160 Meter Transmission Line Study - SWR and Losses

[WA1GFZ]

yes Tom, Thank you fo running the simulation.

[WA1GFZ]

Tom two wires side by side is about 80 ohms. I would think 4 wires would be 1/4 that.
(4 surfaces vs 2) Just run the formula for parallel transmission lines and make the spacing about 10 mils.
When working on my log ant I found this is why Crapadyne went to the square boom to get a lower Z.
I used to have an article on BB transformers and multiple wires to get lower Z but can't remember wher I read it.

[AB2EZ]

Tom
et. al.

I have a copy of EZNEC that I purchased... and have used for fun... but I have a question that I don't know the answer to. Maybe the answer is somewhere in the EZNEC documentation:

When one uses EZNEC to calculate the radiation pattern... and when EZNEC provides the "peak gain relative to isotropic", is it really taking into account estimated ground losses... or is is simply comparing the peak of the radiation pattern divided by the average of the radiation pattern over 4pi steradians?

Stu

[WA1GFZ]

John,
Once I ran a 2 wire inverted Vee and cut one wire 67 feet long and the other 99 feet long so it would load on 40 easier. The flash box didn't like the 75 meter 1/2 wave on 40 meters. Maybe you could hang a little vertical on one wire to tam the ant on 40. gfz

[Steve - WB3HUZ]

Stu, does EZNEC calculate Sommerfield ground coefficients. If not, then the gain numbers are probably a little high, especially when the antenna is close to the ground wrt wavelength. Even with Sommerfield ground, I've heard some claims the number can be suspect. Not sure as to exactly why.

Cebik dicusses at the link below.

http://www.cebik.com/amod/amod60.html

Tom
et. al.

I have a copy of EZNEC that I purchased... and have used for fun... but I have a question that I don't know the answer to. Maybe the answer is somewhere in the EZNEC documentation:

When one uses EZNEC to calculate the radiation pattern... and when EZNEC provides the "peak gain relative to isotropic", is it really taking into account estimated ground losses... or is is simply comparing the peak of the radiation pattern divided by the average of the radiation pattern over 4pi steradians?

Stu

[AB2EZ]

Steve

Thanks for the pointer to the reference document: http://www.cebik.com/amod/amod60.html

I ran the EZNEC program that I have (EZNEC 4.0) using two different ground models: "Real: MININEC" and "Real: High Accuracy". I used the example of a 121 foot long dipole at a height of 10 feet, at a frequency of 3.9 MHz (i.e., the same as the first example in the reference document). I obtained the same results as the reference document described. I.e., the "Real: High Accuracy" ground model showed a peak gain (straight up) of ~ -1 dBi, but the "Real: MININEC" ground model showed a peak gain (straight up) of over +11 dBi. Thus, just as the reference document describes,  the "Real: MININEC" ground model does not accurately portray the ground losses.

I then ran the EZNEC model for a 121 foot dipole, at 40 foot height, at a frequency of 1.885 MHz. With the "Real: MININEC" ground model, the peak gain (straight up) was 9.2 dBi. When I ran the program with the "Real: High Accuracy" ground model, the peak gain was 3.8 dBi. Finally, I ran the program using the "perfect ground" model, and the peak gain (straight up) was 8.7 dBi. Thus the high accuracy ground model predicts that the ground losses for this antenna would be approximately 5dB. In other words... even with a perfect transmission line and a perfect tuner... this antenna would lose approximately 5dB of power because of ground losses. The impedance was 6.973 - j1263 ohms with the "Real: High Accuracy" ground model and it was 1.976- j1265 ohms with the "Real: MININEC" ground model.

I also ran the program for a 242 foot dipole at a height of 40 feet, and a frequency of 1.885 MHz. The feedpoint impedance using the "Real: High Accuracy" ground model was 32.81 - j90.8 ohms.
The peak gain (straight up) was 4.7 dBi. With the "perfect ground" model, the peak gain (straight up) was 8.9 dBi. Thus a full length 160 meter dipole at 40 feet height loses about 4 dB to ground losses.

I also ran the program for a 242 foot dipole at a height of 120 feet, and a frequency of 1.885 MHz. With the "Real: High Accuracy" ground model, the peak gain (straight up) was 6.67 dBi. With the perfect ground model, the peak gain (straight up) was 7.73 dBi.

