Negative gain of antenna

[ashish.mw]

Dear all,
I am in need of help urgently.
I simulated a stacked antenna with slots on the upper patch and feed is also given to the upper patch using HFSS.
On simulation, the S11 v/s freq results showed me quadband nature which i proved theoretically using MATLAB. Zin curves are also almost the same.
But when i saw the dB(GainTotal) v/s freq curve in HFSS, it showed me negative gain in dB at two of the four frequencies.
As far as i know, this signifies that the gain at these two resonating frequencies is less than the isotropic antenna. But some circles around me at my workplace told me that, "in the case of multiband antennas, only the frequencies with +ve dB gain are considered to be the useful resonating frequencies and rest are unuseful harmonics". Upto what limit the above is true. Kindly suggest me something as to what might be the reason for that, as usually Gain(dB) of an antenna should come positive.
Can there be any problem in my design that lead to this characteristic which is unacceptable?

Please help asap.

Thanks in advance.

 

[WimRFP]

From your description I can't imagine what you have exactly, but:

Gain = directivity - mismatch loss - 10*log(efficiency). All in dB.

Efficiency = radiated power / real input power : number between 0...1, efficiency = 0.5 means that 50% of real input power is converted into heat.
Mismatch loss = 10*log(1-|rc|^2) rc = reflection coefficient = S11

I don't follow the reasoning w.r.t. unuseful harmonics.

The low gain can be caused by wrong supersposition of the individual radiation patterns (phase and or polarization issue). If so, the directivity is low. If you have good directivity (directive gain), but gain is low, it must be a mismatch loss issue or efficiency issue.

Ar you looking to total gain or circular gain?

 

[pstuckey]

The first thing to consider is how much antenna gain you need to close your link. It's not uncommon for small antennas to have negative gain, but the system works over a certain range. And if it does not you can make some trade-offs between improving the antenna, transmit power, or receive sensitivity.

The next thing to consider is that your power amplifier is putting a certain amount of energy into your antenna, and if that energy is not radiating it is going somewhere else. Either heat because the antenna is lossy, or reflected back into your circuit because of poor match, or some of both. The VSWR can be tricky, because an antenna might have great VSWR in open space, but poor VSWR due to external objects. Sometimes people require a transmitter to be able to survive infinite VSWR because of that.

 

[ashish.mw]

Let me post the results. please give me few minutes to compile them.

 

[ashish.mw]

View attachment plots.rar

Please have a close look at my results.
Basically it is a quadband antenna. These are best possible results i have obtained and what i guess is that this includes best possible matched feed point.
please help me explain my Gain and directivity patterns

 

[drkirkby]

From your description I can't imagine what you have exactly, but:

Gain = directivity - mismatch loss - 10*log(efficiency). All in dB.

That's wrong. Gain has nothing to do with mismatch loss. It is defined by the IEEE in terms of the power absorbed by the antenna, so if the antenna is poorly matched, it will absorb less power than a well matched antenna, but the gain will not be effected.

There is also a term "Realized gain", which does take into account the impedance mis-match loss. I think there's another term which takes into account polarisation mismatch loss, though I don't recall the name of it.

There's an IEEE document on the definitions of antenna terms. I believe that is the most authorative reference, though others might disagree.

Dave

 

[ashish.mw]

Can anyone please help with my specific problem. I have already uploaded my result file. Please help me explain them, as I don't think it is the result of impedance mismatch, as this is the best possible case i obtained using parametric analysis in HFSS.

 

[WimRFP]

@drkirbly: To avoid confusion about nomenclature:

To avoid confusion about nomenclature:

If I have a directivity 15 dBi (for example from a 4 element stacked structure), |s11| = 0.5 and efficiency of 40% (40% of real input is radiated, 60% is converted to heat), the net gain is:

Gain = 15 dBi - -10*log(1-|s11|^2) - -10*log(0.5)

Gain = 15 - 1.25 - 3 = 10.8 dBi

10.8 dBi is what the user will experience. We can have a long discussion about the actual loss due to mismatch (as a PA doesn't behave as a 50 Ohms source mostly), but this is outside this topic.

Of course, one may add loss due to polarization.

