mf_fx, here is a link to a W8JI j-pole article and a model I tried to duplicate. I don't recall if I scaled it up to 11 meters or not.
http://www.worldwidedx.com/cb-antennas/150476-w8jis-2m-j-pole-article.html
-
You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
Battle of the Ground plane #2
- Thread starter 211inthamag
- Start date
mf_fx, here is a link to a W8JI j-pole article and a model I tried to duplicate. I don't recall if I scaled it up to 11 meters or not.
http://www.worldwidedx.com/cb-antennas/150476-w8jis-2m-j-pole-article.html
That is what the model shows and what I've tried to point out by dividing the currents into 4 units.
Close. It is obvious that the unit representing the radial current is slightly smaller than the others and I made no attempt to distort that fact. The image was blown up 8 times so that individual pixels could be seen and those with the same exact color were selected as the outside edge of the current limits for red shades.
If that is the case and those radial currents are in phase with the upper 1/2 wave, EZNEC should be reporting noticeable gain from the cone radiation. It should also recognize the phase of the vertical has been delayed by 90 degrees since the cone has shielded the bottom 1/4 wave of this vertical from contributing to the pattern and that radiation from the lower area has been replaced by "non apparent collinear" radiation that is phase corrected from the cone.
That also indicates the model should pass my radiated phase accuracy test by adding the extra 1/4 wave electrical length phase shift with another 1/2 wave collinear top and show about 2 more db. Something I would be very suspired if EZNEC were capable of doing without making the phase shift twice as long.
Now I'm a little confused because I thought you just explained your EZNEC model did show about 4/16th of an inch deflection in the current line of the vertical once you increased the source power? At the same time you noted the radials were 1/16th. This sentence seems to conflict with your last unless I'm not understanding you correctly.
Check the model for phase accuracy first with the collinear test. Those 4 wires I've spoken of many times and the length of the phase shift required, will instantly tell you if the EZNEC model is even in the ballpark of interpreting the currents in question.
When I looked at your last model it was evident that out of phase radiation was escaping over the top edge of the cone from the vertical inside of it. That out of phase radiation is a total of 1/4 wavelength. That indicates your cone is too short and the EZNEC model is more like an end fed 5/8 wave from the currents I see on that radiator. It would make sense that the gain in EZNEC peaks in this area since it doesn't show any sign of accounting for phase corrected cone radiation adding to the gain.
I'm sure the results are referenced against the 1/2 wave dipole by the "dbd" specification rather than isotropic which opens the door to more variables like you're asking about. Once you have the dipole as the reference point, the difference between free space and real ground are the ground reflections that change with different heights. The ground gain the dipole sees is effected the same way and most of that variable is eliminated once the dipole is compared at the same "center of radiation" height. The 2 dbd over real ground would be the same although it would have a narrower beam width when compared to the dipole over ground. That's where the gain comes from.
It's also normal that you would find many different omni directional antennas are producing close to unity gain in free space. I would expect everything from a 1/4 wave ground plane to even a 5/8 wave ground plane to be very close in free space. The 1/4 wave groundplane is a 1/2 wave once it sees its reflection in the groundplane underneath. The 1/2 wave is unity gain. The 5/8 wave has to deal with that 1/8 wave of out of phase radiation that combines with the pattern deconstructively and is easy to see in free space without the distortion of ground reflections.
What we see in EZNEC when modeling the Sigma is very unique. It may be the only antenna where EZNEC makes gross mistakes in underestimating 1/3 of an antennas effective radiated wavelength and over estimates the phase shift required to stack sections by 100% in electrical wavelength as a result. That clearly indicates the program has difficulties in identifying where effective radiation on the Sigma is and is not.
For EZNEC to calculate the phase delay being two times the length that works, that means EZNEC is seeing half of that electrical length in the phase delay as deconstructive in phase. That length in question is the 1/4 wavelength of phase corrected radiation emitting from the cone. EZNEC is blind to this aspect and only sees the cone as a piece of tapered coax feeding a 1/2 wave. The proof of this is you have to invert the phase a complete 180 degrees in EZNEC before you can stack another 1/2 wave on top and see gain in a program that is just about completely wrong when modeling the Sigma.
Last edited:
Well Donald we reached another break through. Now my concern is why my Eznec models in free space are not seeing the gain that Sirio reports for their 827, and the NV4K, which notes 4.15 dbi as noted here: CB 27 MHz Base Antenna, NEW VECTOR 4000
My Eznec model is showing 4.06 dbi, but this model is over real Earth, and includes a mast, and when I make the model in Free Space the best the model will show is 2.25 dbi at 17* degrees. This is my dilemma now.
I don't see that fine of definition, but my dimensions are close.
This is the issue Donald, that I'm still having trouble with, but I'm patient.
Donald can you share your Eznec collinear model with me? The other day I attempted to follow your lead and try and duplicate your instructions, but I think I failed at that, and we've not gotten back to that idea.
