Advanced search  
Pages: 1 [2]   Go Down

Author Topic: The 3105 Donut  (Read 27875 times)

Bob Brandenburg

  • TIGHAR member
  • *
  • Posts: 33
Re: The 3105 Donut
« Reply #15 on: September 15, 2011, 09:42:57 AM »

Gary,

How did you do your calculations?  By hand, or were you using a propagation model?

BTW, a couple of points re Chuck's previous remarks:  an antenna radiation pattern has no polarization -- the radiated field is polarized, but not the radiation pattern.  I suspect Chuck meant to say the pattern is vertically oriented, i.e. that its central axis is vertical.  Also, the "back story" of the antenna model change is not related to the pattern dimple -- both models show a dimple, that was never an issue -- but rather that NEC2 much more accurately models an antenna close to ground.  Before 4NEC2, the NEC2 input/output process FORTRAN punch-card format and was unwieldy.   4NEC2 added a GUI that greatly facilitates using the model.   

Bob
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #16 on: September 15, 2011, 03:04:20 PM »


BTW, a couple of points re Chuck's previous remarks:  an antenna radiation pattern has no polarization -- the radiated field is polarized, but not the radiation pattern.  I suspect Chuck meant to say the pattern is vertically oriented, i.e. that its central axis is vertical.

Bob,

Thank you for keeping me on my technical term toes. I have edited the first paragraph in the subject post (reply #10, this thread) to clarify what was intended.

Quote
Also, the "back story" of the antenna model change is not related to the pattern dimple -- both models show a dimple, that was never an issue -- but rather that NEC2 much more accurately models an antenna close to ground.

Thank you, too, for clarifying what I referred to as the "back story". I can't recall all that motivated my supposition, but in part it included Ric's words in a 27 Jan 09 forum post (italics are mine):

"As explained in the most recent TIGHAR Tracks (Oct. 2008, "The 3105 Donut"), computer modeling software that has become available since 2001 has revealed an anomaly in NR16020's transmission propagation pattern which changes the whole ball game about where the Electra could have been at 1912 and 2055 (the time of the last transmission heard by Itasca)."

That was coupled with having read how you viewed, and appeared to have modeled, the dorsal V antenna for so long. Take, for example, these words of yours from a 23 Oct 01 forum post:

"In order for there to be a vertical electrostatic field component, and hence a ground wave, it is necessary that the radiating antenna be vertical.  If the antenna is horizontal, the electrostatic field is parallel to the earth's surface, and is immediately shorted out, causing the entire wave front to collapse, thus preventing ground wave propagation.  And there's the rub.  The Electra's antenna was horizontal, hence there was no groundwave."

The last sentence is incorrect at 3105 and 6210 kHz, of course, as the electric field component at those frequencies was predominantly vertical. (Because electrostatic fields are pretty much reluctant to radiate, I suspect you meant to use the term electric field.)

Quote
Gary, How did you do your calculations?  By hand, or were you using a propagation model?

You asked Gary how he did his calculations. I wonder the same for you. In reply #4 in this thread you told Ken Brookner that you used ICEPAC. My question is How did you use it for an aircraft in flight? Do you have a version that will accept an elevation input?

Chuck
Logged

Bob Brandenburg

  • TIGHAR member
  • *
  • Posts: 33
Re: The 3105 Donut
« Reply #17 on: September 15, 2011, 03:42:54 PM »

Quote
The last sentence is incorrect at 3105 and 6210 kHz, of course, as the electric field component at those frequencies was predominantly vertical. (Because electrostatic fields are pretty much reluctant to radiate, I suspect you meant to use the term electric field.)

I did mean the electric field.   And why was the electric field component vertical at those frequencies?


Quote
You asked Gary how he did his calculations. I wonder the same for you. In reply #4 in this thread you told Ken Brookner that you used ICEPAC. My question is How did you use it for an aircraft in flight? Do you have a version that will accept an elevation input?

There is no version that accepts elevation input, nor is there a need for one.  For skywave propagation, an aircraft's altitude is insignificant in the context of the ionospheric refraction geometry

Bob
« Last Edit: September 15, 2011, 07:20:18 PM by Martin X. Moleski, SJ »
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #18 on: September 16, 2011, 10:30:00 AM »

I did mean the electric field.   And why was the electric field component vertical at those frequencies?

Because at those frequencies the currents in the generally vertical wires contributed more to the radiated field than those in the generally horizontal wires, making the net polarization of the field vertical.

Does incredulity prompt your asking?

Quote
There is no version that accepts elevation input, nor is there a need for one.  For skywave propagation, an aircraft's altitude is insignificant in the context of the ionospheric refraction geometry

OK, thanks. You do use the antenna's at-altitude gain pattern for an aircraft's in-flight transmission, don't you?

