Author Topic: Yagi-Uda antenna driven element impedance question  (Read 7705 times)

0 Members and 1 Guest are viewing this topic.

Offline E Kafeman

  • Regular Contributor
  • *
  • Posts: 99
  • Country: se
    • AnTune VNA software
Re: Yagi-Uda antenna driven element impedance question
« Reply #25 on: October 06, 2019, 11:55:43 am »
Quote
In the website I linked, https://www.qsl.net/dk7zb/6m/212.htm
DE impedances of 12.5, 18.0 & 28.0 Ohms are referred to, (with matching methods described)

That is very much proof of a good amateur. Low impedance ideas are however somewhat misunderstood.
He must have tried to proof his ideas? He must have done basic measurements? No, nothing at all. That is not what HAM do.

His narrow band VSWR ideas are said less correct by this HAM: https://www.g0ksc.co.uk/intro-lfa.html
He is offering a better antenna design.
So which fanclub to belong to?
None of them have provided any facts, no calculations and no source references which else is required in any fact based technology. Why doing exceptions from EM wave technology?
We can only trust their words that their home-cooked theories are the correct one.

They have probably never bothered to read about EM theory and do not even know any entry books in this field.
A RF engineers do often have within an arm-length at least one or two books written by names such as Balanis, Kraus, Stutzman and Thiele.
Such books are often a bit heavy as intention is that you should build your own knowledge based on fact and references all the way down to Hertz and Maxwell.
Such books contains very few precalculated recipes of antennas in opposite to HAM pages.

Someone knowing difference of facts relative opinions do probably not have much in common in a group where most not understand difference between rumors and fact.
It is maybe the reason that very few active HAM have any real RF skills.

Quote
all a far  cry from approx 73.0 Ohms for a dipole in free space.
An ideal resonant dipole at free space have an impedance of 73 + j42.5 Ohm. If never learnt from school about complex numbers or that impedance is a complex value can it be hard to understand why both number parts are equally important when calculating a dipole. An ideal dipole can also be said have an impedance of 84 Ohm at an angel of +30° if you prefer to express it as a vector.
Exception are for HAM's, they are orienting with ease in a 2D projection using numbers for just one of the axes.

Quote
The reason I was thinking the impedance change might not be very large is because often the coupled elements resonating at the same frequency can have that effect, and I thought that the distance between elements was also a fixed fraction of a wavelength (as I said - not designed any yagi-uda antennas myself).

Yes you is correct. It is a common thing also for a lot of other kinds of waves such as sound and mechanical resonances...

For good reasons are often a Yagi-Uda antenna also called Yagi-Uda array (note that first letter in names and terms is many languages are uppercase).
A Yagi-Uda array is an array of dipoles where the passive dipoles are parasitic elements with no feedings point making it relative simple to calculate how they affect each other and the feed element. They affect each other by mutual impedance.

By measuring impedance of a feed dipole while adjusting distance to a passive shortcut dipole of same length can this effect be studied and it is easy to find if there are any "golden" distances where the elements not affect of each other. Directors for a Yagi-Uda are usually shorter then feed dipole and reflectors longer, but it is no absolute rule and it is not important for below shown measurements.

One of the golden distances between these dipoles as you probably already have guessed is lambda/2 and its multiples. Two dipoles at that distance and equally oriented will be blind to each other.
There are also other distances of interest. A Yagi-Uda antenna have a typical distance between each element in the range of 0.1-0.25 lambda. 0.1 lambda is a bit special distance. A shortcut dipole at that distance will not affect feed dipole resistance, but it is the distance for maximum mutual reactance.

I have done some simple measurements shown in curves below to show these effects relattive dipole distances. Measured object is a dipole with its shape adjusted to become a very ideal 50 Ohm resistive load with almost no reactance over a relative wide frequency range. Its actual impedance is not that important for below shown result but gives a simpler reference to see how another element, a shortcut dipole, at different distances affect measured impedance.

