Affect of Nearby Building on Part 15 AM Antennas

[druidhillsradio]

"These models have no other obstructions in their useful propagation paths, e.g., they are in the center of an open area test site on the surface of the earth."

Exactly. So the use of modeling software (which enjoy playing with BTW) has no meaning in applications utilizing flea-power radio transmitters that are not operating in unobstructed locations. I would cheerfully request, that you put down your mouse, turn off your computer and operate a Part 15 AM transmitter. move it around a bit, and see if the "physics" of grass, trees, houses, etc., and see if you still believe the stuff you have been posting is accurate.

The very reason that for years the FCC advocated the use of Open Area Test Sites for compliance measurements is because of the effect of objects within the proximity of the antenna.

 

[rfry]

I would cheerfully request, that you put down your mouse, turn off your computer and operate a Part 15 AM transmitter. move it around a bit, and see if the "physics" of grass, trees, houses, etc., and see if you still believe the stuff you have been posting is accurate.

When my posts are read with understanding, it will be plain that the NEC study is not one showing received field intensity. It is not intended to be, and such may not logically be assumed.

It is a study showing the antenna system gain in dBi for two configurations of a 3-m whip and a building, to show the difference in the system gains of those two configurations.

I referred to this point in Reply #5 in this thread, which I repeat: "In summary, the NEC study shows that buildings and other conductors have nearly the same affect on radiation whether the base of the 3-m radiator is near the earth, or elevated -- not that the groundwave coverage of Part 15 AM stations is almost uniform in all directions."

Again, the purpose of the study was not to show the "coverage" of either system. That analysis would have to include the propagation environment.

RF

 

[PhilB]

Richard,

Could you please post the input data for this model. I need the x,y,z coordinates of the wires and the number of segments for each wire. I would like to check your results. I think a second opinion on your results is in order in view of the contradictory responses that people have mentioned.

 

[rfry]

Could you please post the input data for this model. ... I think a second opinion on your results is in order in view of the contradictory responses that people have mentioned.

The contradictory responses arose from an incorrect understanding of what the NEC study showed -- information about which was included in the paper I linked to. I have also repeated this information several times later in this thread.

The point not recognized, perhaps, is that a wavelength in the AM broadcast band is much longer than the dimensions of the "building" in the NEC model, so the building has much less affect on the gain of a 3-m whip than some people expect.

For an extreme analogy, if a bright light bulb replaced the earth-mounted 3-m radiator, it could not be seen on/near the ground on the opposite side of a solid building. But if that whip was replaced by someone loudly beating a drum, that sound could be heard on the opposite side of that building.

Anyone experienced in NEC modeling and antenna engineering can construct the NEC models in my paper using the descriptions provided there. In general:

Any model of the building with the outside dimensions shown in the paper, and the number of evenly-spaced vertical and horizontal wires shown will be acceptable. The spacing of the wires is much less than 1/10 of a wavelength at the applied frequency, and those wires will behave as a single, solid structure to that EM radiation.

The elevated 3-m whip was modeled with no "ground" or other conductors attached. A series resistance was added at the feedpoint (at the base) to produce the same antenna system gain as when the base of the whip was at earth level. This insures that whatever effect is seen in the results is produced by the structure, and not by differences in the radiation of the driven conductors, themselves.

The gains shown for the two configurations may differ from mine in another NEC model, depending on the resistances used in the connections to the perfect ground plane. But as long as the gain of the elevated whip is set to the gain of the lower whip, and the vertical wires of the "building" connect to ground through the same resistance in the two configurations of the location of the whip, then the difference in the circularity of the nominal gains of the two configurations will show the affect of the structure on the two whips -- which was reason for the study.

Below is the NEC input file of the model with the 3-m whip at earth level. Any modification to it will need to be done with due care if the results are to be valid.

