Callsign
ad: giga-rw
Page 1 of 4 1 2 3 4 LastLast
Results 1 to 10 of 35

Thread: Modeling a "3-D folded" loop antenna?

ad: L-HROutlet
ad: l-assoc
ad: l-rl
ad: l-gcopper
ad: l-Waters
ad: l-tentec
ad: l-innov
ad: l-AmericanRadio
ad: l-BCInc
  1. #1
    Join Date
    Jul 2009
    Location
    Ishigaki Island (Okinawa)
    Posts
    59

    Default Modeling a "3-D folded" loop antenna?

    I'm back at home and back on track... Going into round 2:

    After a lengthy exchange in this forum, concerning an undersize lazy loop antenna that I had envisioned ( see
    http://forums.qrz.com/showthread.php?t=259736&page=4 ), I had come to the conclusion that it would probably be useless to pursue the matter any further, since short loop just don:t do the trick, and so I turned my attention to a kind of antenna called "cobra dipole". But somehow my mind kept working on the loop idea, and today a weird contraption came out - a kind of lazy loop partially folded under itself, that would allow me to put upthe full length of wire, and so I would like to solicit some comments on this new brain burp.

    I used EZNEC Demo to draw the thing up and to run a simulation on it, and I referenced the frequency sweep results to 450 Ohm, a value easily achievable by using a 1:9 balun fed via 50 Ohm coax. The software gave me the following results for highest SWR on each of the bands I checked for (in parentheses: lowest SWR):

    10m: 2.3 (1.42 at 28.7 MHz)
    12m: 1.47 (1.4 at 24.9 MHz)
    15m: 1.67 (1.1 at 21.3 MHz)
    17m: 1.68 (1.1 at 17.8 MHz - out of band)
    20m: 2.77 (1.73 at 14.05MHz)
    30m: >20 - not near a resonance
    40m: >very high SWR and Q; might possibly work in a small part of the band with a 1:1 balun

    EZNEC Demo only permits 20 segments for the calculations, so I am curious whether those results are realistic. If anybody here has EZNEC at hand and wouldn't mind downloading my data file to model the antenna with a much larger number of segments, I would very much appreciate hearing about the results.

    Here is the data file for the antenna:
    http://www.qsl.net/js6shs/yuuri_loop4.EZ
    (just in case, the same file as archive: http://www.qsl.net/js6shs/yuuri_loop4.EZ.zip )

    Some background information:

    - island location; QTH at abt. 24N 124E
    - distances to the closest areas with a larger population: 275km (170mi) to the west and 400km (250mi) to the northeast
    - possibility of high winds at any time of the year; occasional tropical cyclones
    - reinforced concrete building; flat roof at almost 8m/25ft above ground
    - within the roof area I can conveniently make use of an area 15m/49ft long and 8.5m/28ft wide
    - no garden or yard to speak of
    - license conditions: max. 100W on 160m-2m in accordance with the band plan

    What I am looking for (ideally) is an antenna with the following characteristics:

    - capable of operating on all (*) RF bands
    - easy to match to the Rx/Tx
    - high angle radiation on the low bands and low angle radiation on the high bands
    - low wind resistance
    - should fit on the roof
    - easy to build
    - easy to take down (before a typhoon, etc.)
    - low cost

    (*) In view of the limited space, the fact that QRN is high here (near the tropic), and given the very low ham population in this general area, 160m and 80m will be useless in my situation anyway, so I am not concerned about those band - but I definitely want the antenna to cover the WARC bands.

    As always, comments and suggestions are most welcome!

    Thanks in advance and 73s from JS6SHS

  2. #2

    Default

    Quote Originally Posted by JS6SHS View Post
    EZNEC Demo only permits 20 segments for the calculations, so I am curious whether those results are realistic. If anybody here has EZNEC at hand and wouldn't mind downloading my data file to model the antenna with a much larger number of segments, I would very much appreciate hearing about the results.
    I get similar results running the model with increased segmentation - in fact the VSWRs turn out slightly better than the ones you quote.

    There's a quite good 50Ω match on 80m.

    Radiation patterns are complex, as you would expect.

    I'm sure I've seen a design very similar to this, but can't yet track it down.