Measured antenna is center tuned at 2440 MHz
[attachimg=2]
This curve shows resistance part of impedance. Yellow curve is free space. It is when dipole have no objects in its nearfield. Green is with distance lambda/2 between the dipoles which within a reasonable bandwidth shows same resistance. It is not exactly same curve but that is a result of that measurement not is done in any anechoic chamber. Measurements are done 200 mm above my lab bench.
Blue curve, distance lambda/4 is distance resulting in maximal mutual resistance at resonance frequency.
As can be seen from this curve, if an antenna is placed near a conductive surface can it result in an increased resistance.
Reducing antenna distance relative another surface is not automatically causing an reduced resistance. In this case is it actual distance relative wavelength that is determining factor.

Red curve is a bit special, it does not affect resistance at center frequency but is rather tilted within resonance frequency range. Its distance is around 0.1 lambda, a common element distance for Yagi-Uda antennas.

Mutual reactance is just as important as mutual resistance to understand when designing an Yagi-Uda antenna but as written before, it is a bit to complicated factor to handle for many amateurs but it is still just as important factor as resistance.
Mutual reactance do however behave different then mutual resistance when distance between two dipoles varies as can be seen below.

Exactly same measurements as above, but now is it reactance part that is shown.
[attachimg=3]
Green curve is a bit low but that is due to measurement limitations. Ideal should it also here be same as curve for free space.
In opposite to resitive curve where blue curve showed max mutual resistance is mutual reactance just minor affected.
It is red curve that represents max mutual reactance. It is when dipole elements was at typical "Yagi-Uda distance".

If a Yagi-Uda antenna consist of many elements is it of course a mutual impedance between all elements. Here plays also reactance part an important role. It is much same as when designing a bandpass filter with a number of capacitors and inductors. Right length and place for each element will play an important role for total antenna performance.
Playing with reactance for each element is mainly y adjusting element length. To balance high mutual inductance is director element slightly shorter then lambda/2 used, which causes them to be a bit capacitive.

Same measurement as above but represented by its impedance. can be seen in curve below. Table is for yellow curve in free space.
[attachimg=1]
Shown antenna behavior and mutual impedance is typical but not exact repeatable.
Result is depending on a lot of details, including standing waves relative a roll of solder in opposite end of my labbench.

Above measured dipole is rater wide band relative an ideal dipole as can be seen in the curves. According to some HAM logic should that be negative for antenna efficiency due to low Q.
Antenna efficiency expressed in percent is max 100%. Higher is not possible as an antenna is a passive structure.
For that reason can it be interesting to see measured antenna efficiency for this dipole in free space:
[attachimg=4]
Not mentioned regarding these mutual effects is if element affecting the dipole not is a good conductor. Soil below antenna  is a often poor conductor. A mix of reflector and absorber.
An antenna above ground, especial if soil below antenna is wet can it be very absorbing. Absorbing soil act as a variable atenuator for the antenna.
Commersial shortwave radio with real long horizontal dipoles, a few MHz or lower, do often place a metal net at ground below antenna to avoid atenuator effect of antenna. Ground do then become a reflector instead of absorber and total radiated power increases.

Not neither mentioned is if there are more then two dipole elements which often is the case for Yagi-Uda antennas. All elements will affects each other. Not really any complicated then already shown above but it will be more terms in calculations of complex impedances.

Actually is it a lot of other things not either mentioned but it is still at a very basic level. It is better to read a good RF theory entry book from any of above mentioned authors then trust me, as I may be wrong  and is definitive incomplete. There are much more to learn. A book that starts from almost zero assumed knowledge is "Antenna theory" by Balanis.
« Last Edit: October 06, 2019, 12:07:08 pm by E Kafeman »
No signature
 

Offline cuagn

  • Newbie
  • Posts: 1
  • Country: fr
Re: Yagi-Uda antenna driven element impedance question
« Reply #26 on: July 22, 2022, 11:03:11 am »
All,

I was really happy when I found this thread on the Web. The question of the DE impedance is precisely the question I would like to get an answer.

The problem is that, after reading long (too) explanations  about antennas, I still have the question.

Is the only solution to connect an Antenna Analyser (not easy to have one above 1 GHZ) ?
Is there really any HAM analysis of this subject ?

I would appreciate any complementary iformations.

