Radiation efficiency of microstrip antennas

[bingjiang99]

Dear fellows,

I have a question about radiation efficiency of microstrip antennas. I know that surface wave contributes to the low radiation efficiency of microstrip antennas. I want to know if there are other factors affect the radiation efficiency. I noticed that different patch shapes have dramatically different radiation efficiency. Therefore, are there any ways to improve the radiation efficiency without changing the substrate material?

Thanks a lot,

 

[flatulent]

The efficiency is set by the substrate material and thickness. In your case you can make the substrate thicker and use aperture coupling through a slit in the ground plane.

 

[bingjiang99]

Dear flatulent,

Thank you very much for your reply. I agree with your opinions. By increasing the thickness of the substrate, I can increase the radiation efficiency. However, I can't use the group coupled idea due to the design restrains.

Besides these two methods, are there any methods that we can use? I noticed that different patch shapes shows quite different radiation efficiency for the same thickness. What caused it? As a result, what kind of shape should be the best shape with the highest radiation efficiency?

Thanks,

 

[flatulent]

On the feed, you can use coaxial lines to the antenna. The usual problem with thick substrates is that the transmission lines on the substrate will also radiate.

I have not seen any information on the efficiency of shapes. I suspect that the difference in the reports you see are caused by different thicknesses. Do not expect much more than 50% efficiency in the best of situations.

 

[bingjiang99]

Thank you.

I am doing a lot of simulations based on HFSS. I noticed the difference coming from the same thickness. One design gave me almost 50% radiation efficiency, but the other one only gave me 10% efficiency.

Regards,

 

[bingjiang99]

Did anyone have the similar experience? Another question, is the dielectric loss a bigger loss than the surface wave loss? I am not sure, although I believe that dielectric loss is not a big deal.

 

[tomcat_x]

The radiation efficiency is dependent on the antenna size and frequecy.

normally , the small size antenna has lower efficiency ,

the higher frequecy gets more efficiency .

 

[bingjiang99]

Do you have any proof for that?

For a dielectric material substrate, obviously smaller size substrate will give us higher efficiecy. Usually, I know that big size will give us higher gain.

By the way, in my design, I have two similar shape microstrip antennas, the smaller has higher efficency.

Thanks,

 

[jallem]

if you post the moles maybe we can help a little more.
can you post the models where the efficiency of the
smaller one is higher?

 

[savedadogs]

I know that if you place UC-PBG around the antenna it will attenuate the surface waves resulting in a 'cleaner' radiation pattern. Not sure if it will increase efficency though. There are several published papers on this topic.

 

[jallem]

Did anyone have the similar experience? Another question, is the dielectric loss a bigger loss than the surface wave loss? I am not sure, although I believe that dielectric loss is not a big deal.

I did two simulations of the same antenna. I used the same topology(shape) but not
the same size since the dielectric constant were different and adjusted so both
antennas were exactly tuned to the same frequency.
I used FR4 and Rogers 4350. Rogers gave me an efficiency of 94% while
FR-4 gave an efficiency of 67%, which represented a difference in gain of 1.5 dB.

 

[jian]

Hi, Jallem: Normally, larger electrical dimension for the thickness will yield more surface wave loss. When the thickness is large enough, the surface wave loss is very critical. Regards.

 

[bingjiang99]

Hello Jallem,

What is the thickness that you are using? For 62 mil FR4, I can only get 30 - 50% efficiency in 915 MHz band due to different shapes. Thicker substrate gives me better performance.

I can't upload the file due to the company policy. Sorry about it.

Thanks,

Did anyone have the similar experience? Another question, is the dielectric loss a bigger loss than the surface wave loss? I am not sure, although I believe that dielectric loss is not a big deal.

I did two simulations of the same antenna. I used the same topology(shape) but not
the same size since the dielectric constant were different and adjusted so both
antennas were exactly tuned to the same frequency.
I used FR4 and Rogers 4350. Rogers gave me an efficiency of 94% while
FR-4 gave an efficiency of 67%, which represented a difference in gain of 1.5 dB.

