then, you define your CPW fed slot antenna on z = 3.175 mm. The CPW ground is finite.

Best regards,

 

[kbmanick]

Re: clear my ie3d doubt

Respected sir,


1. only one question how u calculated the AEC ratio? plz explain

2. so my structure is cpw with finite ground plane. now the resonance is 5.5 ghz. i think i can go for fabrication? but i did not get the infinite gnd plane structure.
if u send it, send it to my mail id.


with thanks,

KBMani.

 

[jian]

Re: clear my ie3d doubt

Hi, kbmanick:

1. The AEC Ratio is set by the users. Our suggested range is from 0.02 to 0.15.
2. The maximum cell size is based upon the estimated waveguide wavelength (WL). For microstrip structure, we normally can estimate WL very accurately. However, for multiple substrate structures, the estimation may not be very accurate. The AEC cell size is the product of AEC Ratio and maximum cell size. When the estimated WL is not precise, the estimated maximum cell size is not precise and the AEC cell size based upon AEC ratio is not precise. Anyway, the informationi is used in the meshing. It is not the final results not accurate. When the meshing density is good enough and you apply AEC, the results normally are accurate.
3. The infinite ground plane file is in the ver_final_5_r3_v10.geo in my previous upload.

Best regards,