microstrip patch antenna - IE3D Gain problem

[Neil_Preston]

IE3D Gain problem

Hi all,
I'm a student and as a part of my study I learn the IE3D program for antennas' simulation. I try a simulate an antenna with the following parameters:
microstrip patch antenna with L=W=46 mm, the probe feed is at (-17 mm,0) and there is also a shorting pin at (-22.6 mm,0) with a radius of 0.6mm. The patch lays on the substrate with dielectric constant of 4.3 and its height is 1.59 mm.
Now, from the manual of the program we know that
Gain=Directivity-Loss on Antenna-mismatch loss. The antenna loss includes the metal losses, the dielectric losses and the losses from the surface wave which decay as a square root of conductivity. In my simulation I've set the dielectric to be lossless and the conductivity value is 4.9e7. Therefore there should be no losses on antenna and the difference betweeen the gain and the directivity is only the mismatch loss.
However, if we take look at the simulation at the frequency point of 485 MHz, we can see that the gain is -18.1dBi and the directivity is 2.44 dBi. The return loss is -3.51 dB and thus mismatch loss is -2.56dB and we see that directivity minus mismatch loss doesn't equal gain.
I do not understand what's wrong with the simulation and would appreciate any help. I've also attached the geometry and simulation files.
Thanks,
Neil

 

[jian]

Re: IE3D Gain problem

Hi, Neil:

Your antenna is a modified inverted-F antenna. The probe and the shorting pin form a very small loop (compared to wavelength). It is causing very strong current concentration in that region. The strong current concentration there is causing much metallic loss even the sigma = 4.9e7 s/m. You will see the return loss is reduced significantly if you change the sigma to 1e+25 for the strip and the gnd. It is an indication that significant part of the loss is due to metallic loss for such type of antennas. Also, the rest of the loss is due to the surface wave.

best regards,

 

[Neil_Preston]

Re: IE3D Gain problem

Hi Jian,
Thanks for your answer. However, there are still things [/b]that I do not completely understand. I've done as you suggested and set the conductivity to be 1e25. I've got the following results: RL=-0.89 dB and thus mismatch loss==-7.32 dB; Directivity=2.44 dBi and Gain=-16.54dBi. We can clearly see that we loose less for the conductor loss. However, why my RL has become worse?
I also do not completely understand the mechnism of surface waves. If I check "Antenna Theory" by Balanis, I find that G=ecdD (it doesn't include the mismatch loss, but it doesn't matter here). The ecd means the losses due to dielctric and conductors. So, if the dielectric losses are zero, the surface wave losses is a mechanism that is connected to conductor losses?
My final question is:the antenna in my simulation (it is taken from the book "Broadband Microstrip Antennas") is a small antenna which size is less than λ/10 and its gain is very low. Is it possible to build a small antenna with Gain>-8dBi? Can you suggest some articles?
I've attached the files of my new simulation.
Thanks for your patience,
Neil

 

[jian]

Re: IE3D Gain problem

Hi, Neil:

Surface wave is some kind of wave propagating in the rho-direction in layered dielectric environment. Space wave is decaying as 1/r while surface wave is decaying as 1/rho. It is always there whenever there are at least 2 layers of dielectrics with different Er or Mr values. Vertical current such as vias and slots on large planes may create signficant surface wave. For the modified inverted F antenna, it is able to create a sharp resonance at some low frequency with the size of the antenna much smaller than half-wavelength. However, the resonance does not create signficant of radiation. Most of the energy will be absorbed as metallic and dielectric loss due to very high current concentration on the probe and the feed. The vertical current on the probe and feed will also create significant amount of surface wave loss. It is a good resonance while it does not radiate much.

Best regards,