[K1JJ]

Basically, the modeling falls apart below 1/4 wavelength above ground. This has been something to avoid since at least 1987 when I first started using miniNec.

Back in '88 I was suckered into thinking a dipole for 75M at 10' would be fantastic - with huge high angle gain. I believe MiniNec used a perfect metal mirror for the ground modeling.

Anyway, I put a 75M dipole up at 10' and found it was deaf. Terrible ant even for local work.

Years later I did many real-world comparisons on 75M dipole heights and found even a dipole at 30' was not as good as at 60' for high angle local work. Ground losses become the major loss in addition to takeoff angle. Basically, high angles between 80-90 degrees are near useless for average 75M local work.  A take-off of about 50-60 degrees is much more useful in addition to the lower ground losses. This equates to a dipole at between 50'-80' high for optimum performance. (surprise surprise... )

I played with reflectors under dipoles, phased arrays shooting straight up, stacked 75M rotary dipoles at 190' and 80' phased straight up, etc etc.  You can't beat a single flat straight dipole at 50'-70' high for combo local and moderate skip on 75M.

Much also depends on the point in the solar cycle, time of day and condix of ionosphere, of course.

I had a reference dipole out in the woods at 30' for a year and finally cut it down. It was never better than any of the higher dipoles at ANY time. The only advantage it had was being quieter for power line noise cuz it was farther away from the street. 

73,
T

[WA1GFZ]

Tom,
You have power line noise....find the bad insulator and report it. I bet the powerco will change it out because it is leakage......a 22 short also works

[K1JJ]

Frank,

The power line noise problem is something you'd never suspect -  Chuck filled me in on this...

Many streets run straight for quite a distance (like mine here) and the houses are also in a line.  It so happens that our power line terminates at the end of the street here at this house. In effect we have a longwire that has tremendous gain on 75M  pointing here.

As proof, you can tune down the 75M band and hear the peaks and nulls of thre noise as the long wire cancels and adds it's wavelengths.

On certain days the noise can be S9 on 75M, but that is not a problem since I operate at night and the band noise covers it.  It doesn't really affect me.

To confirm this, I built up a 4el portable Yagi for 135Mhz and bought an aircraft band AM scanner to hear the noise. I can go up and down the street and prove that everything is reasonably quiet. I even went to the next streets a mile away and cannot find anything bad. There IS a small residual noise on certain poles, but not enuff to complain about.

The end result is when these noises randomly add up on the longwire, they produce a strapping signal here. I know it comes from down the street cuz that's the only place there is a power line for many miles. Plus my 6M beam peaks up on noise that way.

Strange stuff, huh?

T

[WA1GFZ]

interesting...

[Steve - WB3HUZ]

Good stuff, Stu. You test runs clearly show the problem with MiniNec based modeling sw and low antennas. Another thing to keep in mind with MiniNec is that some versions do have real ground capability, but the feedpoint impedances are still calculated using perfect ground.

If possible use a Nec based (newer the better, like Nec4) modeling sw. But always read the manual thoroughly, since implementations vary. For best results, use the Sommerfield-Norton ground, since it best models ground losses. Finally, bounce your model results off the real word, previous work, written documentation, etc. For example, Tom's 10 foot high dipole with tons of gain should have been suspect. How many guys have we heard on 75 meters with a CONSISTENTLY big signal running a 10 foot high dipole?

As you pointed out in a previous post, the modeling will not show any near-field (or otherwise) losses due to coupling to near to the ground/antenna structure like power lines/wiring, phone lines, metal fences, etc. So, many low antennas may take an even worse hit than the numbers posted. This point really stuck in my head many years ago when I was running an end-fed antenna and was having terrible phone and TVI. I realized that whatever signal was getting into the phone was just that much less signal that could make it to the ionosphere.

[Tom WA3KLR]

Steve and all,

Stu is running EZNEC 4 and I am running EZNEC 3 (which is 5 years old now!) for my antenna simulations posted above.  These programs with “Real/High Accuracy Ground” selected are handling all the pitfalls you and the Cebik article mentioned.  A dipole’s feed point impedance drops from the 53 Ohm range down to 30 Ohms with a perfect ground selected, so the feed point impedance with Real ground is not calculated with a perfect ground.  I think ELNEC was this way though and also had the unrealistic high gain at low heights also.

I recommend that people test their antenna programs per the situations above and get EZNEC 4 if their old antenna programs aren’t handling the low dipole situations well.  Retire the use of the inferior programs.

[WA1GFZ]

So some fine person want to send Tom a boot copy of 4 so he can provide better information since he is doing such a fine job simulating our dream antenna systems....

Ho Ho Ho

[Ian VK3KRI]

....
The simplest example of multi-wire open wire line is 4-conductor, laid out in a square pattern, with diagonal corners strapped together.  I have seen diagrams of line with as many as 16 conductors, along with formulae for calculating the impedance.  This allows for (relatively) low impedance line without ridiculously close spacing and/or large diameter conductors.

Does anyone know of a readily available source of this data, particularly on line?

"Radio Antenna Engineering" by LaPorte lists 21 types of lines . A pdf is available at http://snulbug.mtview.ca.us/books/RadioAntennaEngineering/

[AB2EZ]

Ian

Thanks... this is a great reference. I downloaded it, and saved it on my computer. I skimmed through it, and I think it has a lot of very useful information about the real behavior of real antennas.

Thanks again!

Stu

[flintstone mop]

I'll have to agree the HUZ. Just stick what you can up there and rock on. Everyone hasn't mentioned the loss in the tuner. That can be many dB's. And the aerial has to be WAY up there, if it's a dipole OR do a vertical like Don Chester, and then you'll start radiating a real 160M signal. I'm happy with my ladder line fed 180 foot dipole at 70 feet and the K1JJ tuner. Almost had Phoenix Az on that system last Winter! Careful patient listening would have made a QSO happen.
G'day
Fred

[WA1GFZ]

KB3AHE is making the trip up here with 60 feet of antenna. He is about 10 to 20 dB weaker than Mikey tonight.

[Steve - WB3HUZ]

Wow. So Frank got his ant working on 160! Sorry I missed it. Give us the details Slab!

[W3SLK]

Yeah, Frank was in there for a while but the band was just deplorable. I think it was suffering from the remanants of that flare we had earlier in the week

[Steve - WB3HUZ]

Err... not a god time to get one and try out a new ant.

[WA1GFZ]

Maybe it is time for frank to lay out some radials to drop ground losses a bit.
I would swag that if you can't get longer go lower resistance.

[Tom WA3KLR]

The study continues - I did a quick simulation study of Frank KB3AHE’s 60 foot long folded dipole at 1.9 MHz. as mentioned in his topic “A 60' flat top for 160 and 75m!”
at http://amfone.net/Amforum/index.php?topic=9181.0

I assumed 50 feet height.

Feed point impedance at 1.9 MHz. = 2.1 – j443 Ohms.
Feed point SWR with 600 Ohms line = 441:1
Assume 90 feet of 600 Ohm feed line, transmission line loss = 1.53 dB.
With 1500 Watts applied, at tx end of line 17 kV r.m.s., 10 kv r.m.s. at ant. feed point.
With 1500 Watts applied, at tx end of line 5.5 Amps r.m.s., 23 Amps r.m.s. at ant. feed point.

The antenna gain appears to be about 19 dB below a full-length half-wave dipole at 50 feet.

The impedance seen at the transmitter end of 90 feet of 600 Ohm ladder line is 53 – j3100 Ohms.  A simple matching network of just one series inductor of 260 uH (or two 130 uH inductors, no mutual coupling) would provide a 53 Ohm match to the transmitter.  Inductor Q needs to be 600 or higher for low loss.

[WA1GFZ]

Tom,
Sounds like a job for  lots of 1/2 inch copper tubing

[WA1GFZ]

Once I got loaded on New Years tonight watching football with the kids and thinking way too hard.

Based on Tom's post today I quickly see big -J component means big inductors.
A Fan Dipole reduces this -J component quickly (like JN's on page 2 here)
I would think you can't avoid a lot of loss if forced to use large inductors.
Low R load does not seem as big a deal.

I wonder if Frank AHE would be better off with a multiple wire 60 foot dipole to reduce -j component making it easier to tune???

Happy New Year!!!!!!!!!!!!!