BTW, are you aware of my 11 July post to you?

Chuck
« Last Edit: September 16, 2011, 04:57:51 PM by Martin X. Moleski, SJ »
Logged

Bob Brandenburg

  • TIGHAR member
  • *
  • Posts: 33
Re: The 3105 Donut
« Reply #19 on: September 16, 2011, 02:43:52 PM »

Not necessarily incredulity, but I am curious.   The term " generally vertical wires" suggests one or more such wires other than the one through the starboard cabin wall.   Where would it/they be?  What are the relative amplitudes of the vertical and horizontal currents?

Re: the at-altitude gain pattern, since the fuselage essentially is ground as far as the antenna is concerned, it doesn't matter how far above geographical ground the plane is.

Re: your 11 July post, I wasn't aware until you just mentioned it.   I somehow missed it.   I'm open to insights on the harmonic power paper -- or anything else. 

Bob
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #20 on: September 16, 2011, 07:52:22 PM »

Not necessarily incredulity, but I am curious.   The term " generally vertical wires" suggests one or more such wires other than the one through the starboard cabin wall.   Where would it/they be?  What are the relative amplitudes of the vertical and horizontal currents?

I was using "generally vertical" to refer to all wires exclusive of those forming the legs of the V.

I've attached a screen shot (file: MMANA-GAL_screen) from the antenna simulation software I use most often: MMANA (MININEC3-based), or in this specific case, MMANA-GAL, which is the current iteration. It shows a results summary for the dorsal V at 3.105 MHz and its harmonics. Forget the details of what's being modeled for the moment. Note that the far right column gives the polarization of the radiated field, and that it's vertical for 3.105 and 6.21 MHz, but horizontal for higher frequencies.  That's the depth to which I typically go on the polarization issue. 

I've attached a screen shot (file: 3105_currents) showing an example of the segment currents for the V at 3105 kHz.  The current sums give some feel, but it's how their effects combine that tells the story. That's what the simulation software is for (and I've never yearned to learn how it does it). In this case the "generally vertical" currents sum to a value very slightly greater than those of the "generally horizontal".  I intended to include a screen shot for the 12.42 MHz case, the first one for which MMANA reports a horizontal polarization, but because I used a constant segment size of wavelength/100 it gave far too many lines to include in a screen shot. It gave a sum for the "generally horizontal" segment currents that was 16 times that of the "generally vertical". Again, a feel, but not the whole story.

We can continue with this until you feel less uncomfortable. If it's going to take a "why" to satisfy, I'll have to work at it, and that will take some motivation and time.

Quote
Re: the at-altitude gain pattern, since the fuselage essentially is ground as far as the antenna is concerned, it doesn't matter how far above geographical ground the plane is.

I tend to disagree with that statement in general, but let's look at some specific cases. You indicated in January that you've now modeled the antenna on a wire frame model of the Electra. If you put it at 1000 feet above sea water (I use dielectric constant 80 and conductivity 5 S/m), does it give the same pattern that you get with the fuselage bottom a couple feet above real ground (say 0.6 m and dielectric constant 13 with conductivity 0.005 S/m)? How about above perfect ground at same height? Try it in free space. What does the pattern look like when compared to the other cases?

When I model the antenna over a conductive mat, I get quite different patterns for the three cases.
I've attached a vertical panorama of screen shots (file: dorsal_V_patterns...) of the total, vertical, and horizontal patterns that I got with the dorsal V over a conductive mat 1000 feet above sea water. Do you see anything like this with your model?

Quote
Re: your 11 July post, I wasn't aware until you just mentioned it.   I somehow missed it.   I'm open to insights on the harmonic power paper -- or anything else.
 
Let's finish this thread first, then move on to that.

Chuck
Logged

Bob Brandenburg

  • TIGHAR member
  • *
  • Posts: 33
Re: The 3105 Donut
« Reply #21 on: September 19, 2011, 10:58:46 AM »

Chuck,

A few questions to help me understand your results:

1) Where on the aircraft are wires 5 and 6, which "are vertical and connect to the lead-in and ground"?

2) Did your test runs involve only the antenna in isolation -- i.e. without the fuselage?

3) Does MMANA-GAL have the same antenna height limitation as MININEC-3, i.e. 0.2 wavelength?

Bob
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #22 on: September 19, 2011, 02:08:20 PM »


1) Where on the aircraft are wires 5 and 6, which "are vertical and connect to the lead-in and ground"?

