The 5/8-Wavelength Antenna Mystique

[Marconi]

Here is an article that might be interesting to study. It is by Ronald K. Reynolds, K7DBA, where he reviews a 1924 report on the operations of the 5/8 wave monopole for the Broadcast Radio industry. I've posted this before in the course of other threads, but have never made it the subject of a thread on its own.

The Reynolds review was based on a study, by Stuart Ballantine, "On the Optimum Wavelength for a Vertical Antenna Over Perfect Earth." It was published in "Proceedings of the Institute of Radio Engineers, 1924, Volume 12, p. 833, December 1924.

In addition to discussing the 5/8 wavelength radiator for Broadcast Radio for those times, there is some history noted and antenna range ideas used in testing antennas that you may also find interesting.

The big thing I find interesting in this review is how it spells out clearly all the limitations noted for the results found in the report. I have claimed for a long time these limitations have been ignored or all but forgotten by the CB World. And then we read current day articles from authors like Cebik and W8JI reporting their works indicate far less superiority for the 5/8 wave over the paltry 1/4 wave...which also seems to supports the ideas in the review.

So, what is the truth, is the 5/8 wave idea as good as many in the CB World claim, or are we destined to ignore the truth?

http://www.worldwidedx.com/attachme...n-master-58th-wavelength-antenna-mystique.pdf

IMO, I think the work is the defact'o commentary on the 5/8 wavelength radiator, but for arguments sake, I think there is a mistake in what Fig. 4&5, are suggesting, but Reynolds is (SK) so we can't ask him to explain.

 

[The DB]

I think Figures 4 and 5 are displayed a bit differently than we are used to seeing.

Typically, the pattern with the most gain would define the maximum gain of the circle, and other patterns would be scaled within the circle. In these case the author seems to have taken the maximum gain of each pattern to the edge of the circle instead of scaling them all accordingly. Unfortunately this may lead to some confusion where people think that you get more from the less gain scenarios than you actually do.

I think many people see the difference between 1/4 wave and 5/8 wave antennas in "perfect" theoretical conditions and attempt to apply those differences to their very imperfect setup. Very few people have access to anything in reality that comes close to the conditions needed for the gains they think they will get.

I think part of this idea comes with the idea that more is better. (This also comes into play when many think about cophased antennas, two antennas must be better than one, right? ...)

I, personally, have long been under the impression that 1/4 wave radials are simply not enough for a 5/8 wave antennas. I'll have to see if I can find a study where four 1/2 wave radials (or perhaps slightly shorter) on a 5/8 wave antenna provided actual gain over four 1/4 wave radials on the same antenna. That being said once you got to or just past that half wavelength point for the radials the upper lobe became the dominant. This was shown with modeling but they did not do an actual study using actual antennas.


The DB

 

[Marconi]

I think Figures 4 and 5 are displayed a bit differently than we are used to seeing.

Typically, the pattern with the most gain would define the maximum gain of the circle, and other patterns would be scaled within the circle. In these case the author seems to have taken the maximum gain of each pattern to the edge of the circle instead of scaling them all accordingly. Unfortunately this may lead to some confusion where people think that you get more from the less gain scenarios than you actually do.

I think many people see the difference between 1/4 wave and 5/8 wave antennas in "perfect" theoretical conditions and attempt to apply those differences to their very imperfect setup. Very few people have access to anything in reality that comes close to the conditions needed for the gains they think they will get.

I think part of this idea comes with the idea that more is better. (This also comes into play when many think about cophased antennas, two antennas must be better than one, right? ...)

I, personally, have long been under the impression that 1/4 wave radials are simply not enough for a 5/8 wave antennas. I'll have to see if I can find a study where four 1/2 wave radials (or perhaps slightly shorter) on a 5/8 wave antenna provided actual gain over four 1/4 wave radials on the same antenna. That being said once you got to or just past that half wavelength point for the radials the upper lobe became the dominant. This was shown with modeling but they did not do an actual study using actual antennas.


The DB

I tend to agree, the pattern over an infinite ground should show maximum gain in this case, and unlike noted in image #2, where all three models were over a perfect ground, #4 & #5 do not show the same advantage for the infinite ground plane model. So you're right, and that is why I made the note.

