ad: ARR

My final(?) attempt to improve Cobweb bandwidth

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by SP3L, Jul 12, 2017.

ad: L-HROutlet
ad: l-rl
ad: MessiPaoloni-1
ad: L-MFJ
ad: Left-2
ad: Left-3
ad: Subscribe
  1. SP3L

    SP3L Ham Member QRZ Page

    I described some modifications of the G3TXQ Cobweb in the other thread. I called them Mark 1 and Mark 2. I built Cobweb Mk.1 myself and Elias, EB5KT, built Cobweb Mk.2. Mark 2 was better because it had wider bandwidth on the 10 m band. However, bandwidth improvement came at a cost. The radiation patterns of the modified Cobwebs were closer to a bidirectional dipole-like pattern. Since that time, I modeled a number of other modified Cobwebs but until recently all my ideas did work very well. Finally, I found something worth calling Cobweb Mk3. In my opinion, it compromises very little in the radiation pattern while giving big improvement in bandwidth. See the performance data below.

    G3TXQ Cobweb
    upload_2017-7-12_15-47-12.png

    SP3L Cobweb Mk. 3
    upload_2017-7-12_15-47-54.png

    All antenna gains are referred to a dipole and thus expressed in dBd. If you need the value in dBi, increase the numbers by 2.15. Both antennas were simulated in free space. Gmax in in the X-X direction and Gmin roughly in Y-Y direction in all the drawings that will follow. AGT results are =< 0.2 dB what mean the model is credible.

    The Mark 3 is designed around the same supporting structure as the regular model: four spreaders to keep the corners of square loops and a fifth one to support the 1:4 balun. But in some of the spots where the corners of the loops were normally attached attached to the spreaders, I added vertical conducting wires. My ideas is to use brass threaded rods diameter 4 mm (M4). When using threaded rod, it will not be difficult to attach it to a spreader and to attach antenna wires to the rod. You need just several nuts and washers. Eight 80 cm long and four 66 cm long rods are required. The structure looks like in the drawing below.

    upload_2017-7-12_15-48-39.png

    And the antenna model looks like that:

    upload_2017-7-12_15-49-9.png

    I used Cu wire of diameter 2 mm in the model. 20 m and 17 m band radiators are as in the regular Cobweb. Only the internal radiators had to be modified. All required dimensions are shown in the pictures below.

    10 m band radiator:
    upload_2017-7-12_15-50-7.png

    12 m band radiator:
    upload_2017-7-12_15-52-27.png

    15 m band radiator:
    upload_2017-7-12_15-53-3.png

    Remaining dimensions:
    upload_2017-7-12_15-53-41.png

    The dimensions were adjusted for the antenna in free space. When installed over real ground, the end wires will require some adjustment depending on height of installation and ground properties. Keep in mind that all dimensions matter. For example if you used longer brass rods, you would have to change quite a number of other dimensions. So, better do not.

    And, no, I did not build it. And I am not going to. I have got the Cat’s Whiskers with a rotator and I am more than happy with it. But if I were limited on space and/or could not use a rotator I would build just the Cobweb Mark 3.

    73
    Jacek, SP3L
     
    KD6RF likes this.
  2. M0MNM

    M0MNM Ham Member QRZ Page

    Looks nice....and complicated! Pity you're not gonna build it...

    Nice work! :)
     
  3. KC8VWM

    KC8VWM Moderator Volunteer Moderator QRZ Page

    In my experience, the only band that is really lacking in full coverage bandwidth on the cobweb is 10m.

    The cobweb won't cover from 28.000-29.700 but then again, i'm not sure there's many other antennas that do either.

    The Mark 3 cobweb is an excellent design and solution to limited bandwidth issues, but I think most people simply operate the cobweb where the majority of activity is occurring on the 10m band. In my case, I tune the antenna somewhere around 28.300. That gives me enough bandwidth to cover everything from 28.100 to 28.500 with less than a 1.5:1 VSWR.

