Battle of the Ground plane #2

[Shockwave]

Now I see what you're getting at Bob. I think I may have some of the answers too. The way it was explained to me is that individual currents can be seen by the colors on the respective elements themselves while the combined pattern can be seen as the radiation propagates away from the element and the individual shades of color blend together moving outwards.

It was also stated that the individual currents displayed on the cone radials are divided by the number of radials and the mast. All of which radiate in a constructive phase. Which makes each radial appear weaker than the vertical radiator while the combined field from all radials is significantly stronger then what is displayed on any one.

That was explained to me at the same time the CST models were given to me. After looking at it more, I noticed some interesting things. The loop has been omitted. I believe this was done for clarity since it is a known fact the loop does not alter the performance noticeably.

I also think 2 of the 4 radials have been omitted for the same reason. One would be out of view from behind and the other would be covering the main radiator. If I'm correct then each radial is only supporting 50% of the total currents. This too should not make a significant difference in the total field.

Since I noticed these things long after the project I was working on was completed, I did not reopen the discussion with the engineer at Sirio. I'd really like to find someone else with CST willing to experiment further. I've already asked a million questions from my source and don't intend to wear my welcome out.

Bob, after watching the CST video a little more I spotted something I missed in the original still photo. I believe I can now see the separate opposing currents forming on the inside and outside of the cone. If you look carefully, just as the phase flips and before the currents ramp up, you can see the inner portion of the cone matching the color of the vertical radiator before the currents get strong enough to combine and look like one inside the cone.

 

[Marconi]

My comments at this point are only going to be useful with regards to the EZNEC instructions since we obviously have had much different experiences with the basic Sigma design.

I see these instructions as acknowledging the fact currents can be inaccurate depending on how you define the ends of a given wire. I also see this as an attempt to assure one that this is nothing to be alarmed about as though it doesn't make a big difference.

Perhaps that is true but it does not take into account what is happening in the Sigma model. Here we have a situation where current phase is critically important to the electrical length of a phasing network and causes a difference in length that is 100% greater.

That strongly supports the possibility that the stock antenna modeled in EZNEC may not be interpreting the fact that the cone is a 1/4 wave radiator in itself and that it physically causes a 90 degree phase delay in the currents radiated on the vertical radiator.

Donald, let me repeat, what I and others do when modeling is try and predict how and what an antenna with similar dimensions, height, etc., might do.

The only thing in this world that is conclusive evidence as to what a model will do would necessarily have to be tested in the real world. So if you did that right...as far as I'm concerned that is good enough for me.

I trust you my man. I'm sure I don't know enough to decide at this point anything except what I see and think. If that sounds like an echo, maybe it is probably the way you see things too. If there really is a big difference in the results we see between Eznec and CST, then the ball is in your court and your golden.

You said in you earlier post at 7:09PM this day the following:

What I can say is there are many strong indicators that CST is interpreting the currents in the cone properly. The fact it reports gains around 2 dbd while EZNEC struggles to show much better than unity gain tells me it recognizes the cone as a 1/4 wave radiator.

Could you tell us where you see that gain info?

Has Sirio published more on CST stuff for their New Vector 4000, than I know about?

Too bad the image doesn't indicate all the details that Eznec shows along with its pattern images.

Do you know for sure that the image of the Vector is in free space?

If this is all true then you are right about the gain only being 2.49 dbi for my model in free space. As I have improved on the dimensions for this model I continue to see the gain drop. That has been my observations for a long time, the more error in the model you are likely to see more gain. I haven't read GHZ24's thread for understanding yet, and I told him why. I do recall him asking me if I have been checking the gain for my free space models. The answer is no I haven't, but tomorrow that will change.

I told 24 thanks, but I was into this discussion, and I did not want to burn my candle at both ends right now. I thought the discussion was too important even though we had discussed this before...several times.

I hope to get back to 24 in a few days, when he gets more time off from his work. He is currently working about two jobs.

