sigma4 article is online

[Marconi]

@ Eddie,
I understood completely. I just found a lot of humor in you being so perturbed over the difference in how you saw you and GHZ being handled. . .

Still funny

Homer, I never knew what was going on in this regard until this all came up recently, and I see now the real animus you guys felt towards me...and I don't like it.

Glad I made you laugh.

 

[Marconi]

Marconi, that is the figure I'm referring too of Henry's Sigma model on page 30 of the report. All of that high angle energy is easily refocused to a much more effective angle by adjusting the lengths of the two radiators and would be required before the model could be considered a Sigma.

Donald, I can't be sure, but that high angle results looks similar to the results that I see with GHZ24's model when he used 4Nec2 several years ago. I tried to explain this to Henry. I asked him to send his Eznec model too, so I could compare it to my Vector model using Eznec. I wanted to see what I did wrong.

 

[The DB]

DB, I only mentioned the option of shortening the top radiation because that's the easiest one to change in the model and see the effect. For sure it's the lengthening of the cone structure that provides the best gain in the distance. It is the one area Sirio, Bob and myself all agree on.

I simply adjusted the model itself to see if it conforms to both of these statements, and it does. In that regard, it is operating as you say it should. As a matter of fact, making the listed modifications both increase maximum gain, and lower the angle of radiation. As I said above it is possible to use either method to push the radiation angle below the "horizon" in freespace.

This of course shortens the electrical length of the upper section and requires it be made longer too. Since the cone must be similar to an electrical 1/4 wavelength, lengthening the 4 radials requires the loop be smaller and the upward angle to be sharper.

I haven't yet played with adjusting the diameter of the loop, however, the tools are available to do just that...

Neither of the two radiator lengths can be extended without the use of a capacitive matching network. Whatever gain people assume is being lost in that gamma match is insignificant in comparison to the gain that can be achieved in the distance with a gamma match. Understanding that's only possible since the gamma allows us to extend radiator lengths beyond resonance.

This I don't think is necessarily true. If I make various adjustments in combination with each other, I can maintain resonance, and in the process raise the ring area of the cone as well as shorten the vertical element length. As R is near 35 to begin with, I think I can get it near 50 ohms while still being reasonably close to resonance, and actually, then push R even higher if I wish. This process will also lower the angle of radiation from the antenna, and I think it will also have an increased overall gain as well.

One thing I haven't done yet is change the ring diameter. I mentioned above I think that system could have potentially been better. That, in fact, is the only thing I really don't like about the existing model as given. Instead of applying a multiplier to a 30 inch radius directly, I think it would be better to apply a multiplier to an inch radius. I think it would be easier to make consistent small adjustments who's meanings and lengths are easier to determine... That being said.


The DB

 

[HomerBB]

Homer, I never knew what was going on in this regard until this all came up recently, and I see now the real animus you guys felt towards me...and I don't like it.

Glad I made you laugh.

Now you can laugh at me, because I'm confused.
I like you, Marconi. Always have. and I have to look in the dictionary to see if I know for sure what animus means. What I read isn't how I feel about you, unless you mean the second meaning - like work on antennas.

an·i·mus
ˈanəməs/
noun
noun: animus

1. 1.
   hostility or ill feeling.
   "the author's animus toward her"
2. 2.
   motivation to do something.
   "the reformist animus came from within the Party"

 

[The DB]

No Henry, I won't be the one fixing GHZ24's model. If DB is going to use it to try and optimize it to check out Donald's idea here...then he will be the one to fix the model, and it does need some fixing. I am a little surprise however, that DB is using the Optimizer feature in 4Nec2, because I recall him claiming the feature was not reliable and maybe even a waste of time.

DB, am I wrong again?

I've played with the optimizer in 4NEC2 in the past, but I'm not a fan of it. I just like to control things by hand is all. It takes longer, but I can observe the changes as they happen and make adjustments on my own terms. That being said, I don't remember ever saying implying, or thinking for that matter, that that feature was broken...

Aside from what I mentioned above, I did notice the potential acute angle issue with this model (Henry actually mentioned it in his report). While I don't see this as a big issue as experience shows there to be little difference in what NEC2 results, I will likely address this before doing any serious work on the model. I also intend to add in a height variable to make adjusting the height easier for when I play with the model over an earth... Should I get to this point I'll happily release an updated text version of the file...


The DB

 

[The DB]

I made some changes to the GHZ24 model provided by Henry.