NEC-Win Plus INPUT FILE
------------------------
C:\Program Files\NEC-Win Plus\RJF Antenna Files\Part 15 AM with Nearby Structure.NWP

CM
CE
GW 1 25 0 0 0 0 0 9.84 0.0053543
GW 2 3 -60 -20 10 -60 20 10 0.0045
GW 3 3 -60 -20 20 -60 20 20 0.0045
GW 4 10 -60 -20 0 -60 -20 20 0.0045
GW 5 10 -60 -6.666667 0 -60 -6.666667 20 0.0045
GW 6 10 -60 6.666667 0 -60 6.666667 20 0.0045
GW 7 10 -60 20 0 -60 20 20 0.0045
GW 8 3 -20 -20 10 -20 20 10 0.0045
GW 9 3 -20 -20 20 -20 20 20 0.0045
GW 10 10 -20 -20 0 -20 -20 20 0.0045
GW 11 10 -20 -6.666667 0 -20 -6.666667 20 0.0045
GW 12 10 -20 6.666667 0 -20 6.666667 20 0.0045
GW 13 10 -20 20 0 -20 20 20 0.0045
GW 14 3 -60 -20 10 -20 -20 10 0.0045
GW 15 3 -60 -20 20 -20 -20 20 0.0045
GW 16 10 -46.666667 -20 0 -46.666667 -20 20 0.0045
GW 17 10 -33.333333 -20 0 -33.333333 -20 20 0.0045
GW 18 3 -60 20 10 -20 20 10 0.0045
GW 19 3 -60 20 20 -20 20 20 0.0045
GW 20 10 -46.666667 20 0 -46.666667 20 20 0.0045
GW 21 10 -33.333333 20 0 -33.333333 20 20 0.0045
GW 22 3 -60 -6.666667 20 -20 -6.666667 20 0.0045
GW 23 3 -60 6.666667 20 -20 6.666667 20 0.0045
GW 24 3 -46.666667 -20 20 -46.666667 20 20 0.0045
GW 25 3 -33.333333 -20 20 -33.333333 20 20 0.0045
GW 30901 1 9901.0000 9901.0000 9901.0000 9901.0001 9901.0001 9901.0001 .00001
GS 0 0 .3048
GE 1
GN 1
EX 0 30901 1 0 0.0 1.0
LD 4 1 1 1 25 0
LD 4 1 3 3 5 0
LD 4 4 1 1 200 0
LD 4 5 1 1 200 0
LD 4 6 1 1 200 0
LD 4 7 1 1 200 0
LD 4 10 1 1 200 0
LD 4 11 1 1 200 0
LD 4 12 1 1 200 0
LD 4 13 1 1 200 0
LD 4 16 1 1 200 0
LD 4 17 1 1 200 0
LD 4 20 1 1 200 0
LD 4 21 1 1 200 0
NT 30901 1 1 2 0 0 0 1 0 0
FR 0 1 0 0 1.7 1
RP 0 1 360 1000 90 0 1 1
RP 0 181 1 1000 -90 0 1 1
EN

//

 

[druidhillsradio]

"Just...these things are matter for dicsussion and consideration, not rancor.
Engineers MUST see beyond things that divide people by opinion and transmit as much truthful knowledge as possible."

A wise engineer once told me, "Engineers wonder how things work, technicians make things work."

 

[PhilB]

Richard,

I ran your code. I got similar results showing little effect of the building model on the radiation pattern.

I then modified your code to connect all vertical building wires to ground through 50,000 ohms instead of 200 ohms. I connected one leg, wire 10, to ground through 2 ohms. There was a dramatic effect on the radiation pattern. The results are shown here:

http://www.sstran.com/temp/3m_antenna_with_building_obstruction.pdf

I tried grounding other vertical building wires instead of wire 10 through 2 ohms and got various effects on the radiation pattern. I also tried grounding wire 10 through other values of resistance and got other patterns.

Grounding only one leg of the building through a low value resistance more closely simulates the effect of house wiring on the radiation pattern. House wiring is returned to ground only at one point (service entrance).

My results demonstrate that the NEC model can show detrimental effects of a building near the antenna. It's all in how you set up the model. The dimensions, complexity and resistances of building conductors are unpredictable, so a blanket statement on the effect cannot be made. It would be possible, with a huge amount of effort, to model an individual house and antenna. I suspect the results would closely match actual performance.

 

[PhilB]

Correction to link to model results posted in my last message:

http://www.sstran.com/3m_antenna_with_building_obstruction.pdf

 

[rfry]

... It's all in how you set up the model. ...

A more important statement cannot be written about the use of NEC.

As I wrote when I posted my NEC input data, "Any modification to it will need to be done with due care if the results are to be valid."

The cardioid pattern with nearly a 30 dB h-plane null as posted by PhilB is not possible for vertical wires spaced as closely together in wavelengths as in my NEC model. Wires that close together behave almost exactly like a unit radiator, that is, a single vertical wire of rather large cross-section. The h-plane radiation pattern of such a radiating source is highly circular.

