Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: Prev 1 ... 3 4 5 6 7 [8]   Go Down
  Print  
Author Topic: Small multiturn wire loops for transmitting  (Read 59384 times)
JAHAM2BE
Member

Posts: 287


WWW

Ignore
« Reply #105 on: April 12, 2014, 11:26:46 PM »

Some more analysis of the potential efficiency of a 4-turn square loop (0.84m x 0.84m) with 24mm-diameter conductors at 7 MHz.

I noticed that due to current non-uniformity in the simulated antenna, the detrimental effect of added loss resistance depends on the location of that resistance within the antenna. If the loss resistance is in a low-current area of the antenna, losses will be low; in a high-current area, the losses will be higher. In my previous simulations of the 3D multi-turn loop and the 2D planar multi-turn loop, I was placing the additional loss resistance (200 milliohms) in a low-current portion of the antenna, at or near the capacitor. As we can see, the current is lowest in this portion of the structure (the horizontal bar at the top):


When I place the additional loss resistance in the high-current (red) area of the conductor, for example in line with the bottom corner of the third turn from the left, then the efficiency drops to about 20 percent (for the 3D multi-turn loop).

In real life, when constructing this antenna out of four turns of wide copper strap, any additional loss resistance will likely be in the capacitor or connections to the capacitor, which would seem to be a low-current area of the antenna, so I think it conceivable that 50% efficiency may be achievable.

Now, are these numbers believable? In other words, is it believable to achieve between 20% and 50% efficiency for a 4-turn square loop (0.84m x 0.84m) at 7 MHz? I think the results are believable based on the analytic formulas posted earlier in this thread:

Let's do some calculations for a small multiturn loop. R = (177NS/lambda^2)^2, S = loop area
[...]
Skin depth in copper at 7 MHz is 0.03 mm. The resistance per square is 560 u ohms.

Jasik, Henry, Antenna Engineering Handbook, 1st Edition, McGraw-Hill, (c) 1961, Library of Congress Catalog Card Number: 59-14455, page 6-2:

"6.3. Radiation Resistance, The radiation resistance of a small air loop is:"

"Rrad = 320(pi4A2N2)/lamda4"

Calculating the radiation resistance with the formula provided by WX7G gives:

S loop area=0.84m * 0.84m
N=4 turns
lambda=43m
Rrad=(177 * 4 * .84 * .84 / (43^2))^2
Rrad = 0.073 ohms

Calculating the radiation resistance with the formula provided by W5DXP gives:
pi=3.14
A=0.84m * 0.84m
N=4 turns
lambda=43m
Rrad=320*(3.14^4 * (0.84*0.84)^2 * 4^2)/(43^4)
Rrad=0.072 ohms

So both formulas (which are just rearrangements of one another) are in agreement with a radiation resistance of about 0.073 ohms.

The loss resistance for copper is 560 microohms per square. With four turns each of length 0.84*4 meters, we have 0.84*4*4 meters or 13.44 meters length of copper tubing. With 24mm-diameter copper tubing, the available copper width for current flow is the tube circumference, pi*0.024m = 0.07536 m. So we have a total conduction area of 13.44m / 0.07536m = 178.344 squares. Multiplying that by the loss resistance per square for copper gives us 178.344 squares * 560 microohms/square = 0.1 ohms loss resistance.

Assuming no other loss resistance, we get an efficiency of Rrad / (Rrad + Rloss) = 0.073 / (0.073 + 0.100) = 42 percent efficiency.

Assuming an additional 0.2 ohms of loss resistance, we get an efficiency of 0.073 / (0.073 + 0.300) = 20 percent efficiency.

Analytical results of between 20 and 42 percent efficiency are in good agreement with the simulated results of between 20 and 50 percent efficiency.

There is still the proximity effect to worry about, which will introduce additional losses and is not accounted for in either the 4nec2 simulation or the analytical formulas, but with 4 wire diameters spacing between turns, or with the use of flat strap geometry with turns laid edge to edge, the proximity effect losses should not be overly large.

I also ran 4nec2 simulations on a 42cm x 42cm x 50cm antenna for 20m, which also comes in at about 50% efficiency. I may try to build the 20m version first to verify that the idea works before investing the time and materials into the larger 40m version.

