Resonant vs. non-resonant antennas

[j.thalbach]

Hello,

What’s the greatest difference between resonant and non-resonant antennas? Are non-resonant ones better in performance?

Jo

 

[vfone]

In few words, a resonant antenna radiates more EM waves, when a non-resonant antenna radiates less...
If an antenna doesn't radiate EM waves, is useless.

 

[albbg]

Since, ideally (infinite Q), the resonance phenomenon happens only at one frequency, you could operate the resonant antenna at that frequency. In practice acting on some parameter you can wide the working range. Over this range the imaginary part of the impedance of the antenna will be quite close to 0, while the real part can be brought not too far from the ideal 50 ohm. The VSWR is then acceptable.
A non-resonant antenna in principle doesn't have the problem to be operated over a limited range of frequencies. The problem is that here the imaginary part of the impedance is different from zero and can be both positive or negative depending from the considered frequency, so in order to be able to transfer the power from the amplifier to the antenna a matching network is often required. In this case without a matching network the VSWR will be very high so the most part of the power will be reflected back.

The ability of the antenna to radiate more or less EM waves depends instead from its radiation resistance, that is related to its geometry.

 

[vfone]

If have a non-resonant antenna with huge VSWR, is possible that all the power to be reflected, and remain nothing to be radiated by the (best geometry) non-resonant antenna.
In this case a matching network will help only to make the transmitter's power amplifier happy, but the VSWR after the matching network (at antenna input) will be the same because the antenna is the same.
Non-resonant antennas are used only by compromise, and not as a desired first option.

 

[albbg]

I do not agree with you.
As you said a huge VSWR cause the most of the power reflected back to the TX. A matching network has the task to adapt the TX impedance with the load impedance allowing (ideally) the maximum power transfer.

In we have the general impedance of the antenna Z = R +jX, then resonant means X=0, while non-resonant means X <> 0. Then is R is reasonably close to 50 ohm you can use the resonant antenna "as is", while your matching network will be, simply Zmatch = -X. You can see this as an "artificial" resonance due to lumped component instead of distribuite as in the case of a resonant antenna. The EM field will depends only by the radiation resistance (ideally because in real world you will have losses) regardless the antenna itself is resonant or not.

 

[volker@muehlhaus]

In this case a matching network will help only to make the transmitter's power amplifier happy, but the VSWR after the matching network (at antenna input) will be the same because the antenna is the same.

I don't agree.

If the matching network is low loss, the transmitter power will
- not be reflected
- not be dissipated in the matching network
- be radiated as desired

I'm often designing such "short" antennas with very reasonable overall radiation efficiency. You can't always use full size antennas.

 

[FvM]

I want to question the given definitions of resonant versus non-resonant. I think it's reasonable to equate "non-resonant" with wideband or frequency independent antenna. It can be operated over a wider frequency range without tuning the matching network. A LPA is a perfect non-resonant antenna in this sense.

An electrical small antenna with an almost reactive impedance has no resonance on its own. However, in combination with a low loss matching network, it exposes a high Q resonance with much smaller bandwidth than a classical resonant antenna like a lambda/2 dipole.

If performance refers to efficiency at a single frequency, there's no principle difference between the discused antenna types, all can be theoretically lossless and achieve good VSWR. In practice, electrical small antennas with low radiation resistance have difficulties even at a single frequency.

 

[j.thalbach]

As far as I understand the following comparisons may be considered:

1. Resonant antennas are band limited; non-resonant antennas are wide band
2. Only non-resonant antennas need matching networks
3. Resonant antennas may have greater antenna gain
4. Resonant antennas are more traditional; non-resonant antennas are newer technology

Are all the points correct?

 

[volker@muehlhaus]

I don't agree on most of your points, but maybe that is because your definition of "resonant" is too fuzzy. You really need to define your subject more clearly.

Hertzian dipole is an example of an antenna that does not exhibit a specific current resonance pattern.
A resonant antenna can have an input impedance that requires a matching network.
A resonant antenna can be inefficient with low gain.
A biconical antenna is broadband and known for a long time - would that be a "resonant" or "nonresonant" antenna for you?

 

[vfone]

Can design short, or long, or whatever length antenna, but is hard to say that a non resonant antenna has reasonable efficiency. Not even close.
Years ago in amateur radio bands, using a trans-match I tried my screwdriver as an antenna. Surprisingly I made few radio contacts. The "antenna" was reasonable good? No, was a piece of crap.

A resonant antenna on the desired frequency, have many chances to get the required input impedance (without using any matching networks), when a non-resonant antenna has almost zero chances to get the same required input impedance (as the impedance of the resonant antenna).
When you do comparisons have to compare the SAME antenna, when is working on resonant frequencies and when is working on non-resonant frequencies.
An then, you can see that the same antenna have better gain when is resonant, have better input impedance when is resonant, and have better efficiency when is resonant.

