Question on the Vector

[Henry HPSD]

Think fastes way for u would be:

copy paste the wires provided of "model DB"
into a .txt document (wordpad)

Open eznec.
Go to wires screen "press" other.
Go to import wires from ascii file .....replace exsisting wires
Go find your .txt document

done.

 

[Marconi]

Think fastes way for u would be:

copy paste the wires provided of "model DB"
into a .txt document (wordpad)

Open eznec.
Go to wires screen "press" other.
Go to import wires from ascii file .....replace exsisting wires
Go find your .txt document

done.

Henry I did as noted above.

The txt file uploaded and went into Eznec fine, but it does not look like an antenna, and I can't scan it, because Eznec reports the source is incorrectly placed. I converted from mm to feet and the model shows wires 1749' feet long and wire diameter 549" thick. This looks like the dimensions did not come over correctly.

I put the model into one of my models and selected...to replace the wires in my model. I did not check what dimensions either models were set at.

I'll try it again.

I would try to load it into 4Nec2 like I think you did, but I don't know yet how to manipulate the models easily.

 

[Henry HPSD]

pehaps the txt file is saved in "asci" ?
When I save the text for the first time by using the wordpad
I can save it as a .txt document and select : ansi as "code"
Thats default for me, but perhaps it is with you ?

Not sure if that solves it, i dont have much knowledge of computers

ps the source etc isnt copied using this methode.

Kind regards,

H

 

[Henry HPSD]

Almost forgot about necplot...

For fun:

 

[The DB]

I made a few modifications to the model to compensate for the change to aluminum, and the error message that Henry confirmed was occurring in 4NEC2. I had addressed this before but I'm not sure if Henry read either of those posts.

What I did:

1) I separated the wires some that was causing the error message. They are now twice as far apart as they were.

2) I adjusted the length of the vertical radiator for maximum gain, it went from 9.1 wavelengths to 8.5 wavelengths long. A note on this, it is now withing 0.3 wavelengths of the length that Donald has been recommending for years.

3) I adjusted the length of the mast, unlike the vertical element, this ones new optimum length is longer, at .23 wavelengths long, it is now very close to 1/4 wavelength long.

I do have some other adjustments to make, such as the radial lengths and angles, which are made up of four elements each that all need to be changed the same way, and this takes a lot more work than making an adjusting to individual elements at a time.

Here is a comparison plot...

The blue plot is my newly modified plot.
The red plot was the original model modified to aluminum with no other changes.
The green plot is a dipole reference made out of aluminum. It's center is close to half way up the antenna's height.

One thing I am noticing is in this version of the model, there is a very small amount of current in the basket area of the model. This is a variance to how the model looked before converting it to T6 aluminum. As my next step is making adjustments to the basket area, I wonder if that will make a difference.

Another thing I'm noticing is the efficiency has gotten better with these changes.

A question, how far apart do elements in NEC2 need to be for accuracy? I can't seem to find any data on this on-line.

Also, if anyone is wondering, the Average Gain Test for the current version of the model is 1.005.


The DB

 

[The DB]

pehaps the txt file is saved in "asci" ?
When I save the text for the first time by using the wordpad
I can save it as a .txt document and select : ansi as "code"
Thats default for me, but perhaps it is with you ?

Not sure if that solves it, i dont have much knowledge of computers

ps the source etc isnt copied using this methode.

Kind regards,

H

If you are talking about the file that I posted, I simply took the .nec file 4NEC2 created and posted it. To get the web server to let others access it as a text file I had to add a .txt extension.

When I get a chance to make a few more adjustments I will post another .nec file the same way so whoever wants to play with the updated model can. No telling when that will be at the moment, however.


The DB

 

[The DB]

It's agreed the model is still flawed. I'd like to see it represent the 90 degree phase shift in a collinear model and I believe DB said it does not. I still see success in learning about more than one current on the cone. Seeing we can form a 1/4 wave current with a peak in the center using a pair of currents. Learning that when extra wires are added around the cone for CMC to radiate in the model, gain goes up. I suspect as time permits we may see DB's model get closer. In the meantime these models have given us food for thought and opened insights to others that may not have been apparent before.

After looking at the pattern, I'm not sure what the distribution would look like if the currents were combined into one. The issue here is there are two currents on two separate elements that are supposed to count as two halves of the same physical element. It will take some hard looking and calculating to determine if that is the case or not. Long story short, it is to soon to tell, and I have other work that needs to be done on the model before speculating on such possibility would be possible or even relevant.


The DB

 

[Marconi]

DB, did you do your model that you posted as .txt file using mm or inches.

When I uploaded it to my Eznec...it comes across in MM?

I still haven't been successful at loading it into Eznec, but I've read a post on another forum that you have to add a line of instructions at the top of the .txt file that tells Eznec what the Units are.

Henry, what do you know about this?

 

[The DB]

All of my models are in meters. I just find it easier to work with meters when modeling.

If you would like, I can give you all of the endpoints of all 26 wires and you can enter them manually. Again they are in meters.


The DB

 

[The DB]

I've done some more work on the aluminum version of the model. This version of the model has gain pushed as high as I can by adjusting the dimensions of the antenna itself. Here it is in comparison to an aluminum dipole reference.

As you can see, it doesn't quite have the 2 dB gain over the dipole reference, but 1.77 dB in gain isn't bad. Another thing to note, I have been keeping track of the efficiency numbers, and this version of the Vector model is actually more efficient than the aluminum dipole reference, although not by much. The AGT is 1.005.

On to currents in the basket area...

Donald, when it comes to the currents and the 1/4 (or in this case near 1/4 wavelength sections) on the radials you may be interested to know that the elements for the inside portion of the radials has a current distribution similar to the 1/4 wavelength section with sources 90 degrees out of phase. You can see the characteristic dip in the colors above for those wires. However, the outside of the radials more closely resemble, although not perfectly, two sources on a wire that are 1/4 wavelength apart, and are in phase with each other. You can't see that here, I had to look at the current distribution numbers of the element directly to see it. The peak is closer to the bottom of the element as opposed to the middle, but there is a rise and fall to the currents. In my opinion, this current appears to match the CST image for the outside part of the radials even more than our test models with sources on either side of an element from before.

Another thing of note with this model, and I noticed this before when I was initially working with the model before the aluminum conversion that Henry suggested. When I lengthened the radial sections I expected to have to shorten the mast length to compensate for the additional length, however, to achieve maximum gain with the radial length modification I also had to lengthen the mast section. This is not intuitive to me as I initially thought to compensate for making an element longer I must make another attached element shorter. There is clearly something for me to learn about how NEC2 is handling currents in this layout.

Just to make a note, yep disclaimer time. This model is an experiment designed to test an idea that has been proposed and see what would happen. I, the creator of the model, am not saying this is definitely the solution to what is happening, but it is a step in a direction. I hope it is the right direction, as it is not my intention to complicate a complicated topic even more. As always, I am interested in any suggestion that might help me create a more accurate model to test this idea, or any known flaws I have overlooked.

And finely, as I mentioned I would above, here is a link to the .nec file from 4NEC2 (I renamed it to add a .txt extension so the webserver would let you have access to it).


The DB

 

[Henry HPSD]

Above model in eznec pro4 3,7dBI average gain 0,789....-1,02 dB ( roughly ...2,7 dBI left.

Not sure what the minimum distance is in aspect to closed spaced wires, somewhere in my mind the figure 0,1 wavelength keeps comming up.
I could ask around.

It is my interpertation that the text "difficult to model" is due to the closed spaced wires and the angle that they have.

@ Eddie, no idea m8....
Im sadly im not really into computers i dont understand "computer code" .
Simple things like converting a .txt into something else, makes my head spin hihi.

 

[The DB]

I think I must add another note on the models above, and really all of my models. I don't use the stock NEC2 files to compile data in my models. I use a rewritten version that has some changes. For one it is multithreaded, which means it can use more than one core of my processor at a time, which means it is faster. It also has some bug fixes within the code. One of the bug fixes takes care of a known bug when using a ground that isn't a perfect ground, for example.

When running this model with the stock code I get a slightly higher gain than the version of the code I prefer to use. The stock code shows a gain of 5.35 dBi as opposed to the 5.31 dBi you see above. It also has an additional 0.04 dB in gain over the aluminum dipole reference as compiled with that core.

Just though I should put it out there.


The DB

 

[bob85]

henry,
from what i have read close spaced wires in nec2 should be several wire diameters apart and a few diameters above ground if used as radials over ground,
i don't know how that effects the model as in the real antenna the space at the lower end of the sleeve is about the diameter of the monopole in the sigma and less in the vector,

im interested in understanding what barkley claims about thin sleeve elements and taper increasing transmission-line loss by radiation,

 

[Marconi]

im interested in understanding what barkley claims about thin sleeve elements and taper increasing transmission-line loss by radiation,

Bob, I don't find this specific claim by Barkley in the OSA article perse, but it sounds like what I have described when I talked about a pigtail connection disrupting the proper construction ratio of coax and causing the cone to radiate antenna mode type currents rather than CMC's like has been suggested.

My old antenna mentor use to always tell me that as soon as you made a pigtail connection, the pigtail started to radiate.

 

[Marconi]

All of my models are in meters. I just find it easier to work with meters when modeling.

If you would like, I can give you all of the endpoints of all 26 wires and you can enter them manually. Again they are in meters.

The DB

DB, I'm not complaining. I just wanted to know if you did your models in mm, because that is what I saw, and I wasn't sure why the Eznec failed to copy your wire descriptions correctly.

I think I already have the end points in the text file you posted. Thank you though.