Consider wires 5 and 6 together to be one vertical wire, call it wire 5*, that is 29 inches long. (Wire 5* was split into two wires, 5 and 6--each 14.5 inches long, as it afforded a convenient way in MMANA to place the source at five different heights above ground. [A MMANA option is to place the source at the beginning, center, or end of a wire]).

[9/20/11 additional comment: I typically made the lengths of wires 5 and 6 equal, but sometimes used different proportions to place the source in a desired location. I saved such a case in January, and see that I grabbed that one for the examples attached to my 16 Sep 11 post. For that case, the lengths of wires 5 and 6 were 6 inches and 23 inches, respectively.]

Wire 5* connects to the inboard end of the lead-in and drops vertically to "ground", which in the case of the pattern attachment to my previous message, is a conductive mat (a patch, in Geometry Builder) with dimensions as given in the annotation.

Quote
2) Did your test runs involve only the antenna in isolation -- i.e. without the fuselage?

That's correct. Nothing beyond the mat that allowed modeling it off the ground. The aforementioned pattern attachment to my previous message shows the modeled antenna configuration (in blue) in each of the four pattern presentations.

Quote
3) Does MMANA-GAL have the same antenna height limitation as MININEC-3, i.e. 0.2 wavelength?

If there's a height limitation, I'm ignorant of it. What, specifically, was the 0.2 wavelength limitation for MININEC3--and the effect of ignoring it?

Chuck
« Last Edit: September 20, 2011, 07:24:21 AM by Chuck Varney »
Logged

Bob Brandenburg

  • TIGHAR member
  • *
  • Posts: 33
Re: The 3105 Donut
« Reply #23 on: September 20, 2011, 09:35:36 AM »

Quote

Wire 5* connects to the inboard end of the lead-in and drops vertically to "ground"

Quote
2) Did your test runs involve only the antenna in isolation -- i.e. without the fuselage?

That's correct. Nothing beyond the mat that allowed modeling it off the ground.

Quote
3) Does MMANA-GAL have the same antenna height limitation as MININEC-3, i.e. 0.2 wavelength?

If there's a height limitation, I'm ignorant of it. What, specifically, was the 0.2 wavelength limitation for MININEC3--and the effect of ignoring it?

Chuck

It's interesting that you ran the lead-in directly to ground.  On the aircraft, the lower end of the lead-in connected to the antenna terminal on the transmitter.

Modeling the antenna in isolation ignores the aircraft structure.  The antenna field induces currents in the structure, which generate field components that interact with the antenna field.

The antenna height limitation is discussed briefly in the MMANA-GAL user notes, and more extensively in the web literature. 

Bob
« Last Edit: September 20, 2011, 07:19:50 PM by Martin X. Moleski, SJ »
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #24 on: September 20, 2011, 05:47:10 PM »

It's interesting that you ran the lead-in directly to ground.
 
It's interesting that you didn't note that the path to ground included a source (a "transmitter").

Quote
On the aircraft, the lower end of the lead-in connected to the antenna terminal on the transmitter.

And through the transmitter, and on to a connection to the airframe ("ground"). If I had information regarding the actual path, and path length, from the feedthrough insulator to the airframe--and if I thought that incorporating it would materially alter the characterization of the antenna--I would have used it.

What do you have to offer with regard to the path particulars?

Quote
Modeling the antenna in isolation ignores the aircraft structure.

Of course.

Quote
The antenna field induces currents in the structure, which generate field components that interact with the antenna field.

No argument. I'm interested in hearing some details of what accounting for the interaction revealed to you.

Quote
The antenna height limitation is discussed briefly in the MMANA-GAL user notes, and more extensively in the web literature.
 
Are you referring to differences between NEC-2 and MININEC-3 feed-point impedance and gain results for horizontal antennas modeled close to ground?  If so, do you know that the differences are germane to the issue of modeling the dorsal V, say at 7-8 feet above ground at 3105 kHz? If the answer is in the affirmative, how close would the NEC-2 and MININEC-3 impedance and gain results need to be for you to judge them insignificantly different? Perhaps you could try it yourself and we could compare results.

Perhaps you can clarify something for me. I thought the two topics we were pursuing were:

1) The polarization issue (whether the dorsal V, by virtue of its largely horizontal array of wire, could not radiate a ground wave [that is, was not vertically polarized], as you contended).

2) The radiation pattern vs. height above ground issue (whether altitude is immaterial to propagation modeling as far as the pattern goes, since it's the same in the air as on the ground, as you contended.)

What is your conclusion regarding 1)?

How about 2)?