Regards
73, Marc op F6DNH
 

Offline mag_therm

  • Frequent Contributor
  • **
  • Posts: 605
  • Country: us
Re: Yagi-Uda antenna driven element impedance question
« Reply #27 on: July 22, 2022, 11:48:33 am »
Hi Marc,
There is a model on XNEC2  called "2m_yagi.nec"  , a 6 element classical Yagi, not including matching. It is in the directory "examples"

If you have not used XNEC2, it is a challenge to learn the archaic card method to build a model.
Otherwise XNEC2 works OK.
I think best way is to take an existing model as above, learn it, then spin of your own model while making small adjustments at each run.

Here is a User Guide : https://www.nec2.org/other/nec2prt3.pdf
 

Offline El Rubio

  • Regular Contributor
  • *
  • Posts: 74
  • Country: us
Re: Yagi-Uda antenna driven element impedance question
« Reply #28 on: August 11, 2022, 02:38:35 pm »
The articles and notes of L.B. Cebik are excellent. Most of his articles are centered around antenna designs and modeled performance. The articles are written in a practical manner and not the least bit pedantic. He is no longer with us, but his articles can be found at:
http://on5au.be/Cebik%2520documents.html

One thing he does mention is that many amateurs will use 4:1 baluns inappropriately. The library is extensive but good reading.

 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7053
  • Country: au
Re: Yagi-Uda antenna driven element impedance question
« Reply #29 on: August 12, 2022, 02:25:25 am »
No, it is not easy.

The dipole more than likely when near parasitic elements in a yagi configuration will have different impedance, than free air one.

Half-wave ("Open") dipole does have free space impedance of 73ohm (close to 75, sure), but still you can't connect that to a unbalanced coaxial 75 ohm cable, not to mention, that 75ohm is not the choice for transmitting, 50 is.

For open dipole with a 50ohm feed line, you would still need a 1.5 : 1 impedance balun (baloon is a different thing). 75 ohm unbalanced coaxial  feed line would just get away with a 1:1 balun.

I know, there are many hacks and things that kind of work and kind of don't. I am not interested in those.  I am not targeting "the simplest or weirdest antenna build", I just want to make a good quality yagi antenna. Not a slapped-together one, I have lot of those already.
Even for a plain old half wavelength dipole, the often quoted "73 \$\Omega\$" is only valid if the antenna is around one wavelength or more above ground.
Other heights offer different impedances, & some common heights are a good match to 50 \$\Omega\$.

Strictly speaking, you should use a balun, but there are many HF dipoles happily operating with direct connection to coax.

This guy talks about the change in impedance of the driven element of a Yagi.
https://www.qsl.net/dk7zb/dk7zb-match.htm

Martin also presents designs for a number of antennas, including Yagis on
https://www.qsl.net/dk7zb/start1.htm
 

Offline E Kafeman

  • Regular Contributor
  • *
  • Posts: 99
  • Country: se
    • AnTune VNA software
Re: Yagi-Uda antenna driven element impedance question
« Reply #30 on: August 13, 2022, 02:18:00 am »
>Even for a plain old half wavelength dipole, the often quoted "73 \$\Omega\$" is only valid if the antenna is around one wavelength or more above ground

It is not correct if it is antenna impedance. If we ditch reactance part and only focus on resistance is it somewhat more correct. At one wavelength above ground have the antenna resistance crossed 73 Ohm four times from lower position and upwards.This height can be expressed as [0.23 lambda + (n x 0.25 x lambda)] as first height above ground in air is around 0.23 wavelengths when first crossing of antenna radiation resistance of 73 Ohm occur for a simple center feed dipole antenna.
This resistive value is same resistive value as for antenna radiation resistance in free space why it can be a bit interesting when trying to understand dipole antenna nature.
Reactive part do not share this behavior as can be seen in my above dipole measurements and shown in Smith chart so I assume it is resistance you is referring to, not impedance.

Antenna impedance and resistance are different animals not living in same dimension.
Impedance is a value calculated for alternating voltages.
It is an expression of voltage/current wave ratio and phase difference.