I think you misunderstand me here. That 4/16th was measured at the very bottom of the radiator for segment #1. It is a measurement between the 1/4 out of phase red line in my model that passes thru the cone and out to one side. My ruler measured a bit over 4/16th from the end of this red line back to the base of the monopole at segment #1. When I go up to the top 1/2 wave portion of the radiator and find segment #73, I measure a bit over 2/16th at that point. Now when I change the power in watts to 500 from the Eznec default value for watts...that does not suggest any change in the red lines for the model. Some would wonder why, but it may be only a convenience of reducing code to handle every possible iteration in power available. IMO, this indication related to us as a visual is related and the net value of the power probably does not change as a relationship...so there is no need for the antenna view to change...regardless of power. Just a guess however.
Again, do you think your Eznec model that shows you claim here could be helpful to me? I am by nature persuaded more by the visuals than the words in all cases. My mind's-eye is perceptive, but not very creative.
I hear you Donald, but this is our bottle neck, because I see the currents originating at the bottom of the radial, with the current distribution radiating up each radial just like we see with any end fed 1/4 wave wire.
On both models with 90.5" radials, I see only the bottom portion of the radial being phase and about 1/2 wave up the current collapses and changes to being out of phase. What I see at the very tip end of the radial is very little current flowing into the hoop, and the phase shifting into the hoop switches phase every 5 segments (5x2.35"=11.75" inches) on the hoop.
With the 107" radials, I see the full radial being in phase with the top 1/2 wave radiator, and I would expect that to be better, but the model looks to fail in that configuration, and I have not even a guess as to why that may be.
In fact I was surprised when I discovered the difference to be noted between my S4 to specs and my Vector to specs and that is what lead me to my MOD (Modified) version for the Vector. That said however, the S4 is still showing a bit more gain that the Vector. And, when I check the match the S4 shows a better natural match than the NV4K, so I don't see SWR affecting gain in a positive direction, like I often do with my models.
No comment.
No comment.
No comment.
I don't know how the following info relates, but the radials in my S4 model and my NV4K MOD model shows there is a phase shift on the radial at the
a little past the 1/2 way up toward the top at about segment 16 of 35 on the Vector and at 18 of 29 on the Sigma4. The difference here was an error I made when I made the NV4K's radials shorter and forgot to change the segment count. That may make the difference I see in the gain between these two. I only discovered this when I was checking to answer this issue.
For what it is worth, the NV4K built to Sirio's specifications where the radials are 107" long...shows the radials are fully in phase with the phase of the top 1/2 wave radiator. But that does not provide for better gain, it is the opposite. In fact as I make models more accurate I tend to always see the model show a bit less gain, and that may be the case here.
I wish I understood more Donald, but I'm still stuck I think. But that said, I think I'm getting closer.
You know what really amazes me is the fact that Bob told us that Cebik did not wish to discuss further about the S4 design. I would think if this one does exactly what is suggested here that would be like a break thru in antenna design that shows approximately double the dbd gain on and antenna that is fully plumbers delight, and requires not insulators or a rather large phase reversing stub.
Even though I don't fully understand all the business about the phasing going on in the antenna...you are finally sounding very convincing, both for what it does and how Eznec does not handle it.
I will try and talk to Henry again, but I don't hold out hope I will understand him any better, but I'll give it a try.
Thanks,
The goal here is to see if anyone can possibly build an accurate model of the Sigma in EZNEC. I already confirmed every model I've seen so far is not worth the time it took to make including the ones I've worked with. Why would you want a model that has already failed this test? I could see if adding the 4 wires to any model were difficult but that's not even close to the case. If you can build the Sigma in the program, adding the 4 wires should only take a few minutes.
I'm sure the EZNEC models of the Sigma I have can not be of any help to anyone. Every one has failed the test I'm suggesting. I'm finding it difficult to believe a guy who can model several antennas in a single day is having difficulty adding 4 wires to an existing model. If that's truly the case, you can send me your model and I'll add the 4 wire for you in a few minuets
The more I participate in these Sigma debates the more Cebik seems like a genius to me. To have the insight to understand that most people would not see this "non apparent collinear" action and to see it as generating pages of debate, where would the temptation to engage in further conversation come from?
that's pretty much why I just read the comments and completely ignore a certain "expert"
.
that's pretty much why I just read the comments and completely ignore a certain "expert"
I'm just here for the entertainment factor.
op:We'll earlier you asked me to model a collinear model. I did two different ones trying to get the model as you described so we could continue to talk about what that model showed you. But as of this late date, I was thinking we had never finished discussing what you meant with that idea.
As you know the first model I did...I thought you were describing two S4's, one above the other in a collinear setup. But I guess that one wrong. I'll go back and read this thread again. Maybe I missed something.
So, when I asked if you could send me your model, I was thinking we had not finished that point you were suggesting...whatever it was. I was just trying to cut to the chase.
Donald again, I think you missunderstood what I was asking you to share and for sure why.
Donald I agree that Cebik was a genius...not because he choose NOT to discuss and write countless articles on topics concerning antennas, but because he did do all that stuff we've been reading.
When Bob told me his story, it just didn't sound like the Cebik that I came to know over the years. Among a whole host of ideas that interested Cebik in his life...he was prolific in his production of stuff regarding antennas and much more. He told me that Cebik refused to talk to him any further about what very few in this world understood about the S4 design.