The link below leads to a modified NEC model to illustrate this point. The only wires in that model are the original 3-m whip with its base at earth level, and Wire 10 that PhilB used as the primary vertical radiator in his model of the "building." Wire 10 in PhilB's model was the only wire in the building structure with any significant, real-world radiation. Note in that link below that the net radiation pattern is almost perfectly circular -- and that removing the other wires increased the antenna system gain by a fraction of a decibel..

If radiation patterns similar to that in PhilB's analysis were possible from such closely-spaced conductors, then licensed AM broadcast stations needing directional radiation patterns would not need to install driven, vertical radiators hundreds of feet apart to produce them.

http://i62.photobucket.com/albums/h85/rfry-100/PhilB_Pattern_Check.gif

RF

 

[LibertyNT]

Fry. Its Different dude, Its just different. Part 15 is different.

 

[rfry]

...Its Different dude, Its just different. Part 15 is different.

Regardless of what people want to believe, and/or what would be most useful to them -- the quote above is not supported by the laws of physics (sorry).

The only difference in the h-plane field intensity at a given distance from a licensed AM broadcast station and an unlicensed "Part 15" AM setup is related to their difference in radiated power in that direction (other things equal). Nothing else.

My statements here are supported by engineering textbooks and proven, real-world results going back many decades.

//

 

[Tom Wells]

I am going to hope most adults adults have seen a "standing wave" of light/dark bars in the ends of operating flourescent tubes.
Sometimes moving, sometimes still. These denote current/voltage nodes along the transmission medium, in this case the flour tube, but it also visually shows how such nodes occur in transmission lines.
There is often one very DARK primary ring, with those "outward" weaker and narrower.
It would be akin to the AM station wave prop model where the director or reflector is placed in the "hundreds of feet" model,
because that's the sweet spot for doing so.

Other spots will work, but results are much harder to hit. Try to land your reflector in a "closer" zone, miss the "dark" ring,
and you have to throw in lots of phase shift, and it may never do as well as you'd hoped.

That's why I was over on the engineering board discussing the "river conductivity".
There aren't answers. There are only more considerations, which translate into more terms in a model that
may provide places to enter these factors.
Does the NEC have a place to enter in the massive iron piping system and the cast iron radiators in my house tied directly to ground?
There is a string of 10 or more buildings in each direction immediately N/S in a perfect line all with such grounded piping.

That doesn't make the model wrong, it just shows how quickly things whammy up a model, and why
technicians and engineers are not really two different things, but more points along a spectrum.
I am employed as more of a technician, but oh how much I appreciate the engineering background.
I can dismiss things that I know are "impossible" where exp and physics agree, when soving a problem.
More importantly, one or the other can show flaws in assumption, practice, or measurement, etc.
This goes both ways.

 

[PhilB]

The cardioid pattern with nearly a 30 dB h-plane null as posted by PhilB is not possible for vertical wires spaced as closely together in wavelengths as in my NEC model. Wires that close together behave almost exactly like a unit radiator, that is, a single vertical wire of rather large cross-section. The h-plane radiation pattern of such a radiating source is highly circular.

OK, try this. Make wire "10" 143.4 ft high and re-position it on the X axis (for simplicity) to -41 ft. Here is the NEC wire line:
GW 10 10 -41 0 0 -41 0 143.4 0.0045

Also, for an apples-apples comparison with my previous model, set the ground load resistance for wire "10" to 2 ohms:
LD 4 10 1 1 2 0

The results clearly show the dramatic effect of wire "10" on the radiation pattern. Overall gain is -20.8dB in the direction opposite to wire "10". Gain drops to -37dB in the direction toward wire "10".

It’s clear that a nearby wire can have an effect the radiation pattern. This model is about as simple as it gets, so I sincerely doubt there is a significant difference between the NEC results and the real world of physics.

I know of one transmitter site that was in use back in the ‘30s that used one active radiator tower and two “passive” towers to create a directional pattern.

 

[rfry]

It’s clear that a nearby wire can have an effect the radiation pattern. This model is about as simple as it gets, so I sincerely doubt there is a significant difference between the NEC results and the real world of physics.

The issues are those of how near is "nearby," and the configuration of that nearby parasitic radiator with respect to the directly-driven radiator.

In PhilB's 2nd modification, he changed the height of the parasitic radiator from 0.035 wavelengths to 0.25-wavelengths, located it 41 feet away from the 3-m whip, and used a 2-ohm connection to a perfect ground plane. This configuration is radically different than the original NEC model.

That 2-ohm ground connection is typical of the set of 120 each, 1/4-wave long buried radials used by licensed AM broadcast stations, and would be an unlikely scenario for a Part 15 user. Also the 41-foot spacing between the two wires is 25.5 degrees long at 1.7 MHz, which is not particularly near in that sense.

But under those unusual conditions then yes, it is true that the net radiation pattern becomes quite directional. However the peak gain of that directional pattern is reduced by 27% from that of the 3-m whip without the parasitic radiator.

What happens when that same parasitic is located 5 feet away from the whip, and 50 segments are used in the parasitic so that the Average Gain Test of the NEC model reports that it is "highly accurate?"

The link below leads to a NEC study of that setup. Note that the net radiation from the system with the parasitic then is circular within 1.5 dB of its RMS value.

Also note that the peak gain of the system with the parasitic is about 9.7 dB lower than the 3-m whip used alone -- which results in a reduction in peak radiated field of about 66%.

Neither system with a "nearby" parasitic would be useful in extending the coverage range provided by the 3-m whip alone, as has been suggested by some.

http://i62.photobucket.com/albums/h85/rfry-100/3-m_Whip_with_qtr_wave_parasitic.gif

RF

 

[rfry]

Does the NEC have a place to enter in the massive iron piping system and the cast iron radiators in my house tied directly to ground? There is a string of 10 or more buildings in each direction immediately N/S in a perfect line all with such grounded piping.

Such an assembly with buried conductors could be modeled using NEC-4, but the price of a NEC-4 license is fairly steep: $300 minimum, depending on the user. That doesn't include the NEC software itself.

//

 

[PhilB]

R. Fry:

In my slight modification of your building wire frame model, I demonstrated that that the building can have a dramatic effect on the radiation pattern. A very large null occured in the direction from the antenna to the wire 10 corner of the building. The only changes I made were to change the resistance to ground of all legs of the building, except for wire 10, from 200 ohms to 50,000 ohms, and I changed the resistance to ground in wire 10 from 200 ohms to 2 ohms. Your response to my results was:

“The cardioid pattern with nearly a 30 dB h-plane null as posted by PhilB is not possible for vertical wires spaced as closely together in wavelengths as in my NEC model. Wires that close together behave almost exactly like a unit radiator, that is, a single vertical wire of rather large cross-section. The h-plane radiation pattern of such a radiating source is highly circular.”

And:

“If radiation patterns similar to that in PhilB's analysis were possible from such closely-spaced conductors, then licensed AM broadcast stations needing directional radiation patterns would not need to install driven, vertical radiators hundreds of feet apart to produce them.”

By the way, my modified model passed all geometry and segment checks and the Average Gain Test produced a result of 1. This result is defined in the documentation as follows: “> 0.95 and < 1.05 Model is likely to be accurate”. This is the highest level of accuracy.

After reading your responses quoted above, I created another, simpler, model with just one vertical wire 41 ft. from the antenna on the X axis. This wire is vertical and is ¼ wavelength long. The model shows a pronounced null in the direction of the vertical wire, but not as severe as with the modified building model. This model was done to show that a nearby wire CAN affect the radiation pattern, contrary to your statements. Your response was:

“In PhilB's 2nd modification, he changed the height of the parasitic radiator from 0.035 wavelengths to 0.25-wavelengths, located it 41 feet away from the 3-m whip, and used a 2-ohm connection to a perfect ground plane. This configuration is radically different than the original NEC model.”

And:

“That 2-ohm ground connection is typical of the set of 120 each, 1/4-wave long buried radials used by licensed AM broadcast stations, and would be an unlikely scenario for a Part 15 user. Also the 41-foot spacing between the two wires is 25.5 degrees long at 1.7 MHz, which is not particularly near in that sense.”

I chose 41 ft. because it is at the center of the original building model. That doesn’t seem radically different from the original spacing from antenna to the nearest wall of 20 ft. If closely spaced wires “behave almost exactly like a unit radiator”, then it seems appropriate to use the center of the building.

You essentially say the quarter-wave wire is behaving correctly in the model, but choose not to believe the results of the 50k and 2ohm grounded building model I proposed. You have contradicted yourself and have not proven why the modified building model does not represent the real world of physics.

Here is a composite of 4 models so other readers can see my model results:

http://sstran.com/RFry_building_model_results.gif

 

[rfry]

I chose 41 ft. because it is at the center of the original building model. That doesn’t seem radically different from the original spacing ...

That is not the difference in your modifications that is radical. It is your changes of the height of the parasitic radiator from 20 feet to 143.4 feet (1/4 wavelength at 1700 kHz), and using 2 ohms in the ground connection of wire 10, and of the 1/4-wave wire.

You essentially say the quarter-wave wire is behaving correctly in the model, but choose not to believe the results of the 50k and 2ohm grounded building model I proposed.

Yes, in hindsight I should have included a definition of closely spaced wires. As it turns out, a spacing of 41 feet for the 1/4-wave parasitic is not very close in terms of wavelengths on 1700 kHz, and the net radiation pattern becomes directional.

However when the 1/4-wave wire is located only 5 feet from the 3-m whip then the net pattern is nearly circular, as shown in my plot of yesterday, and supports my statement about the h-plane radiation pattern of such a radiating source as highly circular. Of course, I haven't defined "nearly" and "highly," either, and I guess there are arguments there.

So the phrase "closely spaced" does not apply well for distances from the wires in the structure to the 3-m whip. My mistake.

You have contradicted yourself and have not proven why the modified building model does not represent the real world of physics.

All of the NEC models linked in this thread represent "the real world of physics," according to NEC and the assumptions used in defining the models. What is unknown is how well the models represent the real world, itself. That judgment is left to readers.

RF

 

[rfry]

In my slight modification of your building wire frame model, I demonstrated that that the building can have a dramatic effect on the radiation pattern. A very large null occured in the direction from the antenna to the wire 10 corner of the building. The only changes I made were to change the resistance to ground of all legs of the building, except for wire 10, from 200 ohms to 50,000 ohms, and I changed the resistance to ground in wire 10 from 200 ohms to 2 ohms.

In followup, it should be noted that a resistance of 50,000 ohms to ground in all of the vertical wires of the structure except Wire 10 with 2 ohms means that Wire 10 would radiate nearly all of the vertically polarized, parasitic energy from that structure.

Based on that I removed all but one of the wires in the structure, leaving only Wire 10 with 2 ohms to ground, spaced from the 3-m whip the same as in the original NEC model. The net radiation from that pair of conductors was posted earlier at http://i62.photobucket.com/albums/h85/rfry-100/PhilB_Pattern_Check.gif .

That radiation pattern is very circular, which supports my original contention that the radiation pattern of closely spaced wires is "highly circular," and casts doubt on the accuracy of PhilB's model of the structure using 50,000-ohm ground connections, showing a directional pattern.

RF

 

[PhilB]

R. Fry:

I agree with your results for a single vertical wire at the wire 10 position. However, that is not relevant to the topic of this thread, which pertains to the effect of a nearby building on the radiation pattern. In the building model, current is flowing in all of the wires. They sum at wire 10 and go to ground through the 2 ohm ground resistance in wire 10 to ground.

Here is a geometry view, showing the currents in all the wires:
http://sstran.com/RFry_3m_ant_with_50k_bldg_gnds_and_2ohm_wire10_currents.gif

The complex pattern of currents flowing in all the wires cannot be duplicated with a single vertical wire. With all other legs essentially being disconnected from ground and only one connection (wire10) to ground, the building looks a lot like a vertical antenna with a huge, offset top hat with multiple verticals dropping almost to ground level around the periphery of the huge top hat.

 

[rfry]

In the building model, current is flowing in all of the wires. They sum at wire 10 and go to ground through the 2 ohm ground resistance in wire 10 to ground

Although I suspect that only 2-3 people are continuing to read this thread any more, I'll add another comment to it.

The currents shown along the metal frame in PhilB's recent posting are a higher percentage of the maximum current on the 3-m whip than my NEC program shows for those conditions (see link below).

There may be some differences with the adaptation made by PhilB for 4NEC2, as in my original NEC input file the maximum current applied to the 3-m whip is 1 ampere (at the center of segment 2), and PhilB's version shows 5.85 amperes. Might there be others?

The link below shows that the currents on all of the conductors in the metal frame except for Wire 10 (with the 2-ohm ground connection) are vanishingly small. Therefore their sum at Wire 10 would not have much affect on the total radiation from that wire.

http://i62.photobucket.com/albums/h85/rfry-100/PhilBs_Currents_RFry_version.gif

RF

 

[druidhillsradio]

Am I one of the 3?