---------- Post added at 12:43 ---------- Previous post was at 12:31 ----------

@ashish.mw: Each antenna has directions where directivity >= 0 dBi. You can have directions with directivity < 0 dBi, but this means there must be directions with directivity well abov 0 dBi.

So if you have good match and good efficiency but gain below < 0 dBi, look to the total radiation pattern and you should find directions with gain > 0 dBi.

regarding VSWR and minimum requried gain, this you should check with your client or project manager, as this depends on the transceiver and of course the link budget.

 

[drkirkby]

@drkirbly: To avoid confusion about nomenclature:

To avoid confusion about nomenclature:

If I have a directivity 15 dBi (for example from a 4 element stacked structure), |s11| = 0.5 and efficiency of 40% (40% of real input is radiated, 60% is converted to heat), the net gain is:

Gain = 15 dBi - -10*log(1-|s11|^2) - -10*log(0.5)

Gain = 15 - 1.25 - 3 = 10.8 dBi

That is quite simply wrong.

10.8 dBi is what the user will experience.

Agreed, but it is NOT the definition of gain.

Gain is defined by the IEEE 145-1983 "Standard Definitions of Terms for Antennas" as "The ratio of radiation intensity in a given direction, to the radiation intensity to that obtained if the power accepted by the antenna was radiated isotropically. "

Note the word "accepted" not "incident"

I agree what you are describing is what the user will probably experience, but what you are describing is the realized gain, not the absolute gain. The realized gain is defined as "The gain of an antenna reduced by the losses due to the mismatch of the antenna to a specified impedance"

Dave

 

[WimRFP]

@drkirkby: If follow your reasoning regarding nomenclature (accepted versus incident issue). It was the reason for me to mention efficiency and mismatch loss (and the example after your comment).

Regarding "the" gain definition, it is not just "right" or "wrong". Google the web, and you will see that many people don't follow the IEEE definition (especially in the EMC community where mismatch loss is included in the gain as stated by manufacturers).

 

[drkirkby]

@drkirkby: If follow your reasoning regarding nomenclature (accepted versus incident issue). It was the reason for me to mention efficiency and mismatch loss (and the example after your comment).

Regarding "the" gain definition, it is not just "right" or "wrong". Google the web, and you will see that many people don't follow the IEEE definition (especially in the EMC community where mismatch loss is included in the gain as stated by manufacturers).

Well, with this logic, the gain of the antenna would need to be specified with a specific load or source impedance. The gain of a half-wave dipole would no longer be 2.15 dBi, but somewhat less if a 50Ω systu em is used, or 2.15 if conjugated matched.

The definition I gave is used by the IEEE, it is used by Balanis, it is used in HFSS. In fact, HFSS has what you are describe, and calls it realized gain, just as the IEEE and myself do.

Dave

 

[WimRFP]

@drkirkby: You are correct with respect to the source or load impedance, it should be specified.

Especially wide band EMC antennas have varying VSWR across the band and you can't make a matching network that covers the whole band. It is therefore convenient to specify gain based on incident power (or maximum available generator power).

This document: www.conformity.com/PDFs/0712/0712_F05.pdf deals with EMC antennas. On page 43 (under Directivity and Gain), it mentions why they deviate from the Gain definition as used in IEEE.

 

[drkirkby]

@drkirkby: You are correct with respect to the source or load impedance, it should be specified.

Especially wide band EMC antennas have varying VSWR across the band and you can't make a matching network that covers the whole band. It is therefore convenient to specify gain based on incident power (or maximum available generator power).

This document: www.conformity.com/PDFs/0712/0712_F05.pdf deals with EMC antennas. On page 43 (under Directivity and Gain), it mentions why they deviate from the Gain definition as used in IEEE.

OK, so they admit the term they use is not consistent with the IEEE definition. Furthermore they say it is sometimes called "apparent gain". I know it is called "Realized gain" in the IEEE standard.

If, as you say, at least some in the EMC community chose to use a different definition to the IEEE one, I don't see any reason to follow them. In any case, this post is not about EMC.

I'm aware there are tons of terms in antennas that are badly used - mainly through ignorance. Personally I don't intend to follow that trend.

Dave