===========

Here's another possibly feasible 50%-efficient 7 MHz variant: a 7-turn square loop, side length 0.64m, total length 1m, made of copper pipe 12mm in diameter. A long 20m roll of soft 12mm copper tubing should be rollable by hand into this structure with no joints and minimal loss. Resonating capacitance for 7 MHz is 6 pF.



« Last Edit: April 13, 2014, 12:31:39 AM by JAHAM2BE » Logged

JAHAM2BE
Member

Posts: 287


WWW

Ignore
« Reply #106 on: April 14, 2014, 01:08:52 AM »

the more I think about it the more I am sure that once can readily construct a rather simple model of the helically wound loop.

The following patent presents and analyzes a toroidal helix antenna.

http://www.google.com/patents/EP0043591A1?cl=en

I didn't have time to go through the paper in detail yet, but it looks rather interesting and is quite long.

---

Also, the long multi-turn loop antennas I presented are, I think, properly called "normal mode helical antennas". There seems to be quite a bit of research on them, for example:  http://ap-s.ei.tuat.ac.jp/isapx/2008/pdf/1645313.pdf .
Logged

N3OX
Member

Posts: 8847


WWW

Ignore
« Reply #107 on: April 15, 2014, 05:06:38 AM »

Keep in mind that the analytical formulas you're comparing with probably assume uniform current in all the turns.

Also, have you checked to see what the radiation resistance and efficiency computed by 4nec2 are when you use perfect ground with loss turned off?
Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
JAHAM2BE
Member

Posts: 287


WWW

Ignore
« Reply #108 on: April 15, 2014, 07:46:06 AM »

Keep in mind that the analytical formulas you're comparing with probably assume uniform current in all the turns.

Good point. That would seem to imply that the analytical formulas would predict higher efficiency than the simulation (which shows some degree of current non-uniformity). However, I am getting better efficiency in the simulator than the analytical formula predicts. Go figure. Smiley

Also on the topic of current distribution, one thing that is evident from the current intensity diagrams I showed is that with the long, nearly-self-resonant helices, most of the current is flowing within the self-resonant structure formed by the coil and its self-capacitance. The additional resonating capacitance, which is very small (on the order of 5-10 pF at 7 MHz) and has a comparatively high reactance, has very little current flowing through it because of the near-self-resonant condition of the rest of the structure.

Also, have you checked to see what the radiation resistance and efficiency computed by 4nec2 are when you use perfect ground with loss turned off?

Unfortunately I forgot how to do that - I do recall you mentioned it once here on eham, but I can't find the post. For the 7-turn helix I just now tried setting the ground to perfect and all wire conductivity to "perfect". I re-ran the optimizer to calculate the resonating capacitance such that the reactance was as close to zero as possible, which gave a resonating capacitance of 5.85 pF. Then, the resulting efficiency was 100% and the impedance at the feedpoint (identical location to the capacitor) was given as 1.95 + j 4.03. (The j 4.03 reactive component is close to, but not equal to, zero due to the optimizer not being able to optimize tiny sub-pF capacitance values to completely eliminate the reactive component.) Note that no attempt was made to match the antenna to 50 ohms.

Anyway, what should I be looking for to find the radiation resistance? Is it 1.95 ohms, the real part of the computed impedance? (That seems rather high, but I lack a good intuition about expected radiation resistance of normal mode helices.)

The 7-turn model file is on my blog here: http://qrp-gaijin.blogspot.jp/2014/04/a-simulated-40m-multi-turn-small.html

From a theoretical standpoint I find it more interesting that the smaller-volume, flat pancake coil (with thick conductors) can apparently achieve a similar efficiency to the larger-volume helix. However, the flat and thick-conductor pancake coil is difficult to physically construct.

By the way, I saw on your website that you may "implement some of the short antenna ideas I've worked on over the years in an apartment setting again." Any news on that front? Smiley I constructed a version of your short balcony dipole with large capacitance hats, but I'm wondering about if the concrete on my balcony is more detrimental to the extreme near field of an E-field antenna like a dipole than to that of an H-field antenna like a magloop (where the concrete-balcony-to-antenna distance is less than 0.1 wavelength, placing it in the extreme near field of the antenna). It's these worries about the effect of concrete on the electric field that are leading me to again look at magloops, with their electric field confined to the capacitor area.
Logged

W5WSS
Member

Posts: 1783




Ignore
« Reply #109 on: April 15, 2014, 12:24:54 PM »

Yes it is conceivable  to a limited distance that perhaps a magnetic field can better survive passing through a loss material such as a dielectric MINUS a series of conductor/s but remember that the distance a magnetic field travels is nearby that point which it changes to an all electromagnetic field.

following this thread closely.