Maybe have to define what means "reasonable" antenna efficiency. Reasonable efficiency compared to what? To the efficiency of the first antenna invented by Heinrich Hertz?

 

[volker@muehlhaus]

Maybe have to define what means "reasonable" antenna efficiency. Reasonable efficiency compared to what?

Radiation efficiency is clearly defined: total radiated power / power into the antenna
where power into the antenna is incident power - reflected power

Over the past years, I have designed many antennas for smart home automation that use a short radiating element (not natural resonance). Overall radiation efficiency, including losses in the matching network, was usually 70-80% or better.

But I agree, you need a reasonable input impedance into the antenna. If it is too extreme, somewhere near total reflection, you are lost and losses in the matching network lower total efficiency too much. Even for a non-resonant antenna, you need a reasonable impedance matching ratio. I really like the antenna matching approach by Optennilab, who optimize matching networks for best total efficiency instead of staring at input matching only.

 

[vfone]

I never use Optennilab but the same, when I tune an antenna, first I do this for best antenna efficiency, and later check how input impedance looks like.
Tuning for best efficiency have big chances to get acceptable VSWR, but tuning for best VSWR do not assure an acceptable antenna efficiency.

To get an antenna efficiency of 80% or better from a non-resonant, electrically small antenna, is a miracle. The greatest concern in these days is the compromise between antenna efficiency and reduction in antenna size.

 

[albbg]

non-resonant antenna does't necessary mean "electrically small".

 

[vfone]

non-resonant antenna does't necessary mean "electrically small".

Read (before) with attention all the posts:

"I have designed many antennas for smart home automation that use a short radiating element (not natural resonance)."

 

[albbg]

Read (before) with attention all the posts:

"I have designed many antennas for smart home automation that use a short radiating element (not natural resonance)."

Sorry, you are right, but please give me just a reason why volker@muehlhaus shouldn't be telling the truth when he said "Overall radiation efficiency, including losses in the matching network, was usually 70-80% or better". Have you got any mathematical evidence from the antenna theory the efficiency should be much lower ? Or are you just guessing ?

 

[FvM]

As mentioned before, there's no theoretical reason why electrical small antennas should have low efficiency. But there are practical problems. Antenna theory shows that the matching Q increases while the ratio of antenna size to wavelength decreases. Respectively it becomes more sensitive to hand effect. For the same reason, I think it's a bit misleading to classify electrical small antennas as non-resonant, see post #7.

The second practical problem even with optimal tuned matching network is to avoid losses when the radiation resistance falls below 1 ohm, which easily happens with respective small antennas.

 

[volker@muehlhaus]

I never use Optennilab but the same, when I tune an antenna, first I do this for best antenna efficiency, and later check how input impedance looks like.

Sure. My comment on Optennilab is that they use overall efficiency (including matching network losses) as a measure for the best matching betwork, so they maximize radiated power for an acceptable matching level, instead of trying to go for best return loss (with possibly less radiated power because your loss in matching increases).

To get an antenna efficiency of 80% or better from a non-resonant, electrically small antenna, is a miracle.

That's why my clients love my work - results both simulated and verified by measurement in the antenna lab.
But maybe your misunderstanding is that you mean "wideband", whereas I mean a narrowband antenna with a shorted radiator that is operated outside natural resonance.

 

[E Kafeman]

This is a bit AD for my self but think it is relevant. I develop antennas, mostly embedded cellphone antennas, but do also develop software for controlling instruments used in combination of RF measurement ranges, typical turntables calltesters and VNA.
When I measure a wide band antenna efficiency during its development do I separate Rrad from Rloss and matching losses with equal or better result then for an anechoic chamber or similar. Have it confirmed in several ways.
Do not even need an anechoic chamber, it is enough with a shielded box and a VNA and measurement can be done at lab bench.
I am using a modified version of a method known as "Wheelers Cap".
It is very practical in many ways to do this kind of measurement live at lab-bench.
It is a real live measurement as result is updated as fast as VNA is able to sweep actual frequency range.
If there for example is a dip in any frequency band can it easily be investigated by add/remove small pieces of CU tape from antenna structure or check if a nearby speaker or something else causes the problem.
I am doing a simple demo, tuning of an antenna and then measure its real radiating efficiency here: https://www.youtube.com/watch?v=RyMFun_KhAc
So fare have these modified Wheeler measurement methods and new ways to better analyze information from VNA data only been known by a limited number of RF engineers but it is since a few weeks made more public.

 

[albbg]

E Kafeman, very interesting video. Thanks for sharing.

 

[volker@muehlhaus]

Very interesting and relevant, many thanks for sharing!

The idea with the shielded box makes perfect sense. Can you comment how critical the box size and shape is?