Donald over real earth I see 4.59 dbi or something close in gain at about 32' feet.

This info on the NV4 is more information that I was not aware of. The db gain in free space is also news to me. I did not see that on the pattern image you posted, but that is pretty typical as Internet image of this software stuff goes. I'll have to think of that a bit and start checking to memory all of my other antennas gain figures in FS like I sad.

For the longest time I thought that animated model you posted was over real Earth, and not in free space. It does not look like any vertical free space model I seen. My bad I guess.

See Donald if I don't talk to someone about all this stuff, my ideas get skewed a bit don't you see. This would be a lot easier if you guys would tell us more that look at this image, it explains every thing that happens on the Sigma4 design.

View attachment Vector in Free Space.pdf

 

[Marconi]

what im wondering is are the currents in the lower 1/4wave shown at 1/4 their effective current with the 4 radial vector, or is the cst plot showing total field strength of the combined radiation from the 4 radials,

that would make a significant difference to how impressive radiation from the lower 1/4wave looks compared to the upper 1/2wave on the cst animation.

Well Bob, I can only speak for what I see on Eznec. But that is exactly what Eznec shows in the Currents Log.

In the antenna view we see the red line much closer to the element at the base than at the red line on the base of the monopole. If I have 4 radials then I can imagine the currents are roughly 1/4 of what I see max on the radiator. I made many notes to this effect when I post my Currents Log, but every bodies eyes just cross in disgust expecting that to be read and considered. That is just the way it is.

I think theory tells us if the feed point is balanced the sum of the radials should equal the magnitude of the current on the monopole at the same point. I always compare current on the current log at the first segment, figuring that the currents are somewhat proportional across the current distribution. I would be surprised that you don't know that already.

I also try, in such cases where the elements are close and parallel to get the elements at about the same segment length. I've written about this too.

Lots of "if's" in there, but Bob you have discussed this issue regarding Eznec before and I think you get the picture.

I didn't get you idea even close, but I think you know what the answer is, the Eznec Manual can fill you in on all of the details. Or, you can check this all out when I finish and post the details for the recap I sent you earlier...if you got it.

I'm kind of waiting and hoping you will straighten me out on the math issue I mentioned earlier...the one the recap covers, before I waste every ones time with something that is a mistake.

 

[W5LZ]

Marconi,
I said I wouldn't pay any attention to your posts but I've been following this discussion. I'm going to say this and really quit it. You are paying entirely too much attention and credence to your modeling. You are putting way too much faith in EzNec and it's just not worth it. The author says there are problems with it, so why aren't you giving any consideration to him? EzNec does give a 'direction' to how it all works but it's NOT infallible, it does have 'qwerks' that have to be considered. Don't limit your self to one view point, only looking at things in one way. That 'way' isn't always going to be right.
That's it, I quit, won't comment on anything else in this thread. It's gotten to the point of being ridiculous.
Good luck and have fun. But most importantly keep an open mind about things. Don't get stuck in the wrong 'rut'.
- 'Doc

 

[Shockwave]

Donald, let me repeat, what I and others do when modeling is try and predict how and what an antenna with similar dimensions, height, etc., might do.

The only thing in this world that is conclusive evidence as to what a model will do would necessarily have to be tested in the real world. So if you did that right...as far as I'm concerned that is good enough for me.

I trust you my man. I'm sure I don't know enough to decide at this point anything except what I see and think. If that sounds like an echo, maybe it is probably the way you see things too. If there really is a big difference in the results we see between Eznec and CST, then the ball is in your court and your golden.

I thank you for this vote of confidence. It's also not just me who has tested the design. Many of my clients have conducted their own tests that are in agreement. Some of these tests were done using a calibrated FIM-71 field strength receiver. An oldie but a goodie.

You said in you earlier post at 7:09PM this day the following:

Could you tell us where you see that gain info?

Has Sirio published more on CST stuff for their New Vector 4000, than I know about?

