About dipoles in UHF

[jujun]

Hello,

When building a dipole antenna for UHF, what is the proper way to avoid that the coaxial radiate ? Because I feel that a dipole is a balanced antenna, and the coax is not, so it must radiate, is that true ?

If the dipole has an angle of about 120° it's near 50ohms, but if I want a "straight" (180°) dipole, what is the proper way to match it ? L C SMD components near the feed point at the output of the coax ? Are the losses higher than a simple narrower angle ?


Thank you

J

[Wirehead]

-Current choke near feedpoint of antenna should solve that
-Try to use / get access to a VNA - then you can use the Smith chart to determine matching

[jujun]

-Current choke near feedpoint of antenna should solve that

Maybe I can simply use a choke ferrite (EMI ferrite that I clip on the coax) ?
Is there some SMD componement that do that with less weight ? (It's for a small signal, about 1W)

-Try to use / get access to a VNA - then you can use the Smith chart to determine matching

Yes, it's not a problem.

But it's hard to think about the losses of every possibilities.

[PhilipPeake]

This is the traditional way to do it:

http://www.hamradio.me/wp-content/uploads/2010/08/baluns_sidestub1.png

[PA0PBZ]

This is another way to do it:

[metrologist]

Here is some technical note about lossy stub matching networks.

http://www.dtic.mil/dtic/tr/fulltext/u2/687101.pdf

One of mine uses coax matching stub and another is a shorting stub.

Here are some more ideas:

http://dg7ybn.de/Symmetrising/Symmetrising.htm

[T3sl4co1l]

Anything will do:

- Wound balun (transformer). Has to be pretty small due to the wavelength = poor power handling (good for rx?).  Provides any impedance match you want, too.
- 1/4 wave transformers (not transformers in the traditional (wound) sense): same thing, narrower bandwidth; requires some design to realize.  Downside: you probably don't have exactly the right impedance transmission line to make one.
- Gamma match: the dipole is one continuous rod, with the shield grounded to the middle of it.  Signal is tied to a short arm, which attaches to one side of the dipole, close to the middle.  The loop made by this arm couples into the field of the full dipole.  There may be some residual line current, because the system isn't perfectly symmetrical after all.
- Direct connection plus current choke: a current choke can never null the feedline current, so there will always be residual, and that should be shunted to a large ground plane before connecting to the transmitter.  Radiation pattern will be skewed as a result.  The direct connection gives poor matching, but can be tuned out (give or take feedline losses, mind -- coax is relatively bad at that, compared to twin lead, say), if you don't mind narrower bandwidth in the process.
- Not giving a shit: no impedance transformation or balancing; shield current is terminated to the nearest ground, and the radiation pattern and impedance mismatch is just whatever it is.  Commonly seen in wifi patch antennas in consumer equipment.

[jujun]

Thank you all for you answer, it's great

Is there a way to measure how the antenna is symmetrical ?

Can I see any change with a VNA after I add a balancing "trick" ?
Of course the impedance transformation is easy to see with a VNA, but how about the balancing ?

[xaxaxa]

A vna won't tell you anything about antenna balance, but a good way to test if an antenna is well balanced is by feeding some medium power RF into it (1W or so) and probing the outer shield of the feed coax with a diode detector and led (bat17 plus red led will give you good sensitivity); if the led lights up then you have imbalance.

[jujun]

A vna won't tell you anything about antenna balance, but a good way to test if an antenna is well balanced is by feeding some medium power RF into it (1W or so) and probing the outer shield of the feed coax with a diode detector and led (bat17 plus red led will give you good sensitivity); if the led lights up then you have imbalance.

Great advice !
Thank you

[David Hess]

Current and 1/4 wave chokes work but do not improve the impedance match; since it is only 1.5:1 it can usually be ignored.  I prefer the bent ground plane for simplicity and it solves both problems but it is not a dipole form factor.  The gamma match works great as described by T3sl4co1l but like a folded dipole, it is more suitable for Yagi and quad directional antennas.

If you want something which is lightning and ESD resistant, then I would use an end fed dipole in the form of a grounded j-pole despite the unbalanced to balanced difficulties but it is a whole different discussion.  If you use an antenna with an insulated element, then consider shorting your feedline with a 1/4 wave stub for protection if the antenna will be exposed.

[metrologist]

I liked loops. why dipole?