CM 11 meter cb by ghz24
CM Potential acute angle issue fixed, The DB, 2/3/15
CM Added a height adjustment variable, The DB, 2/3/15
CM Made the ring diameter changes more intuitive, The DB, 2/3/15
CE
SY z = 338    'driven element length in inches (need to add zp variable length to this)
SY rh = 61    'ring height above bottom of antenna in inches
SY zp = 12    'avoid segment errors
SY rsf = 30    'changes ring diameter, radius in inches
SY ah = 0    'changes the hight of the antenna, in inches
GW    1    1    0    0    0+ah    0    0    zp+ah    0.1226
GW    2    67    0    0    zp+ah    0    0    z+ah    0.05260011
GW    6    7    0    3    0+ah    0    1*rsf    rh+ah    0.05260011
GW    7    7    0    -3    0+ah    0    -1*rsf    rh+ah    0.05260011
GW    8    7    -3    0    0+ah    -1*rsf    0    rh+ah    0.05260011
GW    9    9    3    0    0+ah    1*rsf    0    rh+ah    0.05260011
GW    10    1    0    0    0+ah    3    0    0+ah    0.05260011
GW    11    1    0    0    0+ah    -3    0    0+ah    0.05260011
GW    12    1    0    0    0+ah    0    3    0+ah    0.05260011
GW    13    1    0    0    0+ah    0    -3    0+ah    0.05260011
GW    100    1    1*rsf    0*rsf    0+rh+ah    .92239*rsf    .382677*rsf    0+rh+ah    0.05260011
GW    101    1    .92239*rsf    .382677*rsf    0+rh+ah    .707126*rsf    .707126*rsf    0+rh+ah    0.05260011
GW    102    1    .707126*rsf    .707126*rsf    0+rh+ah    .382677*rsf    .92239*rsf    0+rh+ah    0.05260011
GW    103    1    .382677*rsf    .92239*rsf    0+rh+ah    0*rsf    1*rsf    0+rh+ah    0.05260011
GW    104    1    0*rsf    1*rsf    0+rh+ah    -.382677*rsf    .92239*rsf    0+rh+ah    0.05260011
GW    105    1    -.382677*rsf    .92239*rsf    0+rh+ah    -.707126*rsf    .707126*rsf    0+rh+ah    0.05260011
GW    106    1    -.707126*rsf    .707126*rsf    0+rh+ah    -.92239*rsf    .382677*rsf    0+rh+ah    0.05260011
GW    107    1    -.92239*rsf    .382677*rsf    0+rh+ah    -1*rsf    0*rsf    0+rh+ah    0.05260011
GW    108    1    -1*rsf    0*rsf    0+rh+ah    -.92239*rsf    -.382677*rsf    0+rh+ah    0.05260011
GW    109    1    -.92239*rsf    -.382677*rsf    0+rh+ah    -.707126*rsf    -.707126*rsf    0+rh+ah    0.05260011
GW    110    1    -.707126*rsf    -.707126*rsf    0+rh+ah    -.382677*rsf    -.92239*rsf    0+rh+ah    0.05260011
GW    111    1    -.382677*rsf    -.92239*rsf    0+rh+ah    0*rsf    -1*rsf    0+rh+ah    0.05260011
GW    112    1    0*rsf    -1*rsf    0+rh+ah    .382677*rsf    -.92239*rsf    0+rh+ah    0.05260011
GW    113    1    .382677*rsf    -.92239*rsf    0+rh+ah    .707126*rsf    -.707126*rsf    0+rh+ah    0.05260011
GW    114    1    .707126*rsf    -.707126*rsf    0+rh+ah    .92239*rsf    -.382677*rsf    0+rh+ah    0.05260011
GW    115    1    .92239*rsf    -.382677*rsf    0+rh+ah    1*rsf    0*rsf    0+rh+ah    0.05260011
GS    0    0    0.0254
GE    0
GN    -1
EK
EX    0    1    1    0    1    0    0
FR    0    0    0    0    27.18    0
EN

There are a few changes in this code over the GHZ24 version.

1. It ads a height adjust variable
2. It resolves issues that may arise from the acute angles between the radials and the vertical element
3. It changes the ring size code to be more intuitive

For those who wish to play with this model you will see several lines near the top that start with "SY". That is a variable indicator. The variables used in this file are z, rh, zp, rsf, and ah.

• The length of the vertical element is the z value.
• The height of the ring above the base of the antenna is determined by the rh variable in inches.
• The rsf variable is the radius if the ring in inches.
• The ah variable is how high you want the antenna above an earth in inches (should you have one).

Sizes have been adjusted to be as close to the GHZ24 model as possible. Some dimensions are off by approximately .00005 inches.

I hope this is useful to others who wish to play with this model.


The DB

 

[Marconi]

Homer,
Eddie is not talking about vector models, There was no coals, it was a warm soft blanket.

Yes Homer, Bob is right...I think I was modeling the Starduster that DB ultimately did for Bob in a collinear stacked setup...similar to what Donald had suggested to us in the past.

I was telling Bob, that my collinear model using a 1/2 wave SD'r with a 1/2 wave radiator above it connected via a 180* degree phasing stub...was No Way No.

I was also saying that maybe I was going to change my mind about the SD'r really being a center fed 1/2 wave antenna. I was also suggesting that maybe this revelation was what Donald had been trying to describe all the while, but no body was listening...and that went over like a lead balloon. So, I stopped messing with the SD'r collinear model, and continued on with my A/P idea in a collinear setup instead.

DB and Bob both claimed that DB's stacked collinear model showed something like 6.00+ gain at a very low angle to the Earth, and I knew my "Goose" was in the fire again. I wasn't about to start asking questions...because I knew that would be fruitless, and just piss everyone off again...like I did earlier when I was talking about my ideas for the Sigma4/Vector topic.

This way you don't have to go back and rehash all the history. Bob may remember different however...and I suspect he will comment.

 

[Henry HPSD]

Guys, with all due respect....

Could you remain to the Sigma 4 Im allread having difficutlies catching up two pages of info after a couple hrs sleep.

Im going to miss questions, and then ill be reading...he never answered....hihi.

Kind regards,

H>

 

[Marconi]

I've played with the optimizer in 4NEC2 in the past, but I'm not a fan of it. I just like to control things by hand is all. It takes longer, but I can observe the changes as they happen and make adjustments on my own terms. That being said, I don't remember ever saying implying, or thinking for that matter, that that feature was broken...

I don't recall saying the 4Nec2 optimizer was broken either. I'm pretty sure it works, and if a model is pretty close to working already it might even click off a better model. It is like the other tools available...it just takes practice and understanding...that's all.

I don't think the optimizer will fix a really bad model though, and I don't believe that is its purpose either.

Aside from what I mentioned above, I did notice the potential acute angle issue with this model (Henry actually mentioned it in his report). While I don't see this as a big issue as experience shows there to be little difference in what NEC2 results, I will likely address this before doing any serious work on the model. I also intend to add in a height variable to make adjusting the height easier for when I play with the model over an earth... Should I get to this point I'll happily release an updated text version of the file..

DB, I found using this hub design at the base of the S4/NV4K to be almost essential if one wants to make a model with correct and useful dimensions, especially wire diameters and using modestly conservative segment lengths for your models. I also think you might find, that within reason, making rather small physical changes in this area can produce some of the most striking benefits in the models performance.

Im not sure, but I think this hub feature used in conjunction with connections with sharp angles to elements may be the only practical way to model such a designs and minimize segment issues. I have said this before, but I'm not sure how either Homer or Booty Monster will recall...but I seem to remember this area of their homemade Vectors was an area for consideration in building their physical models. So, maybe this limitation is also practical in real world applications too. I see every one using the idea, regardless.

Almost ever model I have to date has been modified and redone at some point...and as I learned more. In the process...virtually every one has shown some reduction in the gain while the match tends to get better...other performance factors not so much, and this is why I might question this idea of our being able to steer or control the TOA as some would have us believe. I'm not even sure we can just slant the antenna and make this angel change dramatically.

I also think better models, in general, show a bit less antenna performance as they are improved. Roy, among other things in his Eznec Manual, gives us guidelines with that in idea in mind.

 

[bob85]

You are right Henry we should stay on topic, sometimes i get lost when we go off at a tangent to the topic ,

going to sleep thinking about antennas sometimes causes me to dream about antenas, i must learn to count sheep or something before i sleep hihi,

i have something on my mind that won't go away about GHZ24's cone model,

Imagine a parallel transmission-line in free space or at least far from any conductor that would upset the balance,
put a dummyload at the far end, and excite it with a transmitter, all currents are balanced and opposite phase so no radiation,

Imagine you taper the transmission-line leaving the dummyload at the far end,
will it not have some small amount of radiation due to the lines not been parallel as explained in the pulling on a rope scenario ?

Lets go back to the parallel lines and remove the dummyload then add a load to one wire of the line,
that load could be the high end impedance of a 1/2wave radiator,

Now we have a situation where one wire of the transmission-line sees the end impedance of a 1/2wave radiator while the other wire sees the very high impedance of air,

The resulting unbalance of impedance causes some radiation from the wire connected to air as nature does a balancing act via common mode on that wire which gives us the lob sided pattern of the j-pole,

Is it not the case that because GHZ 24's cone model is terminated with dummyloads we have zero unbalance,
and any radiation will be due only to the taper in the cone ?,

Does extending the radiator not cause even more unbalance due to the load becoming closer to the end impedance of a 5/8wave?

What am i missing?

thanks.

 

[Marconi]

I simply adjusted the model itself to see if it conforms to both of these statements, and it does. In that regard, it is operating as you say it should. As a matter of fact, making the listed modifications both increase maximum gain, and lower the angle of radiation. As I said above it is possible to use either method to push the radiation angle below the "horizon" in freespace.

DB, I as well have not done the adjustments to the radial angle, radial length, and/or hoop size as yet. I increased the radiator length by 1" inch at a time on my New Vector 4000 model in free space and to the specs that Bob posted some time back.

I did see the gain drop, the max angle increase above the horizon, and the match getting a bit worse with each iteration.

I also noticed similar effects in the model Henry posted by GHZ24, where the radials were way short at about 68" inches for 27.185 mhz, and supporting a 60" inches diameter hoop. Of course those results were even worse as one might expect. I did several more iterations with 10" inch increments and the longer the radiator the worse everything became. All the same things continued to happen until I reached a radiator length of 28.375' feet almost 2' feet longer than the stock NV4K.

I then reversed the process and made the mast shorter starting with the specified NV4K length of 317.5" overall length, bottom of radials to the top of radiator. I also included the top hat basket in the model. I made a 5" and a 10" inch shorter change to the radiator until I saw the FS-TOA drop to 0* degrees...where it stopped dropping.

I continued to reduce the length by one additional 20" inch change, with little effect on the gain or angle compared to the stock Vector as noted below.

The range of changes in this process from the stock to the shortest length:

317.5" Vector at 2.49 dbi @ 3* stock Vector
312.5" Vector at 2.52 dbi @ 2*
302.5" Vector at 2.56 dbi @ 0*
272.5" Vector at 2.49 dbi @ 0*
222.5" Vector at 2.23 dbi @ 0*

At the final 40" inch iteration I reached an overall radiator length of 222.5" or 18.5' feet, with 2.23 dbi at 0* degrees and a terrible match.

I saw no difference in gain at any radiator length that exceeded the values above 2.23-2.56 dbi. And the difference in the range of angles was between 0* degrees for the shortest radiator and 16* degrees for the longest model compared 340.5" inches in overall length.

I see nothing that suggest this process is in any way steering the TOA. IMO these effects, of ultimately seeing a 0* degree pattern at the horizon, are entirely due to the antenna reaching an overall length we would see from a moderately well balanced length center fed 1/2 wave radiator and nothing more.

I sure didn't see anything suggest this process would go below 0* degrees on the horizon however.

 

[jazzsinger]

Sorry I haven't had time to read Henry's report, comment or answer messages.

I have been quite ill recently and had a bit of a fall early last week that resulted in an unscheduled collision with a glass tv stand, a floor standing speaker I decided to glasgow kiss on the way down and a collision with a clr2 that i had lying on the floor. Suffice to say they won.

I'll catch up on Henry's report, answer messages and stick my tuppence worth in soon.

Sorry for not being about as much as I'd like. Looks like I've missed out a fair bit.

73 Jazz.

 

[The DB]

DB, I found using this hub design at the base of the S4/NV4K to be almost essential if one wants to make a model with correct and useful dimensions, especially wire diameters and using modestly conservative segment lengths for your models. I also think you might find, that within reason, making rather small physical changes in this area can produce some of the most striking benefits in the models performance.

Not this again. Do you really think I use umpteen thousand segments in every model I create? My standard segments per wavelength is in the range of 100 per wavelength, and that is simply to make it easier to maintain a given segment length in the models I make. In the models above it is 50 segments per wavelength, which is even less. You have also completely missed the how and why I use the element diameter that I use, and have ignored me on multiple occasions when I said that they are for theoretical and initial purposes only. As I work with models those thin diameters get thicker, to the point of a real world antenna. I have told you these things before...

Im not sure, but I think this hub feature used in conjunction with connections with sharp angles to elements may be the only practical way to model such a designs and minimize segment issues. I have said this before, but I'm not sure how either Homer or Booty Monster will recall...but I seem to remember this area of their homemade Vectors was an area for consideration in building their physical models. So, maybe this limitation is also practical in real world applications too. I see every one using the idea, regardless.

I remember Homer discussing it and saying he though it made a difference based on the versions of the Vector he made, and Donald saying otherwise. I have found on multiple occasions that the versions of models that include a hub have a slight increase in gain as well.

Almost ever model I have to date has been modified and redone at some point...and as I learned more. In the process...virtually every one has shown some reduction in the gain while the match tends to get better...other performance factors not so much, and this is why I might question this idea of our being able to steer or control the TOA as some would have us believe. I'm not even sure we can just slant the antenna and make this angel change dramatically.

I also think better models, in general, show a bit less antenna performance as they are improved. Roy, among other things in his Eznec Manual, gives us guidelines with that in idea in mind.

I was steering the TOA in a freespace environment. I have not yet tried to do the same with this model over an earth. I think attempts to steer the TOA over an earth will have far less of an effect if I can even do it at all. That is why one of the modifications I did to the model was a height adjustment, so I can easily set the height over an earth. When it comes to models, performance isn't all I look at. there is plenty of other good info available forma model.


The DB

 

[The DB]

I see nothing that suggest this process is in any way steering the TOA. IMO these effects, of ultimately seeing a 0* degree pattern at the horizon, are entirely due to the antenna reaching an overall length we would see from a moderately well balanced length center fed 1/2 wave radiator and nothing more.

I sure didn't see anything suggest this process would go below 0* degrees on the horizon however.

In freespace with most antennas you have a maximum gain at "horizontal", or straight out from the antenna. With this antenna that isn't the case. Making changes to this antenna in freespace will change the angle of maximum radiation. You don't see this as steering the pattern, yet it is possible to, at least in freespace, aim the pattern at a certain angle. To me that aiming is steering the pattern. If you choose not to call it that, that is fine. All of this may or may not be moot over an earth for this model anyway...

I was able to manipulate this model in two completely different ways to get a "below horizontal angle" in freespace, so it is definitely possible to do. When I get a chace to get an earth under this antenna I will see if, and potentially how much, this model will allow manipulating the angle of radiation...

Now when it comes to exploring this aspect of the antenna, I will have to vary the height of the antenna as I hypothesise that an antenna that is mounted a wavelength or closer to the earth will be much more difficult to manipulate the angle of radiation than an antenna that is multiple (in most cases likely 10 plus) wavelengths high that Donald is referring to...


The DB

 

[HomerBB]

DB, I found using this hub design at the base of the S4/NV4K to be almost essential if one wants to make a model with correct and useful dimensions, especially wire diameters and using modestly conservative segment lengths for your models. I also think you might find, that within reason, making rather small physical changes in this area can produce some of the most striking benefits in the models performance.

Im not sure, but I think this hub feature used in conjunction with connections with sharp angles to elements may be the only practical way to model such a designs and minimize segment issues. I have said this before, but I'm not sure how either Homer or Booty Monster will recall...but I seem to remember this area of their homemade Vectors was an area for consideration in building their physical models. So, maybe this limitation is also practical in real world applications too. I see every one using the idea, regardless.

Marconi, I clearly remember our discussion regarding using the hub design at the base of the vector models. This occurred during the time i was building my second V4k from the parts of the S4 I had made (I have built two V4k and one S4). I had by this time become convinced that extreme accuracy in the dimensions of the antenna was critical to optimized performance - first because of the difficulty I had encountered with getting the match correct on the first antenna by guessing the dimensions of the gamma match, and secondly, because much discussion had centered around tweaking this antennas performance by inches of change in the length of the 1/4 wave cone and the center radiator. We (Marconi and I/Booty) had decided the best approach to building the antenna was to start as closely to the manufacturers dimensions as I could to establish a base line and then attempt to make the changes from there to see if I could perceive improvements by doing so (I had also gotten an analyzer by this time). Marconi had been wrestling with his models of this antenna so we discussed whether the model was as sensitive to the design particulars as the real antenna seemed to be. Marconi added the hub (like the real antenna has) and the models began to accept the slanted radial configuration like a homely girl does a bouquet of flowers.