Sirio has published the gain of this design at 2 dbd for many years now. It appears they are not publishing the CST models for this 30 year old design and are using this software to promote the new designs they hold patents on like the GM. They use both the CST software and the on site test range to confirm these gains. To confirm free space gain in field tests, you just compare the antenna under test to a 1/2 wave dipole at the same height so ground gain is equal. The results should be close.

Too bad the image doesn't indicate all the details that Eznec shows along with its pattern images.

Do you know for sure that the image of the Vector is in free space?

I know the published 2 dbd gain is without any ground reflections adding to the gain. Free space modeling is the industry standard for all computer modeling. Anything done over real ground would be highly misleading with the ground gain added and the fact it would vary at every different mounting height. It would also be showing gains several db higher if it were over earth ground.

If this is all true then you are right about the gain only being 2.49 dbi for my model in free space. As I have improved on the dimensions for this model I continue to see the gain drop. That has been my observations for a long time, the more error in the model you are likely to see more gain. I haven't read GHZ24's thread for understanding yet, and I told him why. I do recall him asking me if I have been checking the gain for my free space models. The answer is no I haven't, but tomorrow that will change.

At 2.49 dbi in free space the program is placing it at a mere 0.35 dbd. Not to far off from what GHZ has found with his EZNEC models at 0.5 dbd. I've also noticed the closer we make the EZNEC model to the real antenna, the more the gain takes a noise dive in the program. One common example I've seen is we have to model the main radiator shorter than it is in the field to get the gain to peak in EZNEC.

I told 24 thanks, but I was into this discussion, and I did not want to burn my candle at both ends right now. I thought the discussion was too important even though we had discussed this before...several times.

I hope to get back to 24 in a few days, when he gets more time off from his work. He is currently working about two jobs.

Donald over real earth I see 4.59 dbi or something close in gain at about 32' feet.

This info on the NV4 is more information that I was not aware of. The db gain in free space is also news to me. I did not see that on the pattern image you posted, but that is pretty typical as Internet image of this software stuff goes. I'll have to think of that a bit and start checking to memory all of my other antennas gain figures in FS like I sad.

For the longest time I thought that animated model you posted was over real Earth, and not in free space. It does not look like any vertical free space model I seen. My bad I guess.

I'm not sure this particular plot would look much different in free space or over real earth ground although, it would increase the gain if it were over ground. Since this is a near field model, I don't think you would see much interaction with the ground. That may only be apparent in the far field as it elevates the pattern above 0 degrees and the ground reflections combine with the pattern. From what little we can see as the signal propagates away from the antenna, it appears the maximum radiation is perfectly inline with the 0 degree plane. Characteristic of free space models.

See Donald if I don't talk to someone about all this stuff, my ideas get skewed a bit don't you see. This would be a lot easier if you guys would tell us more that look at this image, it explains every thing that happens on the Sigma4 design.

View attachment 11226

I'm doing my best to share everything I know about this design in an attempt to learn more. Even if that information may not be beneficial to me.

 

[Marconi]

I think today was a good Lord's day.

Thanks

 

[bob85]

that is what i meant donald thanks for clearing that up,

eddie,
i had to view that email on my old laptop as its an old email addy,
the first one you sent with the attachment had nothing there to view so the second made little sense either.

 

[Marconi]

that is what i meant donald thanks for clearing that up,

eddie,
i had to view that email on my old laptop as its an old email addy,
the first one you sent with the attachment had nothing there to view so the second made little sense either.

Hey Bob, that's alright. Once I hit the send button, I know nothing about what happens on its way over there...I don't even understand how these emails work.

It's about the same with antennas too...don't you think?

I want to try and explain my ideas for how I see currents working with Eznec too, and I guess I'll have to use the PM feature here, but I don't think it will handle PDF files, and I know that the email feature won't allow anything but text, so maybe I'll have to post it on the forum instead.

I'm not sure I want to do that either right now. Maybe I will wait a little longer...until I can find someone that can talk to me about how they think these currents should be interpreted and what they mean in these modeling programs.

It still amazes me how easy Eznec is to work, and hardly anybody around here understands how to enter in the data...that will make the thing at least show an image of the antenna described.

You guys do know Eznec and 4nec2 are free to get started, and I'll help anybody that wants some help? I'll help you guys with how easy that is.

I can help you with this simulation modeling program also similar to CST if you like pretty colors and graphic images of stuff. It's free too:

PRODUCTS » :: AN-SOF :: Antenna Modeling & Design Software :: Simulation has never been so easy.

Ole Grampa bearing gifts and it's not even Christmas yet.

 

[Shockwave]

If you're into the pretty colors, a box of Crayola's and a coloring book have been scientifically tested and proven to produce 90% more satisfaction with only 10% of the controversy. Less if you stay in the lines

 

[Marconi]

Now I see what you're getting at Bob. I think I may have some of the answers too. The way it was explained to me is that individual currents can be seen by the colors on the respective elements themselves while the combined pattern can be seen as the radiation propagates away from the element and the individual shades of color blend together moving outwards.

It was also stated that the individual currents displayed on the cone radials are divided by the number of radials and the mast. All of which radiate in a constructive phase. Which makes each radial appear weaker than the vertical radiator while the combined field from all radials is significantly stronger then what is displayed on any one.

That was explained to me at the same time the CST models were given to me. After looking at it more, I noticed some interesting things. The loop has been omitted. I believe this was done for clarity since it is a known fact the loop does not alter the performance noticeably.

I also think 2 of the 4 radials have been omitted for the same reason. One would be out of view from behind and the other would be covering the main radiator. If I'm correct then each radial is only supporting 50% of the total currents. This too should not make a significant difference in the total field.

Since I noticed these things long after the project I was working on was completed, I did not reopen the discussion with the engineer at Sirio. I'd really like to find someone else with CST willing to experiment further. I've already asked a million questions from my source and don't intend to wear my welcome out.

Bob, after watching the CST video a little more I spotted something I missed in the original still photo. I believe I can now see the separate opposing currents forming on the inside and outside of the cone. If you look carefully, just as the phase flips and before the currents ramp up, you can see the inner portion of the cone matching the color of the vertical radiator before the currents get strong enough to combine and look like one inside the cone.

Donald, I missed this post and just now saw it.

Here are 2 images of my model for the NV4K that might shed a little light on the issues you're talking about above, but regarding how I see Eznec showing the same thing.

Below are 4 images of my Vector plus the tabular currents log that shows the details. I also made check marks at the points I noted data values.

The first set of two images is positioned just like I see the CST model positioned...with two radials hidden by the center monopole. One image has the currents turned off, so we can see the wires more clearly. The other model has currents and phase tunred on.

In this model I did not have to remove the wires to get this view, the wires are as you say...in front or behind the main radiator and hidden from view. So I don't think Sirio removed two radials for their image. The same is true of the radial loop at the top. There is very little current up there and that is why we don't see much going on regarding the currents in that area. To me either omission would mess up the results...if that is what you meant?

The second set of models shows my antenna view skewed (turned a little), so all the wires can be seen. The first image in this set is with currents turned off, and the second is with the currents and phase turned on.

If we check the currents log we can see that the currents on the bottom of the radiator are also clearly noted, and similar to the red line in the Antenna View. The monopole current is singular, and is very strong. This current is nearly twice as strong at the currents on the 1/2 wave radiator at the max point on segment #73, for wire #2, as noted on my notes.

I think we all can see similar currents indicated in the CST animated image as well, but there we see AC current alternating as the name implies.

On the radials we see similar currents flowing in my Eznec model, but maybe 1/4 the magnitude of the red line current as noted on the monopole. I have always understood that the currents on multiple radials are summed in order to see the full current effects they present to the far field. So you are right on that score too, so we need to multiply what we see on each radial times 4, in this case.

We can't see the phase of these currents noted in the images with Eznec, but that is one reason we have the tabular currents log.

On my currents log for wire #2, the monopole we see at the bottom segment #1, the current shows a (-) phase sign with 1.1108 amps. On the radials, wires #13-#15, we see currents at the same points, segment #1, where we see a (+) phase sign = .30862 amps. So the monopole and the radials are out of phase in the bottom of the Vector.

Theory tells us that radial currents are summed. Here is that sum calculated.
.360862 x 4 = 1.23448 amps with a (+) phase sign.

In this case the relationship of these wires is very close to being fully coaxial...just like Sirio tells us in their description of their NV4K.

IMO, this coaxial condition also suggest we will see cancellation just like we see in Maxwells work on coaxial line theory, and we can't get around this fact.

According to my math seen on the first image, we see that the results indicates a (-) sign phase difference of .12368 amps of current remaining after cancellation, and that current sign tells us this remaining current is in phase with the top 1/2 wavelength of the radiator, and we have constructive addition of the bottom of the Vector and the top.

However, due to the cancellation of currents in the bottom of the Sigma4 design...this magnitude left over is relatively small rather than being significant as you suggest. In fact this current magnitude from the radials is less that the current flowing at the tip of the antenna and/or the tip of the radials...and those current points too are very small.

Donald assuming we see the current maximum RF at 2.37 amps/m with the CST model like it indicates, what would you figure the maximum relative current radiating off of the 4 radials might be...being as you see these currents in the radial area as significant?

I have some more evidence that suggest how alike my model is compared to the CST results reported also. I will post that information latter if you are interested. You will be amazed how close my models gain results are compared to the CST model. I just saw this today and it surprised me!

View attachment New Vector 4000 demonstration.pdf

 

[Shockwave]

Donald, I missed this post and just now saw it.

Here are 2 images of my model for the NV4K that might shed a little light on the issues you're talking about above, but regarding how I see Eznec showing the same thing.

Below are 4 images of my Vector plus the tabular currents log that shows the details. I also made check marks at the points I noted data values.

The first set of two images is positioned just like I see the CST model positioned...with two radials hidden by the center monopole. One has the currents turned off, so we can see the wires more clearly. This other is with currents on.

In this model I did not have to remove the wires to get this view, the wires are as you say...in front or behind the main radiator and hidden from view. So I don't think Sirio removed two radials for that image. The same is true of the radial loop at the top. There is very little current up there and that is why we don't see much going on regarding the currents. To me either omission would mess up the results...if that is what you meant?

Guess what? I think you are correct and I should have stuck with my original thoughts that the CST model showed an individual radial current that was 4 times less than the total sum. That is the is the main reason I was excited to see the very first computer simulated model of this antenna that actually did reflect what I was seeing in the field for what was 13 years when I first received the CST model. It is also going to be a key point in my final attempt to convince you that your thoughts on this antenna are incorrect my friend.

Don't be offended either. I know where you're at with the time invested in EZNEC and how difficult or convenient you may think it is that I say no other program has demonstrated an ability to model the antenna that I have seen, other than CST. It costs something like $2,500 for some reason. Maybe because when you throw that 1 out of 100 designs at it that requires tons more software programming code to handle correctly, it doesn't get confused?

Rather than look at the pretty colors the CST model shows and applying opinion to what it all means. I'm going to try and show you why it's the "goose laying the golden egg" here. Maybe you'll see why it explains with a blink of an eye what took well over an hour to say in words. Lets break the color shades down into "units" so that we have something to compare.

If you look at the bottom of CST video frame #21 you will see 5 lines dividing 4 "units" that are fairly close in size. The first line is the beginning of the intense red shades emitted from the top half wave. The second line is the beginning of those same shades from one cone radial. The third is the end of the red shades for one cone radial. The fourth line is simply an equal spacing between the last two lines. The fifth line is the end of red radiation from the top 1/2 wave.