Antenna radiation resistance is neither same as measured antenna resistance, as the later includes antenna resistivity losses.
1-10 % of total resistance are normal losses for a typical well designed horizontal wire-dipole in range 10-100 MHz at decent height and common type of ground but can be a lot more if ground below antenna is a less ideal electrical reflector.
In controlled environment can losses be lover.

>Other heights offer different impedances, & some common heights are a good match to 50 \$\Omega\$.


Maybe are you thinking about 0.18 lambda which is the only height above an ideal ground which an ideal horizontal antenna dipole resistivity radiation is reduced to around 50 Ohm due to ground height.
Even lower height -> lower resistance. It is however a steep resistance curve not really good for stable antenna design.
All other higher heights for an ideal dipole will all have an resistance higher then 50 Ohm and will stabilize at an impedance of 83300 Ohm for heights closer to free space.

Impedance variation dependence with height above ground is a bit more complex then the resistive variation.
Both your statements are covered by above by me previous shown Smith chart. It is an dipole antenna which impedance is measured at different heights above ground.
Check these measurements and compare how it agrees with your statements.
Smith chart is mayor method to show antenna impedance.

Basic knowledge about "plain old half wavelength dipole" impedance behavior needs some theory to be understand.
I know that the word "theory" is scaring for many lacking antenna knowledge but with minor knowledge can big mistakes be avoided.Below 7 minutes can be well invested dipole knowledge intended for beginners in EM-theory.
Dipole antenna impedance math formulas can be a bit scaring when first seen but it is very simple explained in below video.
Knowledge about AC circuits and reactive component math in general is assumed prior knowledge.
https://youtu.be/QSnFiFReKyg
Linked site below is well known because it explain many aspects of antenna theory in simplest possible way without making it wrongly oversimplified and very little prior knowledge is needed:
https://www.antenna-theory.com/basics/impedance.php
« Last Edit: August 13, 2022, 03:06:18 am by E Kafeman »
No signature
 

Offline profdc9

  • Frequent Contributor
  • **
  • Posts: 285
  • Country: us
Re: Yagi-Uda antenna driven element impedance question
« Reply #31 on: August 13, 2022, 04:07:49 am »
EZNEC 7.0 is now free, and 4NEC2 also works reasonably well.  These are method-of-moments simulators that can be used for calculating feedpoint impedance the radiation pattern, and should produce a result that is fairly close to reality, though inevitably a design may require minor tweaking depending on deviations of the environment and antenna materials from the idealized method-of-moments calculation.  These programs can take a little of the sting out of writing NEC2 models.

A conventional current balun for HF would be something like wrapping a coaxial cable several times around a ferrite toroid.  At VHF/UHF, the capacitance between adjacent turns of the winding, even though it is on the order of a picofarad, greatly decreases the choking impedance of the balun.  A better choice might be to pass a thin coaxial cable, for example RG-174 or RG-316, through several binocular core apertures.  Such binocular core apertures are used to make the common 75/300 ohm 1:4 balun used between 75 ohm RG6 cable video transmission coax and 300 ohm antennas to receive terrestrial broadcasting.

It would be fairly easy to measure the feedpoint impedance of an antenna using a VNA, even a common one such as the NanoVNA, but quite a bit harder to measure the gain and radiation pattern of the antenna.  Shield rooms, goniometers, and biconical antennas are often used for such measurements, as they are often used to measure other possibly unintentional radiators for EMI.  It may be very difficult to eke out the last dB out of a homebrew design without a good measurement setup.  Even still, if the ground is nearby, or there are other objects nearby, this could well frustrate achieving the best gain anyways. 

I will note that the idealized calculation of dipole impedance with antenna length (which I have included a graph of below from "Antenna Theory and Design" Figure 8.16 by Balanis) a point at which feedpoint reactance is zero is actually for an antenna slightly shorter than 0.5 waves, actually around 0.47 to 0.48 waves, at which the feedpoint resistance is actually fairly close to 50 ohms.  It's this reason why generally one gets a pretty good match trimming an antenna, because the resistance is going to be closer to 50 ohms when you trim it for minimum SWR.




« Last Edit: August 13, 2022, 04:11:43 am by profdc9 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf