Question on the Vector

[Marconi]

I may have over simplified my analogy with any two waveforms combining since all 1/4 wave radiators start with maximum current at the source and always go to minimum at the far end. Now what happens if we put two currents on the same 1/4 wave BUT their sources are at opposite ends of that 1/4 wave radiator?

Is it not possible for both ends of that radiator to come to a current minimum since each is an end for one current? As we reach both ends of the radiator, the phase shift between both currents might just be nulling the currents out at the ends. It is the only logical explanation I can arrive at for what we see although, I'm all ears if there is another plausible theory.

Hey Donald the discussion currently going on here is about speculations...NOT THEORY.

 

[Shockwave]

Hey Donald the discussion currently going on here is about speculations...NOT THEORY.

Marconi, run the "speculation" through your trusty EZNEC and when you get the results, be sure and place them where I suggested in a previous post you doubting Thomas. Now it's becoming fact rather than an anomally.

 

[The DB]

Thanks DB, I'll try and upload this to Eznec. Is this your model before you modified it with the extra wires?

I know that Bob and Donald are excited about your new information that might help explain to us how the S4 design might work, but you must remain objective and don't go being subjective on us.

I appreciate your being objective here and having questions, I'm sure both Donald and Bob would agree.

Thanks.

DB, I see you doing a lot of other work trying to justify what we see in your report, but you have an error in your application, and it is not so obvious to show some glaring results...except you see a gain that is too high for this antenna. At this point it is easy to become subjective.

If you check and recheck your model you might find out something is going on that you don't realize yet. I think you know the problem if you see it, but all of this other stuff...that might come into question later, is right now a distraction.

I mentioned an error above about segmentation in the wires of the radials, and why I allowed the occurrence to stay. The segmentation error I saw is in fact not accurate, and I think has to do with the angle of the wires in question. The wires being compared are exactly the same length, exactly the same thickness, and have exactly the same number of segments, and by extension the exact same segment length. It may or may not be a problem in the end, but it came about when I was adjusting the angle of the radials. Honestly, the gain difference is minimal (the range of less than 0.2 dB difference if memory serves) from an angle that does not have that error to the current angle. If, however, you are referring to a different error, I'm not seeing it.

When it comes to objectivity, my posts above should be enough to confirm that I am not 100% behind this model, even though I made it. Certain parts I will defend, others I won't. I have mentioned several observations that show a level of unease with this model. It is, in my opinion, the best model as of yet to describe what Bob and Donald are talking about. It was/is an experiment.

Actually, I wasn't fully ready for the model to be mentioned or released to the public yet. What happened happened, and now I find myself in the hot seat...

I see Henry is here. If he could see the pattern that makes the gain you report, he would likey know in a second what the problem was.

I would love to know Henry's thoughts on this model. That being said, at the same time I wouldn't blame him on not wanting to dive into this mess...


The DB

 

[Marconi]

Marconi, run the "speculation" through your trusty EZNEC and when you get the results, be sure and place them where I suggested in a previous post you doubting Thomas. Now it's becoming fact rather than an anomally.

Donald you have no idea what has happened. This is not an anomaly, it is an error in reporting. Your bias has so overwhelmed your thinking that you do not even question any results now much less Eznec results, results that show more gain than Sirio reports using CST.

You are going to be embarrassed again.

You won't even consider the truth here after I describe the error that DB has made. I would think this idea was great to learn about Eznec...but there is something going on that DB does not realize at this point.

I would rather he discover his error before I have to point him in the right direction. Doing so is always a better lesson learned.

 

[Shockwave]

Donald you have no idea what has happened. This is not an anomaly, it is an error in reporting. Your bias has so overwhelmed your thinking that you do not even question any results now much less Eznec results, results that show more gain than Sirio reports using CST.

You are going to be embarrassed again.

You won't even consider the truth here after I describe the error that DB has made. I would think this idea was great to learn about Eznec...but there is something going on that DB does not realize at this point.

I would rather he discover his error before I have to point him in the right direction. Doing so is always a better lesson learned.

I was unaware of being embarrassed the first time. It seemed you had the market cornered in that area with all your misinformation. Let's see who is embarrassed when you make a 1/4 wave radiator in EZNEC and place a source on both ends Mr. Speculation. You might learn everything you don't understand is not a simple error or anomaly.

 

[The DB]

DB, Interestingly enough, I tried to model my idea in EZNEC and I see some results. Turns out if you make a single 1/4 wave element and apply a source at BOTH ends, you'll get something very close to that odd shaped 1/2 wave pattern over a 1/4 wave radiator just like I expected. Not sure why I didn't think of this idea before.

That is a thought, I did the same in 4NEC2 and got this result:

Note, the current never comes to zero, and its not shown here but the entire element is in phase as well. It also looks like both the tip of the yellow line near the maxima in its sign wave out to 1/8 wavelength in both directions (1/4 wavelength total) in my previous graphs above, and I think it more closely resembles what I see in the Dominator animation as well. They never looked exactly like I would expect from a half wavelength element, but instead, now I see the radiation from the radials more closely resembles the middle part of such an element.

This pattern might assume both sources are in phase with each other, which is questionable weather that is the case. I have some thoughts on this going both ways, but I need to do some thinking and perhaps some more analysis on this.


The DB

 

[Marconi]

DB, I sent you a PM.

 

[Shockwave]

For some reason EZNEC+ does show both currents coming down to zero. Although the curve is more sudden here than on the CST model. I think that's due to the absence of the opposing current inside the basket but not sure. In any event, EZNEC+ does show the ends currents down to zero and your 4NEC2 is showing maximum current in the center of a 1/4 wave element. It does not appear that you need to add a phase shift between the sources since the phase shift between the two currents is a result of their sources being at different ends.

 

[Marconi]

DB, I sent you another PM.

Now you guys are trying to put a source at both ends of a wire to see if that proves some point.

If you can't get to it right away...I will post my report. Here is the report I intend to post in the thread if there is a long delay. IMO, it shows you that you likely made and error in your model's settings when you added the wires.

This is speculation too.

 

[The DB]

For some reason EZNEC+ does show both currents coming down to zero. Although the curve is more sudden here than on the CST model. I think that's due to the absence of the opposing current inside the basket but not sure. In any event, EZNEC+ does show the ends currents down to zero and your 4NEC2 is showing maximum current in the center of a 1/4 wave element. It does not appear that you need to add a phase shift between the sources since the phase shift between the two currents is a result of their sources being at different ends.

I saw the same thing initially, then I revised the model in question to the one I posted above. The reason for the revision is that wire tips that are not connected to anything always have 0 current. They are voltage nodes by definition. The proximity of the wire tip appears to be part of the overall pattern, when in fact it is outside of the area of importance, which is the area between the sources. To show only that area I used additional wires in 90 degree angles to move the current data that is outside of the sources off of the main line.

Changing the phase of the sources does effect the pattern we are seeing. for example, if the sources are 180 degrees out of phase, there is a current null and associated phase shift between the sources. However, this is an extreme example, and I believe outside of what this antenna should be capable of simulating using only a near 1/4 wavelength element.

More likely a 90 degree phase difference between sources would be accurate. I'm still thinking about what I'm seeing with that one.

In spite of all of what was said above, for me to see this line of thought as plausible I need to know what would create such a source where there is no feedpoint. I am also curious what would cause a current maxima where we would expect to see a current null?


The DB

 

[Shockwave]

DB, I sent you another PM.

Now you guys are trying to put a source at both ends of a wire to see if that proves some point.

This is speculation too.

Other than placing a source at both ends, how do you suggest we simulate the CMC flowing down the outside of the cone? You could always open EZNEC and make a 1/4 wave to find the maximum current in the middle doing this, just like on the Vector cone.

 

[The DB]

Other than placing a source at both ends, how do you suggest we simulate the CMC flowing down the outside of the cone? You could always open EZNEC and make a 1/4 wave to find the maximum current in the middle doing this, just like on the Vector cone.

At 90 degrees of difference between the sources above instead of a peak between the sources you see a valley, although not a deep one.

Further I came up with a what if, dangerous thoughts I know, but that is how I analyze... What if the wires outside of the sources are different in length than these small stubs I used, so I modeled it. I made one of the small stubs 1/4 wavelength long. The currents between the sources are exactly the same.

What does all this this tell me? The currents present on the 1/4 wavelength wire between the sources only appear to be based on the phases of the sources in relation to each other, and not to the wire lengths outside of the sources.

I need to do some more thinking and experimentation on this, however, my head feels like its going to explode at the moment... Your making me think to much...


The DB

 

[Henry HPSD]

I would love to know Henry's thoughts on this model. That being said, at the same time I wouldn't blame him on not wanting to dive into this mess...


@ DB,


Perhaps u could run the model again.
But this time include material losses.

- Go to the new editor menu
- Select source / load (select "show loads")
- Adjust "type" to wire conductivity
- Adjust "cond" to aluminium t6

Thats it...now im moving myself away of the mess again hihi.

hope it is of use...

Kind regards,

Henry

(added: "select show loads" to make it more clear)

 

[Marconi]

DB, I have no problem with speculating on problems, but my whole point is you have your model set with the Ground Type = Perfect. I'm not referring to soil types.

Set your Ground type to Zero, and leave the soil type alone. Then remove one wire at a time and check the results. I think you will see very little change in the gain, angle, match, and SWR. If you remove the other 3 wires I think you will see the same or very similar results from the model.

If this happens with you like it does with me...you should realize the added wires do nothing.

The model I posted does not have added wires.

 

[Marconi]

I would love to know Henry's thoughts on this model. That being said, at the same time I wouldn't blame him on not wanting to dive into this mess...


@ DB,


Perhaps u could run the model again.
But this time include material losses.

- Go to the new editor menu
- Select source / load
- Adjust "type" to wire conductivity
- Adjust "cond" to aluminium t6

Thats it...now im moving myself away of the mess again hihi.

hope it is of use...

Kind regards,

Henry

Henry, I guess I was wrong about you seeing this problem in a second, and that was unfair of me to state like I did. I apologize.

I could not see the problem with DB's Average Gain results at first sight either. DB has not published this model yet, so nobody could really tell. I was speculating as what his problem was...after I attempted to duplicate what he might have done by adding some random 1/4 wave wires out beside the radials on his Vector 4K model.

DB's model for Ground Type is set to Perfect and that is why he see the gain he reports.