73
Logged
PA1ZP
Member

Posts: 296




Ignore
« Reply #110 on: April 15, 2014, 02:43:52 PM »

Hi to you all

My experiences purely practical.

Small loops are best in a circle and with only 1 turn.
More turns lead to more resistive losses only.
The best is to use the extra copper for more area of the loop then more turns.

I build 1 loop with a diamater of 1 mtr (3.3 ft) and i build 1 with 2 mtrs diameter (6ft).
I tested both a while and found the bigger loop much better on 40 mtrs lets say a 4 dB.

But it was still lacking about 12 db over my vertical GP to Canada on 40 mtrs (300 miles)

Do not forget that very large amounts of electric charge currents are going in and out of the capacitor and use big bulcky leads with very low losses to your capacitor.
And use a high voltage and high current capacitor, I got it easily to loaded to 3KV with 100 watts.

My findings, they are small those loops , expencive to build enourmes expencive to buy.

i did not find any magnetic magic in them , it just behaved as any other small loop or dipole.

I had nice experiments with them, but these things are only an option if you have no room for other more traditional antennas.
A simple GP will beat them big big time.

73 Jos
Logged
N3OX
Member

Posts: 8847


WWW

Ignore
« Reply #111 on: April 22, 2014, 06:20:38 AM »

Quote
Unfortunately I forgot how to do that - I do recall you mentioned it once here on eham, but I can't find the post. For the 7-turn helix I just now tried setting the ground to perfect and all wire conductivity to "perfect". I re-ran the optimizer to calculate the resonating capacitance such that the reactance was as close to zero as possible, which gave a resonating capacitance of 5.85 pF. Then, the resulting efficiency was 100% and the impedance at the feedpoint (identical location to the capacitor) was given as 1.95 + j 4.03.


It's good that you're getting 100% efficiency. I wanted to make sure you weren't getting 120% or 150%

Quote

Anyway, what should I be looking for to find the radiation resistance? Is it 1.95 ohms, the real part of the computed impedance? (That seems rather high, but I lack a good intuition about expected radiation resistance of normal mode helices.)

You need to be a little careful. Some people prefer a definition of radiation resistance where the impedance is computed at current maximum. It's okay to do it somewhere else, like at the low-current capacitor location, provided that you also *correct the loss resistance* for the impedance transformation you get from the current distribution (and which you've already mentioned... )

The location next to the cap where you're feeding is like feeding a dipole near one of the ends. The real part of the impedance is still quite high and it's due entirely to radiation, but if you insert a 50 ohm resistor in the middle of a dipole fed 2/3 of the way toward the end, you're not going to see that resistance you're measuring go up by 50 ohms Smiley

If you feed at the current maximum, what is the real part of the impedance? This is one common definition of the "radiation resistance," the value at current max. How does it compare with the analytical number for a multiturn loop?


Quote

By the way, I saw on your website that you may "implement some of the short antenna ideas I've worked on over the years in an apartment setting again." Any news on that front? Smiley

Nah, I've had very little interest in ham radio since moving to NYC. Trying not to do things that keep me holed up in my apartment these days.
Logged

73,
Dan
http://www.n3ox.net

Monkey/silicon cyborg, beeping at rocks since 1995.
JAHAM2BE
Member

Posts: 287


WWW

Ignore
« Reply #112 on: April 22, 2014, 06:24:08 AM »

I found a nice article by W6NBC about using normal-mode helix antennas on 2m:

http://w6nbc.com/articles/2011-06QST2mhelices.pdf

He uses a smaller coupling loop to excite the helix, just like the coupling loop in a normal "magnetic loop". His helices are self-resonant, requiring no resonating capacitor. This does make initial tune up somewhat tedious, and limits the loop to one band.

He also mentions the same principles can be used at HF, which is basically what I was doing with my multi-turn loop models shown in my previous postings.

Strangely, there don't seem to be many amateur examples of normal mode helices at HF - an eham search on "normal mode helix" gives only a handful of results.

What I find interesting about the self-resonant, normal mode helix is that it acts like a magnetic loop (maximum radiation in the plane of the loop, use of a small coupling loop for excitation, ground independence, acceptable performance even at low heights), but with the major advantage that it requires no high-value, low-loss tuning capacitor, unlike a normal magnetic loop. The radiation resistance is also higher than a single-turn loop.

I am slowly moving forward with the plans for a four-turn 40m normal mode helix (about 1m diameter, 10cm spacing between turns), to be built with wide copper strap wound around a support frame. I think I can tune it by altering the spacing or angle between the second and third turns - in effect, making the helix into a variable inductance (i.e., a physically large variometer). The whole antenna would fit into a 1m x 1m x 1m volume and, according to my simulations, can achieve greater than 50% efficiency at 7 MHz and about -7 dBi gain at 15 degrees above the horizon. This same efficiency and low-angle gain are also achievable with a similarly-sized traditional single-turn magnetic loop that uses a large-diameter-conductor or multiple-parallel-conductors, but such an approach would require a high-value and low-loss variable capacitor and careful attention to minimizing loss in the connections to the capacitor. The self-resonant normal-mode helix seems like an elegant way to sidestep the requirements of "low-loss capacitor" and "low-loss connections to the capacitor".

Has anyone here experimented with normal-mode helix antennas at HF? They seem like a good solution for a single-band, compact, high-efficiency antenna.
Logged

JAHAM2BE
Member

Posts: 287


WWW

Ignore
« Reply #113 on: April 22, 2014, 06:51:41 AM »

If you feed at the current maximum, what is the real part of the impedance? This is one common definition of the "radiation resistance," the value at current max. How does it compare with the analytical number for a multiturn loop?

For the 7-turn helix (0.64m x 0.64m, conductor diameter 12mm; the file is on my blog) at 7 MHz, and with wire conductivity set to perfect and over perfect ground, 4nec2 reports the real part of impedance at the current maximum (bottom of the middle/fourth turn) as 0.11 ohms. The analytical formula reports the radiation resistance of such a multi-turn loop as 0.076 ohms. Huh Close, but not the same... In free space, by the way, the real part of impedance at the current maximum is reported as 0.18 ohms.

For the 0.84m x 0.84m 4-turn loop at 7 MHz, 4nec2 reports the real part of impedance at the current maximum to be 0.2 ohms, while the analytical formula gives 0.074 ohms. Again, a somewhat unsettling discrepancy.

There was a recent post by WX7G (http://www.eham.net/ehamforum/smf/index.php/topic,96348.msg752285.html#msg752285) about getting dubious radiation resistance values for small loops in EZNEC... perhaps I'm running into a similar issue? I tried increasing the segmentation of my wires but that didn't change the results significantly.

Quote from: JAHAM2BE
By the way, I saw on your website that you may "implement some of the short antenna ideas I've worked on over the years in an apartment setting again." Any news on that front? Smiley

Nah, I've had very little interest in ham radio since moving to NYC. Trying not to do things that keep me holed up in my apartment these days.

Well, then, I shall have to do my part to try and push forward new ideas for apartment antennas, by building and reporting on the small 40m normal-mode helix. Stay tuned...
« Last Edit: April 22, 2014, 07:20:41 AM by JAHAM2BE » Logged

JAHAM2BE
Member

Posts: 287


WWW

Ignore
« Reply #114 on: Today at 02:26:53 AM »

I found a couple of recent unique-looking loop designs using loops in parallel, similar to AA8C's loops.

The way the top of the loops are angled to converge on the capacitor is interesting and unusual. Perhaps the use of the angled connections minimizes the current path length. If only right angles were used, the current from the outer-most loops would need to make a right angle turn to go inwards toward the capacitor, whereas with the angled connections, the current takes the shorter diagonal path to the central capacitor.

http://qrz.com/db/G3JKF


http://mw1cfn.blogspot.jp/2014/12/building-g3jkf-magnetic-loop-array.html
Logged

Pages: Prev 1 ... 3 4 5 6 7 [8]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!