    PS: Found it - "The 3D Folded Loop Antenna" by WX7G in Antennex Magazine:
    http://www.dxzone.com/cgi-bin/dir/jump2.cgi?ID=15199

    73,
    Steve G3TXQ

  3. #3
    Join Date
    Oct 2003
    Location
    Middle Georgia USA
    Posts
    7,722

    Default

    I looked the other day at folding a loop and I ran into phase cancelation problems between close spaced wires either on the lowest band, or on all higher bands. I did not spend a lot of time with it but at first glance it looked impossible to correct.

    Depending on how I folded the loop the antenna either had very low radiation resistance on the lower band, or on all the higher bands.


    At lowest SWR points I've got:


    4 ohms resistance on 3.8 MHz
    25 ohms on 7.2
    150 ohms on 10.6 MHz
    255 ohms on 14.05 Mhz
    500 ohms on 17.85 MHz
    415 on 21.5 MHz
    600 ohms on 25.2 MHz
    730 ohms on 28.5 MHz



    I'd either use a bent dipole and ladder line, or figure out a way to open and short a half wave loop for 80 meters.

  4. #4
    Join Date
    Jul 2009
    Location
    Ishigaki Island (Okinawa)
    Posts
    59

    Default

    Thanks for the feedback again!

    Quote Originally Posted by G3TXQ View Post
    I get similar results running the model with increased segmentation
    Good to know - even the demo version is obviously a good investment - as promised.

    the VSWRs turn out slightly better than the ones you quote.
    That's interesting...

    There's a quite good 50Ω match on 80m.
    I got something like 4Ω, which would make the whole thing very lossy. Will check once more. But, anyway, not to worry about 80m - I can't use that here anyway. Just to double-check: what resistive and reactive values did you get for 40m? I think I saw something aorund 150Ωfor the resistive component. I also wonder whether the demo version is perhaps not sufficient to deal with situations where the wires are relatively short with regard to the wavelength.

    Radiation patterns are complex, as you would expect.
    Yeah, there will be what people call "vertical components" and "horizontal components", and parts of the thing will perhaps work like a phased array, except that that effect is "random" and not changeable in the space I have.

    I'm sure I've seen a design very similar to this, but can't yet track it down.

    PS: Found it - "The 3D Folded Loop Antenna" by WX7G in Antennex Magazine:
    http://www.dxzone.com/cgi-bin/dir/jump2.cgi?ID=15199
    Oh - most interesting! That surely is essentially the same idea! Now, he uses it as a vertical radiator on its fundamental frequency and feeds it on one of the long legs, but the principle is the same. Well, it looks like we might have much wider applications for this concept now. And in recognition of what is obviously "prior art"and seeing the name he gave his thing, I'd might call my contraption "3-DFLL" ("3-dimensionally folded lazy loop") and not "quirky loop" as I had been inclined to at first. Anyway, great fun, this antenna stuff - and I appreciate your help!

    Now, once the typhoon induced bad weather lets up, I'll get on the roof and put some wire up to see how this quirky loop works in reality.

    73s!

  5. #5
    Join Date
    Jul 2009
    Location
    Ishigaki Island (Okinawa)
    Posts
    59

    Default

    Tom, thank you, too, for the feedback. Looks like things are falling into place after all.

    Quote Originally Posted by W8JI View Post
    Depending on how I folded the loop the antenna either had very low radiation resistance on the lower band, or on all the higher bands.
    I see. That would mean my current design is the one I want to stick with, since I can let go of the lower bands.

    At lowest SWR points I've got:

    4 ohms resistance on 3.8 MHz
    25 ohms on 7.2
    150 ohms on 10.6 MHz
    255 ohms on 14.05 Mhz
    500 ohms on 17.85 MHz
    415 on 21.5 MHz
    600 ohms on 25.2 MHz
    730 ohms on 28.5 MHz
    That's quite like what I got!

    I'd either use a bent dipole and ladder line, or figure out a way to open and short a half wave loop for 80 meters.
    Not to worry about 80m - I won't need that. Seeing how promising the values are from 20m upward, I think it will be good enough to use a balun ( 450Ω to 50Ω) and let a regular (unbalanced) ATU take care of the SWR . I will, of course, fine-tune the length of the antenna on the roof first, to get the best compromise for the resonances.