Regarding using radials with a full wave diameter below the 5/8 wave radiator, I would think that would overload the radials with excessive currents and that would tend to push the RF available up into a higher lobe that might even dominate the pattern. This is assuming horizontal radials.

There is a Ham operator here in the area that published an Internet article on his Website one time that was modeling something similar with an Imax, using steeply slanted down 1/2 and 5/8 wavelength radials. He also claimed to have tested the idea in real world testing and it showed a big advantage in gain, but I'm not sure. I was curious, so I visited his location. He did have an Imax up, but it had no radials on it at all. I tried to talk to him, but he brushed me off. I think I still have his presentation on this subject which has been removed from his Website, but it was done on blue print paper...and it does not scan well. I may still have the work, and if so I'll try again and posted a couple of those old images with his text attached where he makes his claims.

 

[The DB]

Regarding using radials with a full wave diameter below the 5/8 wave radiator, I would think that would overload the radials with excessive currents and that would tend to push the RF available up into a higher lobe that might even dominate the pattern. This is assuming horizontal radials.

Pulling from memory here so staying away from absolutes.

That is something that happened starting near the one wavelength diameter point, not sure if it was at, just above or just below. The point just before that happened is the point that showed the most gain on the horizon. If I recall correctly, somewhere before the groundplane was at two wavelengths in diameter (or perhaps just beyond that point) the radiation pattern favored the lower horizontal lobe again so I believe that high radiation angle point is temporary.


The DB

 

[Marconi]

Pulling from memory here so staying away from absolutes.

That is something that happened starting near the one wavelength diameter point, not sure if it was at, just above or just below. The point just before that happened is the point that showed the most gain on the horizon. If I recall correctly, somewhere before the groundplane was at two wavelengths in diameter (or perhaps just beyond that point) the radiation pattern favored the lower horizontal lobe again so I believe that high radiation angle point is temporary.


The DB

I may have misunderstood what you said earlier, but I'll go back and check. I think I made the 5/8 wavelength model with 4 x 18' radials for a full wavelength in diameter instead of 4 x 9' radials. If so, I'll also make the radials a wavelength each, as in 4 x 36' foot radials, for a 2 wavelength diameter groundplane.

Making the radials 18' feet each did increase the gain a little compared to 9' radials, and improved the match a bit. It also about doubled the maximum current on the longer radials, to say nothing about the radial span being 36' feet wide.

However, in making the radials 36' each...it dropped the gain noticeably and additionally the match got even worse than with 9' radials.

As a note, I also used a model that was 18' high where the gain for the 9' radials was 3.51 dbi @ 12* degrees, and the 18' foot radials showed 3.96 dbi @ 12* degrees. The 36' foot radials showed 2.4 dbi @ 12* degrees.

I don't want to get too far off the subject of this thread, but do you know the source you saw? If you have it, I'll post my model results and maybe we can compare.

 

[HomerBB]

I may have gotten lost somewhere between my ears along the way to the meaning of figures 4 and 5, but my understanding of what he was attempting to show was referenced to his idea of what he called the "edge defraction".
I see the graph depicting, along the 270° and 90° line the demarcation between the radiation above the raised radials and the radiation below the raised radials. I could be way off.

From the article:

"Regardless of the diameter of the ground plane, even if it were 100 wavelengths in diameter, as long as it is hoisted up in the air, and the observer is beyond the edge of the ground plane, there will be radiation below the "horizon" plane. The principal cause is edge diffraction."

 

[radioreddz]

i know in testing i have done and with what i have up in the air right now i see know real kick "A" difference between the 3 main antennas A-99 Imax 2000 or maco V5000. right now i have all three antennas up at 42' to the feed point all with 112' of 213 coax each with two ground rods and then all joined together with #4 copper to the main utility ground all on a MFJ selector and when switching between the 3 with stations any where from 20 miles thru 70 miles i see really no difference with only a ever so slight edge to the imax 2000 at 60 to 70 miles. at my locaton i don't see any difference in the noise floor between the aluminum and fibrglass antennas. also each antenna is spaced 30' apart. i think its more cb tails just like drill the hole vs a mag mount. also i think many times some folks just don't have there antennas installed correctly like the guys that knock stardusters but when you look at there installs the radials a dropping around a chiminey or below a roof line or an imax or maco V58 that does not have 18' of clearance around it. Marconi correct me if i'm wrong i know you have tested a many of antenna with seemingly same results. is it myth,lore tales or what.

 

[The DB]

I don't want to get too far off the subject of this thread, but do you know the source you saw? If you have it, I'll post my model results and maybe we can compare.

Give me a day or three, I'll see what I can find... I do know the main part of the article was about vertical 2 Meter antennas and adjusting the vertical element size to adjust for the size of the horizontal groundplane, the feedpoint impedance the gains as well as radiation angle resulting. There was a short section on a 5/8 wave 2 meter antenna and ground plane radial size at the bottom, and this is what I was referring to.


The DB

 

[Marconi]

i know in testing i have done and with what i have up in the air right now i see know real kick "A" difference between the 3 main antennas A-99 Imax 2000 or maco V5000. right now i have all three antennas up at 42' to the feed point all with 112' of 213 coax each with two ground rods and then all joined together with #4 copper to the main utility ground all on a MFJ selector and when switching between the 3 with stations any where from 20 miles thru 70 miles i see really no difference with only a ever so slight edge to the imax 2000 at 60 to 70 miles. at my locaton i don't see any difference in the noise floor between the aluminum and fibrglass antennas. also each antenna is spaced 30' apart. i think its more cb tails just like drill the hole vs a mag mount. also i think many times some folks just don't have there antennas installed correctly like the guys that knock stardusters but when you look at there installs the radials a dropping around a chiminey or below a roof line or an imax or maco V58 that does not have 18' of clearance around it. Marconi correct me if i'm wrong i know you have tested a many of antenna with seemingly same results. is it myth,lore tales or what.

I think you're close radioreddz. That is similar to what I see too...there just ain't that much difference in my comparisons.

That is not to say this is the absolute straight skinny or the only truth of the matter, but it is what I see when I get my antennas on par for a fair comparison.

This idea also seems to me to be supported by the guys that most around here might consider good sources, but the old stories about antenna gain just keep on coming up. This is why I posted the new thread on the 5/8 wavlength Mystique, that IMO indicates why there is probably less difference than theory predicts.

I had the thought the other day when I was posting my new thread on the 5/8 wavelength Mystique, could I find some videos of guys comparing antenna signals that showed some of the big differences we hear reported. I know I don't have any such videos. Plus all of my comparisons are A/B testing, side by side like you did, and that too is a big issue with many of those crying foul...when they hear about A/B reports.

 

[Marconi]

Give me a day or three, I'll see what I can find... I do know the main part of the article was about vertical 2 Meter antennas and adjusting the vertical element size to adjust for the size of the horizontal groundplane, the feedpoint impedance the gains as well as radiation angle resulting. There was a short section on a 5/8 wave 2 meter antenna and ground plane radial size at the bottom, and this is what I was referring to.

The DB

DB don't go to a lot of trouble on my account. I looked back at my post and I can't recall, find the models, or remember what was going on in my mind at the time. I think it had to do a some far out idea with modeling and I don't think I saved anything because of it. If you remember the idea I was talking about...you might refresh my memory however.

I hate to admit it DB, but sometimes my memory plays tricks on me like that, if a day or two go by without discussion on my part.

 

[The DB]

DB don't go to a lot of trouble on my account. I looked back at my post and I can't recall, find the models, or remember what was going on in my mind at the time. I think it had to do a some far out idea with modeling and I don't think I saved anything because of it. If you remember the idea I was talking about...you might refresh my memory however.

I hate to admit it DB, but sometimes my memory plays tricks on me like that, if a day or two go by without discussion on my part.

To late... Link...

Near the bottom there is a section that starts with "What about 5/8 wave antennas?".

Funny, I remembering there being a bit more of a difference...

The page is a good read, especially if you are just interested in 1/4 wave antennas.


The DB

 

[Marconi]

To late... Link...

Near the bottom there is a section that starts with "What about 5/8 wave antennas?".

Funny, I remembering there being a bit more of a difference...

The page is a good read, especially if you are just interested in 1/4 wave antennas.


The DB

DB, I have referred to that link by Walter Fair before.

In my experience with Eznec I've had a thought about how SWR affects the gain shown. Maybe that doubling the radial length does increases the mismatch, and maybe that is what causes the gain to appear to increase at some point, but note the change in angle. The effect on the pattern makes the primary lobe at a high angle, but I don't see it returning to being low again like you note.

So, if we match this mismatch using a device we will probably be adding loss back into the antenna, and maybe that causes the gain to be reduced back down to where we started or worse.

 

[The DB]

DB, I have referred to that link by Walter Fair before.

In my experience with Eznec I've had a thought about how SWR affects the gain shown. Maybe that doubling the radial length does increases the mismatch, and maybe that is what causes the gain to appear to increase at some point, but note the change in angle. The effect on the pattern makes the primary lobe at a high angle, but I don't see it returning to being low again like you note.

So, if we match this mismatch using a device we will probably be adding loss back into the antenna, and maybe that causes the gain to be reduced back down to where we started or worse.

I wasn't worried about the matching, just the pure gain figures, which I thought were higher than was actually listed. It has been a while since I've seen the page... My mistake...


The DB

 

[Marconi]

I wasn't worried about the matching, just the pure gain figures, which I thought were higher than was actually listed. It has been a while since I've seen the page... My mistake...


The DB

What mistake? You could be right DB. Maybe an Imax with 1 wavelength long radials could just about double the gain over a regular Imax. Now we have two pieces of evidence that agrees with you vs. my model that doesn't agree.

This guy in Houston said his model with 2 x 5/8 wave slanted down radials was much better. I would guess a lot of folks will be trying this modification if you can confirm such results in your own testing or another reliable source. Since I recall getting bad results with this idea, I'll go back an recheck my model, maybe I did something in error.

View attachment Imax with 2 x .625w radials.pdf

 

[The DB]

What mistake? You could be right DB. Maybe an Imax with 1 wavelength long radials could just about double the gain over a regular Imax. Now we have two pieces of evidence that agrees with you vs. my model that doesn't agree.

This guy in Houston said his model with 2 x 5/8 wave slanted down radials was much better. I would guess a lot of folks will be trying this modification if you can confirm such results in your own testing or another reliable source. Since I recall getting bad results with this idea, I'll go back an recheck my model, maybe I did something in error.

View attachment 10143

I have argued multiple times that as you angle the 1/4 wave radials down on a 1/4 wave antenna that the antenna as a whole acts less and less like a ground plane vertical and more and more like a dipole. I see no reason why the same wouldn't be true for a 5/8 wave groundplane antenna with 5/8 wavelength radials and an extended double zepp antenna.

My only concern for this last link is it is in freespace, and not near a simulated earth ground. It would be interesting to see how a real ground being some distance under such an antenna would affect it.

Another thing to mention, I have several times in the past on this forum mentioned that I would like to play with a vertical extended double zepp, and every time someone says it would be a cloud warmer. I was never so sure about that as how I apply what I know of antenna theory disagrees with that postulate. The article you linked in the first post of this thread also disagrees with what I have been told by multiple people. In the "Where 5/8-Wavelength Pays Off" section it does state that such a vertical antenna will still have real gain on the horizon, however it does not provide any pattern of such a setup at any distance over real earth.

All that being said, there is definitely a difference between radials of any size and a theoretical "perfect" ground plane of "infinite size and infinite conductivity". Even with radials of 2, 3, 4, 5, 10, or 20 wavelengths long, it is still not the same, therefore the "mirror effect" that such a ground plane presents is incomplete at best. Therefore, if you limit yourself to 5/8 wavelength radials and angle them down as mush as you can, would that not help complete the "mirror effect" as you are physically (and electrically if I am not mistaken) closer to the ideal 5/8 wavelength antenna and its electrical mirror?


The DB