Maybe Cebik was sick.
You told me earlier that you discovered that Eznec would not or could not predict the botton radial cone as a 1/4 wave radiator when you modeled the S4 as a collinear...I wanted that model and maybe by looking at what you had done and how you described what you saw...just maybe I could get a minds eye on exactly what you've be trying to tell me for some time.
Does this clear up why I might ask to see your collinear S4 Eznec model?
Here is a commet that I posted some time back before post #51 in this thread.
Marconi said:
Marconi, please refer to post #22 in the "GHZ24 what is your opinion of the Sigma 4?" thread. After seeing the two Sigma's tip to tip it became clear you did not add the four wires described. You'll find a picture in that post of the four wires to add on your model.
He also knew when a topic would be productive and when it would create countless pages of meaningless debate. That shows wisdom in other areas.
I have to say when people share information here and you refer to it as a "story" it implies you think someone is not being honest. I find it offensive when you do it to me and I suspect Bob may feel the same way. The reason Cebik found it to be non productive has been demonstrated on this forum countless times already. What could he gain from trying to explain someone else's design that was so controversial and difficult to model prior to CST? Maybe a headache?
No, I said EZNEC fails to predict the cone radiation in the Sigma. The collinear version simply proves this statement in the field since EZNEC wants to see a full 180 degree phase inversion with a collinear. What that means is half of the electrical length of the phase shift is now wasting 90 degrees of in phase radiation in the delay when it should be freely radiating. That also places 50% of the next 1/2 collinear top out of phase due to the phase shift being twice as long as it should be.
This all proves the cone is a significant 1/4 wave radiator since the top collinear section that has half of its length bucking the phase of the cone can barely overcome the gain of the cone with a overall effect of no gain added in the collinear modification EZNEC suggests.
Not really. The picture in the other post will provide more than enough information for you to confirm EZNEC needs to see a 180 degree phase inversion to drive another 1/2 wave vertical radiator. What that means may be harder to determine without doing your own field testing to see how the model fails in the field and what it takes to correct it. On the other hand you might just believe my "story".
i never said CEBIK refused to talk to me any more about the sigma, we went back and forth a few times in a very pleasant manner i thanked him and went away looking for a mechanism for the in phase radiation,
the poor guy went silent key before i had a chance to ask him if the sigma could radiate from the lower 1/4wave by the same mechanism as open sleeve antennas described in the arrl antenna book,
it was the "pages of meaningless argument such antennas can generate because most people don't understand how they operate" that he said he did not wish to get involved with,
you can draw your own conclusions about the cst plot showing what CEBIK was talking about or not.
the poor guy went silent key before i had a chance to ask him if the sigma could radiate from the lower 1/4wave by the same mechanism as open sleeve antennas described in the arrl antenna book,
it was the "pages of meaningless argument such antennas can generate because most people don't understand how they operate" that he said he did not wish to get involved with,
you can draw your own conclusions about the cst plot showing what CEBIK was talking about or not.
Sirio GainMaster 3,5dbi no more??
My ASTROPLANE more than 8dbi !
nospec, I think if you check your ground settings, you'll see your model is producing the gain over Perfect Earth, with and Infinite groundplane. Try the real Earth or free space settings, or both, and you'll see the gain as more realistic per the antenna gain we see with our vertical CB antennas.
Here is my GainMaster set to over Perfect Earth, Free Space, and Real Earth according to Eznec's ideas and features.
View attachment GainMaster over real Earth, free space, and Perfect Earth..pdf
thank you very much Marconi.
Its plot is interesting, but the real-world antenna often will be much different if the model ignores the feedline and mast??
nosepc.
thank you very much Marconi.
Its plot is interesting, but the real-world antenna often will be much different if the model ignores the feedline and mast??
nosepc.
Yes, nosepc I agree, but if you want a pure free space model of the antenna design itself then the mast and FL have to go.
When I want to see the model that has all the obvious losses removed, I delete the mast and feed line if they are present. Then I set the model to NO losses that are due to material used...like aluminum, copper, etc. Then I remove the effects of Earth, by setting my model to Free Space.
With Eznec I can get an Average Gain result by also setting the plot type of the model to the three dimensional setting. This calculation gives the modeler an idea in a range of real number values just how effective the antenna design is. Using such numbers values, one can then recalculate the gain and resistance for the model...both upward and downward from the values reported in AG error results.
If we were relating to ham operators that generally consider theory more important than experience when modeling is discussed, then we probably would be reporting our models in purely Free Space most of the time. If the viewer doesn't understand this...then they don't understand modeling.
I know that published guru articles about modeling "cringe" at the idea that CBr's seldom include the mast or feed line in their models...and thus such models are worthless, and are all wrong. This is what drives their idea to be very cautious and critical of models produced by inexperienced modelers. I wish we had a tutoring program that could teach us everything we need to know in order to model effectively in just 10 days...like we hear can be done with a new language as advertised. Do you believe that?
Here is my simple model of a 1/4 wave that shows this difference, with and without the mast and feed line included in a FS model. The first model is over Eznec's feature for real Earth.
View attachment .25 wave antenna.pdf