See anything interesting yet? Knowing that the individual radial current that we are looking at is multiplied by 4 times makes those combined radial currents almost equal to the upper 1/2 wave. I say "almost" because the unit that represents the radial currents is slightly smaller than the others. Almost equal in CST is vastly different than 10 times less in EZNEC. CST replicates what I see in the field with my clients every day. EZNEC did so poorly on this design I gave up trying to figure out why. That's Roy's job.

I know the cone currents don't cancel for many reasons. There is no other place for the gain to come from. CST clearly shows it is the source of this 2 dbd. The engineer that built the model in CST and the new Vector specifically said the cone "confines radiation from the vertical" and did not say it cancels. More evidence of this is when remove some of the radials, gain goes down. If cancellation was the dominant factor here, one radial with all the current on it would produce the effect you're thinking of.

The word "confine" is more like "shield" or "block" than it is "canceling". Shielding like you might put over the VFO tube in a transmitter rather than canceling in a parallel ladder line of equal lengths. That is an entirely different effect causing the same collinear result and is why Cebik called it "non apparent collinear". It removes the unwanted currents from the vertical by blocking them so the cone energy is not wasted in cancellation and allowed to contribute significantly to the upper 1/2 wave.

This is why I say you'll probably have to forget about most of these cheaper programs if you want to learn anything about the Sigma.

 

[bob85]

what you need is a model with 4x the current then splice that lower 1/4 with this upper 1/2wave, that would have eddie doing back flips lol.

 

[Shockwave]

what you need is a model with 4x the current then splice that lower 1/4 with this upper 1/2wave, that would have eddie doing back flips lol.

Easy Bob.....I want to break him in slow so it's not as shocking. Four times the radial currents would obviously be almost the entire group of units combined and equal to the upper 1/2 wave. That's the same message I tried to convey to CDX-007 two years ago and he never followed. I think the visual representation makes it easier now.

 

[Marconi]

.................



I have some more evidence that suggest how alike my model is compared to the CST results reported also. I will post that information latter if you are interested. You will be amazed how close my models gain results are compared to the CST model. I just saw this today and it surprised me!

Bob, I don't think you made your comment to Donald in gist. I understand that you have your opinion too. In fact, as far as I know, you're the only one that ever imagined the S4 might somehow be different. You developed plausible reasons, and you argued your point well over time.

What you told me made perfect sense because I believed you. But at some point in my modeling I got the idea that something just didn't quite make sense to me. I even spent part of a summer real world testing your idea, but was unsuccessful at coming to the good results you promised.

What I saw in modeling the S4 is the only thing I ever found that could be convincing and I wasn't ever totally sure of what I saw. It has taken me a long time to get to this point, but unlike you my friend...I humbly admit I could be wrong.

I would like both you and Donald to give me one last chance to try and at least convince you both that my model of the Vector shows almost exactly the same results as CST indicates in the pattern that Donald posted. Forget for now our differences in how the bottom of the antenna works.

I would further argue that this proof would not be convincing if the values for the maximum amps noted in Sirio image of their New Vector 4000, at the current maximum was not the same as my model shows.

Would you both consider my model of this Vector to at least be as accurate as the CST model for Sirio's NV4K...if I could prove that important factor between them both to be almost virtually the same?

Don't bother to look this up on the currents log I posted earlier...that model uses Eznec default power for the model, and Sirio controlled the power for their model. If you both promise to at least give this last effort your consideration, then I will fix my model to the same power and re-post the currents log for you both to check.

 

[bob85]

i don't know how much in phase current flows eddie,
enough current flows on the wide spaced j-pole to skew the pattern in the direction of the short leg,
enough current flows on the open sleeve antenna to compress the pattern from the upper 1/2wave,
if the vector manages a 360 degree version of that id be happy,

sirio claim 2dbd for all their sleeve type antennas and at lower radiation angles than a 5/8wave, i can live with that claim, i doubt the sirio engineers are far from the truth.