    Many thanks again!

    73s from a wet and windy island

  6. #6
    Join Date
    Jul 2009
    Location
    Ishigaki Island (Okinawa)
    Posts
    59

    Default

    A drawing showing the outline of the antenna

    http://www.qsl.net/js6shs/3-dfll.gif

  7. #7

    Default

    I seem to be seeing something different - not sure why:

    3.8MHz: 36+j34
    7.1MHz: 78-j101
    10.1MHz: 255-j541
    14.1MHz: 305-j15
    18.07MHz: 535-j22
    21.2MHz: 478-j228
    25.0MHz: 558-j198
    28.5MHz: 766-j222

    Try using "High Accuracy Ground".

    We may be running foul of one of EZNEC's "quirks" - I often see strange impedances during the first few 'runs' which can't be repeated when you re-visit exactly the same conditions later on.

    Try looking at the feedpoint impedance on 3.8MHz; then try changing the ground type, the copper losses etc, and then return to the original settings. My feedpoint resistance shifts from 4 Ohms to 34 Ohms.

    IIRC it may have something to do with the Ground File tolerance setting.

    Steve G3TXQ
    Last edited by G3TXQ; 10-23-2010 at 01:39 PM.

  8. #8

    Default

    I was almost speechless when I saw this! I tried this almost a year ago, BUT on a much smaller scale and with more twisty bends. I figured you could make this thing on the order of maybe 5 or 6cm, for 2m/1.25m/70cm. Or maybe use it as a loading coil for a longer antenna. Check this out! It took FOREVER... I thought I was nuts, yet I had to continue making it...



    It didn't work so well (it worked allright on 70cm) so I went back to the bending-board:



    THIS was the result. A giant hilbert-curve. I dunno if I still have this thing around, though I kept the smaller one because it looks cool.

    I'm going to try this...

  9. #9
    Join Date
    Jul 2009
    Location
    Ishigaki Island (Okinawa)
    Posts
    59

    Default

    Quote Originally Posted by G3TXQ View Post
    I seem to be seeing something different - not sure why:

    3.8MHz: 36+j34
    7.1MHz: 78-j101
    10.1MHz: 255-j541
    14.1MHz: 305-j15
    18.07MHz: 535-j22
    21.2MHz: 478-j228
    25.0MHz: 558-j198
    28.5MHz: 766-j222
    Yes, your's is certainly a more manageable value for 80m - but it looks like 80 and 40 won't be workable if i use a 1:9 balun (450 Ohm) to match the upper bands. However, with a carefully chosen length of feed line it may be possible to make this antenna work OK on all bands without a balun. That would be the next thing to figure out.

    Try using "High Accuracy Ground" - it makes a massive difference to the 80m feedpoint impedance.
    Need to check how to do that. For my model I arbitrarily chose a ground conductivity of 1S (several magnitudes higher than the default value).

    In any case, the real-life version of this antenna here will be at fairly low height on the flat top of a re-inforced concrete building. No idea yet sure how good/bad an RF ground this will turn out to be and how the impedance values will turn out in that situation, but I would not be surprised if all Z values were to turn out somewhat different than the model.

    Thanks & 73s!

  10. #10
    Join Date
    Jul 2009
    Location
    Ishigaki Island (Okinawa)
    Posts
    59

    Default

    Quote Originally Posted by VA3CQC View Post
    I was almost speechless when I saw this! I tried this almost a year ago, BUT on a much smaller scale and with more twisty bends. I figured you could make this thing on the order of maybe 5 or 6cm, for 2m/1.25m/70cm.
    Problem is, the smaller an antenna is in relation to the full size, the less efficient it tends to be. My manipulated lazy loop won't be as good as a full-size loop, and the only reason I needed to find something like the 3-DFLL is that I want a loop that easily matches on the harmonics (and not a dipole, vertical, or other antenna with ends "sticking out") but don't have the space for a full size 80m loop.

    A giant hilbert-curve
    It DOES look cool! But it also looks like it might induce a headache when you try to think about the impedance, the phase relationships of the various segments, and the resulting radiation pattern.

    73s!

Page 1 of 4 1 2 3 4 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •