Don't be so bitter Marconi. I'm not going to bother explaining how 2dbd is two decibels over a dipole regardless of what medium you place the antenna in anymore. Just like Sirio wouldn't play your games anymore. Don't pretend I'm dishonest to cover your lie that "I said Sirio won't talk to you because you ask them questions they can't answer". You already had your answer but still don't understand there is no difference yet.
-
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.
Any Astro Plane Fans ?
- Thread starter tuner
- Start date
Eddie,
Im not asking you to imagine anything,
when you consider the idea is to modify a vector by inverting the cone like the top-one, the drawing becomes self explanatory imho,
2 x 1/2waves with a 1/2wave delay between them, DB had no problem understanding the drawing,
the bottom pic is a stretched coil to put more distance between the 1/2wave radiators for a little more gain.
Im not asking you to imagine anything,
when you consider the idea is to modify a vector by inverting the cone like the top-one, the drawing becomes self explanatory imho,
2 x 1/2waves with a 1/2wave delay between them, DB had no problem understanding the drawing,
the bottom pic is a stretched coil to put more distance between the 1/2wave radiators for a little more gain.
Well Bob, that makes sense now. the bottom model has a phasing coil. I thought it was just space between them.
See, like I told you earlier without details about your idea...I could have guessed wrong and my crystal ball is broken.
Why don't you ask The DB to model it, and if I can find my model...then you will have two models in two different software programs to compare.
Last edited:
Well this is what I did back when I was discussing with somebody about the Big Hair collinear antenna.
They are not matched well, but I just guessed at the coil configurations. They do show a little low angle to the patterns and some gain.
Is it just an A/P with a full 1/4 wave radiator with a 1/2 wave and a phasing coil above it?
I have a handle on a 180* degree phasing stub that might work, and I figure it should tune the antenna which my coil models just don't accomplish at this point. I never did try and tune them however.
I made three antennas based on bob's picture. I made a standard Starduster with 15 degree radials, a model of 180 degree phasing stub pictured and a model of the other using a RL (resistive/inductive) load.
The stock Starduster model shows a gain of 4.66 dBi.
The collinear with the RL load shows 6.46 dBi in gain, or 1.8 dB in additional gain. Not knowing how much resistance to expect the coil to have I put one ohm of resistance for that part of the load. I could redo it with no resistance in the coil, but I think that would be even less realistic.
The phasing stub model shows 6.67 dBi in gain, which is an additional 2.01 dB in gain over the stock Starduster design.
In both collinear cases, both the Starduster portion of the antenna, and the half wavelength collinear section above it are in phase.
I also tried to get the same effect using a capacity hat between the upper and lower sections, however, I could not get the phasing between the sections to line up properly with four 1/4 wavelength or smaller elements...
One thing to note is impedance. The standard Starduster model has an R of around 60 ohms near resonance, however both collinear models have an R of between 170 and 180 ohms at resonance. Some form of matching will be required for these designes.
The DB
The stock Starduster model shows a gain of 4.66 dBi.
The collinear with the RL load shows 6.46 dBi in gain, or 1.8 dB in additional gain. Not knowing how much resistance to expect the coil to have I put one ohm of resistance for that part of the load. I could redo it with no resistance in the coil, but I think that would be even less realistic.
The phasing stub model shows 6.67 dBi in gain, which is an additional 2.01 dB in gain over the stock Starduster design.
In both collinear cases, both the Starduster portion of the antenna, and the half wavelength collinear section above it are in phase.
I also tried to get the same effect using a capacity hat between the upper and lower sections, however, I could not get the phasing between the sections to line up properly with four 1/4 wavelength or smaller elements...
One thing to note is impedance. The standard Starduster model has an R of around 60 ohms near resonance, however both collinear models have an R of between 170 and 180 ohms at resonance. Some form of matching will be required for these designes.
The DB
DB, I made my Starduster at 18' feet over Average soil and over Real Earth as you can see below. I used 18' feet because Bob said your model was at 1/2 wavelength high.
You indicate your stock SD'r model shows a gain of 4.66 dbi, but I can only get my model to show 1.53 dbi over real Earth at 18' feet, and IMO that would be pretty close to what I might expect.
So, I figure, even if I modeled my SD'r with a phasing stub and a resonant 1/2 wave dipole radiator above it a few inches, and I can could get the full theoretical 3 db gain, which I won't, I would not even come close to what you note as the gain for your stock SD'r at 4.66 dbi, much less any gain near what a typical phased collinear setup should show us.
I have never noticed this much difference between several models you and I have posted here. Did I get your model's height wrong?
If your model was set at a wavelength or two, then I might understand some difference as the maximum gain typically shows more as we increase the height of a model.
BTW, like I suggested to Bob...I was working on a collinear model using my A/P model with a full 1/4 wave radiator in the top. I haven't fiddled around much with tuning, but I can not get this model to act even close to what I thought a 1/2 wave over a 1/2 wave should look like and preform like I expected.
The dipole part on top looks fine, but the A/P does not seem to contribute like I would expect a center fed 1/2 wave should work. I may be changing my mind about the A/P being nothing more than a 1/2 wave radiator.
This also suggest to me that maybe Donald was right about his 4 wire setup added to his Vector model. I have never noticed what I see here, but I could have something simple working here I don't know about too.
Will you talk to me about Bob's idea DB?
Last edited:
I'm happy to discuss what I did.
The top of the radials, and the base of the 1/4 wavelength vertical element, and the top of the mast are all one wavelength in height with my models. The radials are near 15 degrees off vertical and 1/4 wavelength long give or take a small amount for the angle and length. The vertical goes up 1/4 wavelength above the radials. The mast goes straight down from the feedpoint 1/2 wavelength, and is not connected to anything, such as an earth.
Double checking I just noticed that it was over moderate ground, which is different than the average ground that I use.
If I change to average ground and lower the model to 1/2 wavelength to the top of the mast and radials, and the base of the vertical element, I get 1.9 dBi in gain (the mast is now connected to the earth below, which I don't think makes that much of a difference as in either case there are almost no currents on my mast).
One thing that I noticed is that as I lowered the antenna the efficiency dropped by about 18%, from 61.73 to 43.85%. That and a different radiation pattern I think is the biggest difference between the heights
After lowering the antenna pattern's mine looks much like the one in your .pdf, although, the lower lobe in mine is dominant, which is different than the higher lobe in yours.
I did use my standard very thin wires, actually .1 millimeter diameter, and I use 100 segments per wavelength (or 50 for 1/2 wavelength, and 25 per 1/4 wavelength, ect.). I have not changed the diameter in this case because I was simply looking to duplicate a phase/current pattern for Bob. When it comes to segments per wavelength, I am careful to keep to this segment standard with all my models as much as possible.
It was also made as close to 10 meters in frequency as is customary for me instead of CB frequencies.
I'm surprised Bob didn't post the current'phase results and the gain plot results.
Yes, the antenna does act very similar to a center fed dipole, and vertical center fed dipoles don't like to be close to an earth. I had typed out an observation I made with center fed vertical dipoles and how their gain drops off near an earth and suggested getting the antenna as a whole up a certain amount of height. Looking back I am not seeing that now so I guess I removed that part of my post. Somewhere I have a chart that shows the gain of a center fed dipole with its base (not the feedpoint) starting at 0.1 wavelengths above earth, and graphs it to 10 wavelengths above average ground. I'll dig it up later if you are interested.
The DB
The top of the radials, and the base of the 1/4 wavelength vertical element, and the top of the mast are all one wavelength in height with my models. The radials are near 15 degrees off vertical and 1/4 wavelength long give or take a small amount for the angle and length. The vertical goes up 1/4 wavelength above the radials. The mast goes straight down from the feedpoint 1/2 wavelength, and is not connected to anything, such as an earth.
Double checking I just noticed that it was over moderate ground, which is different than the average ground that I use.
If I change to average ground and lower the model to 1/2 wavelength to the top of the mast and radials, and the base of the vertical element, I get 1.9 dBi in gain (the mast is now connected to the earth below, which I don't think makes that much of a difference as in either case there are almost no currents on my mast).
One thing that I noticed is that as I lowered the antenna the efficiency dropped by about 18%, from 61.73 to 43.85%. That and a different radiation pattern I think is the biggest difference between the heights
After lowering the antenna pattern's mine looks much like the one in your .pdf, although, the lower lobe in mine is dominant, which is different than the higher lobe in yours.
I did use my standard very thin wires, actually .1 millimeter diameter, and I use 100 segments per wavelength (or 50 for 1/2 wavelength, and 25 per 1/4 wavelength, ect.). I have not changed the diameter in this case because I was simply looking to duplicate a phase/current pattern for Bob. When it comes to segments per wavelength, I am careful to keep to this segment standard with all my models as much as possible.
It was also made as close to 10 meters in frequency as is customary for me instead of CB frequencies.
I'm surprised Bob didn't post the current'phase results and the gain plot results.
Yes, the antenna does act very similar to a center fed dipole, and vertical center fed dipoles don't like to be close to an earth. I had typed out an observation I made with center fed vertical dipoles and how their gain drops off near an earth and suggested getting the antenna as a whole up a certain amount of height. Looking back I am not seeing that now so I guess I removed that part of my post. Somewhere I have a chart that shows the gain of a center fed dipole with its base (not the feedpoint) starting at 0.1 wavelengths above earth, and graphs it to 10 wavelengths above average ground. I'll dig it up later if you are interested.
The DB
Bob has recently pointed out why the 4 wire 90 degree test worked in the Vector collinear. The original antenna did not have a 1/2 wave upper section. It's closer to 5/8 wave. The top collinear section also peaked its gain just over 1/2 wavelength. The net result is we have stacked a 5/8 wave over a 5/8 wave over a 1/4 wave. Now it makes perfect sense why the 90 degree delay worked well here.
I can now conclude the 4 wire test is unable to determine if the stock antenna is collinear or not. This means the phase delay needed to add more collinear sections is only dependent on the length of the monopole as we would expect. The phase shift that makes it possible to be a full 3/4 (to 7/8) wave radiator takes place in the cone and this does not alter the delay needed to stack more sections.
I can now conclude the 4 wire test is unable to determine if the stock antenna is collinear or not. This means the phase delay needed to add more collinear sections is only dependent on the length of the monopole as we would expect. The phase shift that makes it possible to be a full 3/4 (to 7/8) wave radiator takes place in the cone and this does not alter the delay needed to stack more sections.
Thanks.
I started with my model at 36' feet. The match at that height was near perfect, and the gain was a tad better than a 1/2 wave CF dipole and I tried to duplicate all the measurements for my real antenna. In this model I did not use the SD'r hub, and that does make some difference in results. I lowered it down to 18' feet to the hub...and the match, pattern, gain and angle changed as one would expect. My model was also on a 18' foot mast that was grounded, but I get no where near the 4.66 dbi gain and I won't at 36' feet either. so I figure Bob should just go with your collinear model.
I just use average ground in all my models too, unless I'm checking the effects of changes in the soil.
My model shows the maximum TOA with 1.89 dbi gain at 50* degrees and the lowest angle at 14* degrees and 1.53 dbi over real Earth. In Free Space without the mast in the model, it shows 2.27 dbi at 0* degrees and I think that shows me the model is about what I should expect a little better that a CF 1/2 wave dipole. The Average Gain report also show 1.043 = 0.18 db. If I had tuned the model to resonance at this height...the A.G. report would likely improve, another good indicator for the model. This tells me that our starting models are very close.
Eznec does not have and efficiency report perse, but the Average Gain is about the same, and I saw a nice reduction in the quality of the match on lowering the antenna too.
This may well be the difference in the soil conditions for our models. I changed my model's soil to dry/sandy (a worse soil condition) and I saw the same results you do with 1.93 dbi maximum gain at 14* degrees with the maximum lobe being lower too.
Due to my 500 segment limitation with my version of Eznec...I try and set my models around 2" - 3" per segment where you set yours at about 4" per segment if I figured yours right. We are close if I'm correct and I too try to keep a standard. The only difference is I try and test the Geometry/Segment checker to the limit and I always try and use the reasonable wire diameters for my models. IMO, if you check close you may find that very thin wires and very high segment counts can generate higher gains and ill-effect models results in other ways possibly. I have not checked if this use affect on the Average Gain report, but I wouldn't doubt such procedures may well also minimize the Segment/Geometery warning notices somewhat.
I would like to see that if I could. I learned something trying to set my Starduster model as a collinear stacked 1/2 wave. I haven't tried to set this model with all wire connections according to the NEC RULE, as I refer to it, with: "wire #1 end 1" connected to "wire #2 end #2" only, and visa versa of course.
Seve, I hope I can explain this so it is easily understood, but I noticed in the stacked model when I scanned it...the model's pattern for the SD'r part at the bottom of the stacked and phased setup does not look anything like the SD'r model by itself.
The currents do not flow on the SD'r radials making a 1/2 wave pattern for the model...they look to be cancelled out and only the top 1/4 wave radiator is radiating into the far field with the top 1/2 wave. This may all have to do with the NEC Rule that I describe above, and I've never seen this effect in any collinear I ever done with phased
stub attached between the two radiators.
In light of what I said above...I am puzzled at your observation that the pattern for the SD'r looked fully like a 1/2 wave and I see the opposite.
I would like to see the report you made if possible. I have done something similar using Eznec using an A99 model, but in that case I could not track the match with such a badly matched antenna...so I only watched pattern, gain, and angle.
Thanks Steve.
i can see a few issues with the astropane co-linear eddie,
the source is in the upper monopole,
you left the hat loaded 1/8wave there instead of removing the hat and extending to 1/4wave before adding the 1/2wave delay and upper 1/2wave,
the mast is a bad length for cmc,
i would isolate it 1/4wave below the hoop and take it from there.
the source is in the upper monopole,
you left the hat loaded 1/8wave there instead of removing the hat and extending to 1/4wave before adding the 1/2wave delay and upper 1/2wave,
the mast is a bad length for cmc,
i would isolate it 1/4wave below the hoop and take it from there.
I'll tell you Bob, this A/P model will allow for the feed point to be set at both places, and the Average Gain will show good results, and the model that has the source on the radial leg under the feed point on the real antenna does not produce good results in the collinear configuration. This is why I was saying earlier...that this antenna may not be a 1/2 wave after all.
I talked to somebody about this the other day, and I know we decided that the source should be where we see it on the real antenna, and it was said at the time...that if it was not located there...the antenna was not an A/P...just a plan and simple 1/2 wave dipole instead. Maybe I said it...I don't really remember anymore. I do remember that I quoted something in the Patent about being able to use different designs for feeding and it was evidence of something to do with circuits being similar will produce similar results. I even posted the words from the Patent. It may have nothing to do with the issue about where the source goes on this antenna however, it was just my thinking about "why."
Well, I tried the collinear idea with the match under the feed point and the model is no-way-no. So, I don't know what to do.
The top hat issue is about as strange. At first I was using a full 1/4 wave radiator in the top of the A/P for the model and it would not work worth a hoot. Early this am I fixed the model using just three legs of the top hat and setting the phasing stub where the other radial went, and that worked the first time...just like you see it now...that is if you are looking at that model. I have done nothing more to the model today, because I knew it was not showing the gain that DB's model shows so I figured I would never use it.
Since then however, I have talked to Steve and found out why my model of his standard Starduster was not work with anywhere near the gain. So I just stopped with the models.
Bob I think I set both of the collinear models at 18' feet, like you told me that DB did. I see I did not change the title to read 18' feet and it still shows 36' feet, but that is wrong.
Also I did not isolate these models either. You should be able to see a blue dot on the model where I was isolating your model however...so that will be one touch change and it is isolated. I just didn't get that far, I was just trying to get the models to work like I read DB's model was working with the SD'r...and that never did work for me. So, after I found out DB was doing a SD'r, I dropped the idea about the A/P.
Thanks for the tips...I hope I was able to explain well enough. If you see anything else...just let me know before somebody gets mad at me and stops talking.
BTW, I sent you an email explaing some of this I think, so look in your other computer you just got fixed.