What did you find out with regard to the following, which I asked about in Reply #20 four days ago:

You indicated in January that you've now modeled the antenna on a wire frame model of the Electra. If you put it at 1000 feet above sea water (I use dielectric constant 80 and conductivity 5 S/m), does it give the same pattern that you get with the fuselage bottom a couple feet above real ground (say 0.6 m and dielectric constant 13 with conductivity 0.005 S/m)? How about above perfect ground at same height? Try it in free space. What does the pattern look like when compared to the other cases?

Chuck
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #25 on: September 21, 2011, 03:11:12 PM »

Bob,

This post regards my "3105 Donut back story" ( Reply #9, paragraph three) and your revision to it in Reply #16, where you wrote:

Quote
. . .Also, the "back story" of the antenna model change is not related to the pattern dimple -- both models show a dimple, that was never an issue -- but rather that NEC2 much more accurately models an antenna close to ground.

In Reply #17, I accepted your revision to my version and gave two examples of what led me to believe what I had written. There were others that didn't come readily to hand at the time I responded. One of them follows.

These are your words, as appended to a Ric Gillespie forum post from 9 January 2009,12:25:25:

Quote
. . .With the increased speed available on a modern PC, and with the GUI, NEC4WIN95 was the most practical tool for TIGHAR's purposes at the time. As it turned out, the model missed the "dimple" that was later found by 4NEC2.

In 2009 you said the "dimple" was revealed when 4NEC2 (NEC-2-based antenna modeling software) replaced NEC4WIN95 (MININEC-3-based antenna modeling software).

In 2011 you said the dimple was never the issue--both software programs showed it.

Please explain.

Chuck
Logged

Bob Brandenburg

  • TIGHAR member
  • *
  • Posts: 33
Re: The 3105 Donut
« Reply #26 on: September 22, 2011, 10:21:59 AM »

The phrase "dimple that was missed" should be read in the context of the fact that NEC2 handles antennas close to ground better than NEC4WIN95.  The dimple is there in both, but is deeper and steeper in NEC2. 

Bob 
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #27 on: September 22, 2011, 01:38:54 PM »

The phrase "dimple that was missed" should be read in the context of the fact that NEC2 handles antennas close to ground better than NEC4WIN95.  The dimple is there in both, but is deeper and steeper in NEC2.

Bob,

The following is from your The Post-Loss Radio Signals: Technical Analysis paper. You are describing the radiated power patterns you obtained for operation at 3105 kHz and 6210 kHz.

". . .It is interesting to consider the effective radiated power from the NR16020 antenna used in this analysis.

At 3105 kHz, the radiated power is 0.3 watts at takeoff angles less than 15 degrees, with a smooth nonlinear taper up to 6 watts at takeoff angles above 80 degrees. Perpendicular to the main-lobe axis, the radiated power is 0.003 watts at takeoff angles less than 5 degrees, tapering up to 6 watts for takeoff angles greater than 75 degrees.

At 6210 kHz, the radiated power is 0.2 watts at takeoff angles less than 5 degrees, with a smooth nonlinear taper up to 40 watts at takeoff angles greater than 80 degrees. Perpendicular to the main-lobe axis, the radiated power is 0.06 watts at takeoff angles less than 5 degrees and 40 watts at takeoff angles greater than 80 degrees."

These are not descriptions of dimpled patterns.

Please explain.

Chuck
« Last Edit: September 22, 2011, 02:27:26 PM by Chuck Varney »
Logged

Bob Brandenburg

  • TIGHAR member
  • *
  • Posts: 33
Re: The 3105 Donut
« Reply #28 on: September 22, 2011, 03:02:12 PM »

That was eleven years ago, Chuck.  What's your point?

Bob
Logged

Chuck Varney

  • T2
  • **
  • Posts: 94
Re: The 3105 Donut
« Reply #29 on: September 23, 2011, 11:17:41 AM »

That was eleven years ago, Chuck.  What's your point?

Bob,

My point is that you would have the general reader think that, because you now understand a thing, all the papers you authored reflect that understanding. They don't. The excerpt I gave from your The Post-Loss Radio Signals: Technical Analysis paper is one example.

Chuck
Logged
Pages: 1 [2]   Go Up
 

Copyright 2019 by TIGHAR, a non-profit foundation. No portion of the TIGHAR Website may be reproduced by xerographic, photographic, digital or any other means for any purpose. No portion of the TIGHAR Website may be stored in a retrieval system, copied, transmitted or transferred in any form or by any means, whether electronic, mechanical, digital, photographic, magnetic or otherwise, for any purpose without the express, written permission of TIGHAR. All rights reserved.

Contact us at: info@tighar.org • Phone: 610-467-1937 • Membership formwebmaster@tighar.org

Powered by MySQL SMF 2.0.15 | SMF © 2017, Simple Machines Powered by PHP