 

[jallem]

I designed for 2.4GHz. I was using 120 mils but I get the same
results for 60 mils. Only difference is the bandwidth. But, shape
is not relevant in this particular example. Just wanted to point the
difference due to loss tangent. I would guess for 900 MHz the size
you are using must be a "small antenna" and therefore the eficiency
is small. But make sure you got a good matching at the
design frequency.

Added after 5 minutes:

I had experimented with different shapes and some
are better than others but I did not attribute them
to the shape intrinsically but due to the fact that some
shape will radiate more near field than others. In
others words some are more prone to radiate less
near field. It is a conjeture but I dont have the way
to prove it.

 

[bingjiang99]

I know there is a bandwidth difference between different thickness. I also had experience different radiating efficency. I fabricated 60, 90, and 120 mil thickness antennas, the thicker substrate gives me the better performance.

I forgot to mention. The antennas that I designed are half wavelength antennas. So, they can't be treated as small antennas. The circularly polarized antennas give me bad performance, while the linearly polarized ones have better performance.

Your last point is very interesing. I will investigate it and let you know the results.

Best,

I designed for 2.4GHz. I was using 120 mils but I get the same
results for 60 mils. Only difference is the bandwidth. But, shape
is not relevant in this particular example. Just wanted to point the
difference due to loss tangent. I would guess for 900 MHz the size
you are using must be a "small antenna" and therefore the eficiency
is small. But make sure you got a good matching at the
design frequency.

Added after 5 minutes:

I had experimented with different shapes and some
are better than others but I did not attribute them
to the shape intrinsically but due to the fact that some
shape will radiate more near field than others. In
others words some are more prone to radiate less
near field. It is a conjeture but I dont have the way
to prove it.

 

[bingjiang99]

I simulated the antenna with low radiation efficiency, and investigated the near field. I used the HFSS software. I didn't see any stronger near field radiation.

By the way, I got lower radiation efficiency with truncated corner p a tch antenna. Does any one have some theoratic explanation?

Thanks,

 

[jian]

Hi, Bingjian99:

From what my experience, a typical resonance based antenna will store some energy in its proximity. The feed continuous provides energy to the patch with field distribution in the proximity while there is energy radiated out into the space. For a high Q antenna, the energy may not be stored in a very big volume. However, the energy density can be quite high. For a low Q or wide band antenna, the stored energy may not be very big while it needs a bigger space to store it. Patch antenna with cut-corners normally are for CP and their resonances are normally very narrow bandwidth. It means that much energy is stored in a very small space and the field intensity is very high here. Any dielectric or metallic loss in the material may cause significant loss of efficiency. I think my comments should explain what is happening in your case.

Best regards,

 

[bingjiang99]

Hello Jian,

Thank you very much for your answer. I noticed that there are a few vey high field intensity spots along the cutted pa tch, not distributed as a big area, which concides with your points. So in this case, we should say it is the dielectric loss, not the surface wave loss?

Another point is that it may not be relevant to the bandwidth. We can still get very high radiation efficiency with even narrower bandwidth.

Thanks,

Hi, Bingjian99:

From what my experience, a typical resonance based antenna will store some energy in its proximity. The feed continuous provides energy to the fix with field distribution in the proximity while there is energy radiated out into the space. For a high Q antenna, the energy may not be stored in a very big volume. However, the energy density can be quite high. For a low Q or wide band antenna, the stored energy may not be very big while it needs a bigger space to store it. fix antenna with cut-corners normally are for CP and their resonances are normally very narrow bandwidth. It means that much energy is stored in a very small space and the field intensity is very high here. Any dielectric or metallic loss in the material may cause significant loss of efficiency. I think my comments should explain what is happening in your case.

Best regards,

 

[jian]

Hi, Bingjian99:

Frankly, surface wave may be one of the most important sources for loss of efficiency of microstrip antennas with substrates of large horizontal sizes. Dielectric loss and metallic loss are also very critical for narrow bandwidth antennas. Everything is relative. It is possible that you can get higher efficiency for some even narrower bandwidth antenna if the material's loss is not so critical. Regards.

 

[kevinj]

I've made pcb antenns with Microstrip Feeds and have only simulated 40 to 50% efficiency.

Has anyone done much better?