Best current balun for dipole

[AF5C]

I have a homebrew G5RV type dipole, and need to replace the balun at the twin lead/coax junction (I am currently use a 4:1 voltage balun there).  From what I understand, a 1:1 current balun would be much better in eliminating RF traveling on the shield of the coax. 

Any recommendations as to what the best current balun (at a reasonable price) is out there?

73 John AF5CC

[VE7TIT]

> Any recommendations as to what the best current balun (at a reasonable price) is out there?

From what I've read (I don't have one), a ferrite bead balun looks like a good choice. But don't you still need the 4:1 impedance conversion?

[WB6BYU]

But don't you still need the 4:1 impedance conversion?

No, the G5RV is supposed to have reasonably low SWR on 80, 40, 20 and 12m
with a 1 : 1 balun.  Using a 4 : 1 instead will increase the SWR on those
bands and slightly increase the losses in the coax.

On the other bands the coax losses will generally be fairly high (unless the
coax is short, or otherwise has very low loss to start with) so, even if it
can be matched with a tuner, efficiency won't very good.

[W5DXP]

a ferrite bead balun looks like a good choice.

My original G5RV came with eight FB-77-5621 beads inside a waterproof piece of PVC pipe. A 4:1 balun is a terrible choice for 80m operation as it lowers the already low resonant impedance from 16 ohms to 4 ohms thus increasing the SWR on the coax from 3:1 to 12:1.

[AF5C]

My dipole isn't exactly a G5RV. It is approximately 102 feet long, but uses a run of Radio Shack 300ohm twin lead to a 4:1 balun, then a run of RG-11 75 ohm coax to the shack. I am not sure how long the twin lead run is, probably 20 feet or so.  I am planning on moving the new balun closer to the shack so I can cut down the length of the coax run to 20 feet or so.

John AF5CC

[WB6BYU]

Whenever you use a combination of transmission lines like that, it's good to
calculate the impedances and SWR at various points in the system.  That's
because the wrong combination can cause very high losses.

For example, say you have a 90 ohm impedance at the antenna (which is
a ballpark value for 20m.)  If the twinlead is 1/4 wave long, it would
transform the impedance up to 1000 ohms at the coax junction, so the
SWR on the 75 ohm feedline would be 13 : 1.  But if the twinlead was
1/2 wavelength (as it it is with the G5RV feed system) then the impedance
at that point would be 90 ohms, for an SWR of 1.2 : 1.

When you consider the range of impedances that the tuner might encounter
at the end of a random length of coax, it could be between 6 and 1000 ohms.


Of course the feedpoint impedance varies with frequency, as does the
electrical length of both feedlines.  So you have to do this analysis on each band
that you want to use, like this:

http://vk1od.net/antenna/G5RV/index.htm

If your interests are primarily 80, 40, 20 and 12m, then using the standard 20m
half wavelength of twinlead is a good approach.  If you operate more on 17m,
15m and 10m then putting the balun on the table beside the tuner in the shack
is a better choice, so the coax length is a matter of inches.

[W5DXP]

My dipole isn't exactly a G5RV.

A G5RV antenna has design specifications. It is a 1.5WL dipole on the design frequency and is fed with 1/2WL of parallel line. It works on the design frequency and on 1/2 that design frequency and on 1/4 that design frequency. Any other dimensions and it is, as you say, not a G5RV.

[K9SRV]

My advice is to listen to Cecil! He has taken much of his own time to help me.
Thanks! K9SRV

[KB5UBI]

Run an appropriate length of ladder line all the way to the tuner. Cecil's webpage addresses the appropriate feeder lengths.

An old rule of thumb:
The electrical length of the feeder plus one side of the doublet should be an odd multiple of a 1/4 wave for the frequency. That should be the current lobe maximum (low imp) on the feed line.

[W5DXP]

Run an appropriate length of ladder line all the way to the tuner. Cecil's webpage addresses the appropriate feeder lengths.

Note that when one runs an "appropriate", i.e. random length of ladder-line, it will transform the impedance to a random, i.e. unknown value, that many built-in autotuners will not tune. A wide-range tuner is usually required.

For example, the resonant 80m feedpoint impedance for a standard G5RV is about 16 ohms with an SWR on 30' of 450 ohm ladder-line of about 28:1. If the total length of ladder-line from the 102' dipole to the shack is 30'+60'=90 feet, the impedance presented to the tuner will be about 28(450)=12.6k ohms and very difficult to choke and match.

An old rule of thumb: The electrical length of the feeder plus one side of the doublet should be an odd multiple of a 1/4 wave for the frequency. That should be the current lobe maximum (low imp) on the feed line.

That old rule of thumb assumes a straight-line relationship between 1/2 of the length of a dipole and the length of the feedline and can be about 20% inaccurate in some cases. Following is a graph that is more accurate than the old rule of thumb. For instance, if 1/2 of the dipole is 0.35WL long (dipole total length is 0.7WL)  the old rule of thumb says the feedline should be 0.4WL long. The graph indicates that 0.33WL long is a better choice. The graph is based on EZNEC simulations.

http://www.w5dxp.com/majic1.gif

I wrote a DOS-BASIC program that runs under XP or DOSBox on later versions of Windows. It displays graphically the optimum length of ladder-line for variable length dipoles. It can be downloaded at:

http://www.w5dxp.com/IMAXGRAF.EXE

[KB5UBI]

Cecil,

Easy, easy. By appropriate I mean the proper electrical length for the frequency. Appropriate doesn't mean random.

Yes, the "old rule of thumb" is for a near resonate current fed antenna and meant to get you in the ballpark of the maximum current lobe. It works well in a pinch.

Steve

[W5DXP]

By appropriate I mean the proper electrical length for the frequency. Appropriate doesn't mean random.

Oops, I inferred that "appropriate" meant "from the antenna to the shack" which is often a "random" length. Sorry.