Perplexing Arrow J-Pole tuning question...?

[ko6jw_2]

The zombie j-pole thread rises again. It won't die.

Consider the following:

1. J-poles are also called end fed Zepps. They are half wave antennas with a quarter wave matching section. Why? An end fed half wave antenna has (theoretically) an infinite impedance. The quarter wave section inverts that impedance to something like 50 ohms.

2. A j-pole can be either direct fed (open stub) like the Arrow antennas. Yes, they are really j-poles. Or, a j-pole can be shunt fed with the bottom shorted and the feed point moved up the matching section a bit.

3. Either design is perfectly valid. However, the open stub is the only configuration that can be directly fed with coax. The shunt fed j-pole antenna presents a balanced load and needs a balun. Connecting coax directly to the shunt feed is asking for problems and this is why many operators don't like j-poles.

4. The balanced feed concept in mostly ignored and/or hotly disputed. Don't like the idea? Don't build a j-pole and blame the antenna. Don't blame me for telling you how j-poles work.

By the way, I'm sure Arrow Antenna would be happy to hear your views on their products. Clearly they know nothing about antenna design.

Back to the present issue. Open stub j-poles are great, but they are difficult to design and are nearly impossible to tune since there is no adjustment point other than changing the lengths of the elements. So the first question is, did you know what you were doing when you changed the center frequency to 160Mhz? How did you arrive at the element lengths? If there is a problem here, you will not be able to recover no matter what coax you are using. Arrow actually knows how to make these antennas. I've tried the same thing as you and never totally succeeded.

You might be better making a shunt fed j-pole. Use a coaxial balun - not a choke and I guarantee that you will get better results. The balun solves issued with weird patterns, interaction with nearby objects and problems with resonance. Certain folks out there will dispute this. They should not use j-poles. By the way, I discussed this with a friend who has been a ham for over 50 years, has built many j-poles and has a Phd in electrical engineering. His advice is that of course you need a balun.

How to make a coaxial balun? Look in the ARRL Antenna book.

Good luck.

 

[Mikejo]

Ok, Sounds good U-Da-Boss!

...The shunt fed j-pole antenna presents a balanced load and needs a balun....

Guess I need to put a Balun (Balance-Unbalance) on a Balanced Antenna, not a Choke? Or was that a Typo?

 

[nanZor]

deleted most -- sorry guys but got too long winded once I looked at it.

 

[prcguy]

A highly effective feedline choke becomes a 1:1 balun by virtue of it isolating the antenna side of the feedline from the radio side and it can allow unbalanced coax to play nice with a balanced antenna.

Ok, Sounds good U-Da-Boss!



Guess I need to put a Balun (Balance-Unbalance) on a Balanced Antenna, not a Choke? Or was that a Typo?

 

[nanZor]

Mikejo - here's what's going on and it is interesting and easy to fix. Just know that there is an additional problem relating to scanner use. Let's fix the problem first.

Unless your antenna is a purely resistive impedance, (which it isn't), then to measure your swr with any accuracy at all means you'll need to make a special measuring jumper long enough to be convenient not to be affecting it with your body.

You can't get an accurate measurement off a long random piece of feedline, but you make your measurements at the end of an *electrical half-wavelength* of coax. Coax basically acts as a transmission line transformer along it's length and we want a length to mimic exactly what it seen at the feedpoint. Thus the need for the electrical half wavelength.

1) Find the length needed:
492 / Freq mhz = half-wavelength in feet. NOW, multiply that by the VF, or velocity factor, which is around 0.66 for RG58.

492 / 160mhz = 3.075 feet. Now, 3.075 x 0.66 = 2.0295 feet.

That might be too close to make a measurement with by body effect, so you can make it multiples of that. Since we aren't trying to be nasa close for transmission, let's just say you could use a measurement jumper of RG58 of about 2/4/6/8 feet lengths for the railroad band at 160mhz. Dont' go too long with multiples or you'll run into cable loss issues as well. Just get it long enough to be convenient and not be a body-effect closeness issue.

Measuring with any other lengths, especially those that are long will result in major errors due to the cable transmission line transformation and loss. This was the "CB" trick of tacking on small random lengths of coax to an existing feedline. All you are doing is fooling the *transmitter* with false swr measurements by coax transformation, but the feedpoint impedance isn't changed at all!

But be sure to find the real VF or velocity factor for your coax before you make these cuts and assemble the test jumper.

tbc next message...

 

[nanZor]

Issues with the homebrew open stub j-pole:

The open stub version differs from the more common types that have a sliding jumper going up and down the stub, which are admittedly easier to adjust to center frequency then the open stub version by virtue of having to trim the elements exactly.

Thing is, many users of both don't use a short electrical half-wavelength jumper at their frequency of interest and get their swr readings wrong too. But now YOU will get them right, or very close to it.

Problems and indications that may have left you scratching your head:

1) When you removed the short coupler to form the "stub", what is left?

Answer: A 3/4 wave spike attached to the coax which uses the common mode of the coax braid as basically a very long-wire version "OCFD" type of dipole! The swr and resonance will wander around close to the 3/4 wave frequency up and down as the coax, not cut to any specific length, transforms the impedance up and down and up and down as it crosses the quarter wave lengths to the half-wave multiples and back again.

Of course, this "stubless" j-pole, now an OCFD, is a very big longwire with screwy directional lobes. But of course you don't want that, you want the J with the stub!

So you replace the short element to form the stub, but you are still back in the ballpark!! What gives??

*** AT RESONANCE *** near your center frequency, the open-stub J-pole is now trying to feed an end-fed half-wave like it is supposed to.

BUT, unless you cut your special measurement jumper to an electrical half-wavelength the screwy results will be somewhat as head-scatchy nuts!

Nearing the end here...

Now that you have your measurement jumper properly cut, you'll find that the lengths of the stub and the driven element itself VERY critical. If you get them wrong, or you go so far out of band that the "stub" is well away from a quarter-wavelength, what you have is a very long "OCFD" dipole, with one very short end, a tiny neaby parasitic element, (no longer doing any impedance transformation) with screwy lobes.

And this is the problem for scanners with typically weak front ends susceptible to broadband desense and overload. Unlike an amateur transceiver, which typically (or used to) have a very narrow front end, signals well out of your desired band of the open j-pole, will happily couple to your "electrically model-changed" broadband ocfd and may present issues. The stub is just ignored and is now a parasitic element well outside of resonance.

Sorry to be so long winded - fix your swr measurements with the special jumper, and trim the elements carefully watching the changes in bandwidth and swr. If you get it right - congrats! Just be aware of the non-resonant "OCFD" effect that might overload the scanner depending on your environment.

AND yes, use a choke of some sort. Typically #43 type ferrite (good from 30-300 mhz) at the feedpoint. Even with all of this right, the J-pole, no matter it's variety, has common mode issues going from unbalanced (coax), to a balanced stub, and the back to an unbalanced end-driven element.

Again - sorry to be so wordy about it, but just thought you might want to play with your new creation, knowing some of the perils behind it - which may or may not affect your happiness with it.

 

[nanZor]

TIP: You can use your antenna analyzer to find out the length of an electrical half-wavelength pretty easily too. There are many discussions, but this one has an example using an MFJ analyzer for quarter waves and read on for half-wave sections:

https://www.hamradio.me/antennas/measuring-quarter-wave-coax-stubs-using-mfj-analyzer.html

Once you have that cut to your operating frequency, or center of band, you can start to fine-tune that open stub j-pole to your railroad freqs - which is important to know, because if you are too far off, you have an ocfd with a teeny parasitic element.

 

[W5lz]

1. SWR has nothing to do with resonance.
2. Get that resonance first, then worry about matching impedances.
3. Resonance depends on antenna (radiating elemenyts) and is definitely wanted. (Resonance = efficiency)
4. Matching input impedance is also a very nice thingy! Get it if you can.
5. A loading coil affects antenna length (electrical length). It's typically found up from the base of the antenna but ma be enclosed in the impedance matching coil's housing AT the base of the antenna.
6. An impedance matching coil is located at an antenna's base where the feed line connects.

There are variation in all of that, just like variations between manufacturers antennas!

 

[W5lz]

If changing the feed line's length by a foot makes that much difference then there's a problem with the antenna, not the feedline.