    Certainly useable beyond those points on the band too, but that's where maximized performance is likely to occur.
     
    Last edited: Jul 24, 2017
  4. SP3L

    SP3L Ham Member QRZ Page

    Well, it seems that my final attempt was not a final one. Here is yet another idea how to improve the Cobweb bandwidth. The concept is to use capacitance hats at the ends of 4 radiators (20 m radiator is left unchanged). My simulations revealed that the biggest improvement can be achieved if different cap-hat types are used in the neighboring radiators. Evidently such configuration reduces the coupling effect in closely spaced radiators. The Mark 4 Cobweb looks like in the drawing below.

    upload_2017-11-30_10-35-34.png

    At the ends of 10 m and 15 m radiators, I used T-shaped cap-hats. In the 12 m and 17 m radiators - trident-shaped cap-hats. The 20 m band radiator is left unchanged. So the overall X-Y dimensions are the same as in the original G3TXQ Cobweb. The table below compares the SWR and gain performance of both antennas. Average Gain Test (AGT) results are very low, so the models can be trusted.

    upload_2017-11-30_10-40-32.png

    Improvement in the SWR at the 15 m and 10 m band ends is undeniable. Also the minimum gain (Gmin) did not suffer and is very close to the original design. At the same time, Gmax slightly improved. So, the improvement in bandwidth did not happen at the cost of making the antenna less omnidirectional.
    I think that Mark 4 is easier to build than Mark 3 while providing similar performance.

    I am quite satisfied with these results. It is time to archive the models and stop thinking about the Cobwebs.

    Jacek, SP3L
     
  5. KD6RF

    KD6RF XML Subscriber QRZ Page

    Interesting work - Tnx!
     
  6. SM6CJB

    SM6CJB Ham Member QRZ Page

    Really interesting!

    How about making another set of spreaders and place them above the fist set.
    This would make it possible to install folded dipoles with a great distance between the horisontal parts.
    If you connect the feeder only to the lower (or upper) part of the folded dipole, it will work as the G3TPW design, and will not need the balun.
    The stucture would resemble a cubical quad resting on its back.

    Jan/SM6CJB

    PS.: Of course the name of the antenna would be "Lazy Quad"!:)
     
  7. M0MNM

    M0MNM Ham Member QRZ Page

    I extended the dimensions of my G3TXQ, to make the gap between the ends of the elements larger, and it doesn't seem as good as it used to be. Am I imagining it? Should I put it back how it was?

    Your experiments are certainly interesting, but I don't really need mine for 10m. :)
     
  8. SP3L

    SP3L Ham Member QRZ Page

    I modeled a number of various shapes. Also something similar to what you are describing as a Lazy Quad but with not so big distance between the upper and lower parts. I did not get interesting results. I can not claim that better designs than Mark 3 or 4 do not exist. Simply, I did not find them.
     
  9. SP3L

    SP3L Ham Member QRZ Page

    Increasing the original Cobweb in order to get larger gaps between the radiator ends improves SWR a little bit but the change is not dramatic. At the same time, the antenna becomes more bi-directional (so less omnidirectional). Its minimum gain (i.e. the gain in Y-Y direction) goes down. So, doing only such a change does not produce very noticeable improvement.

    If you still can do that, return to the original size and try the Mark 4 modification. Be the first to build it. If you are interested, I will give you all dimensions of the cap-hats. But you will have to develop the mechanical part yourself.
     
  10. M0MNM

    M0MNM Ham Member QRZ Page

    If I needed it for 10m, I would definately have a go at it. But I have a vertical, and the improvements on 15m wouldn't be worth all the fiddling for me. :)

    Thanks for your answer! I only moved the intersections out by an inch at each corner, so I'm guessing I've not messed it up much! :)

    It's so difficult to tell with such varying conditions...but it's probably in head as I thought!

    Keep it up! Have you considered modelling 2 phased cobwebs on top of each other? That I may be able to get on board with! :cool:
     

Share This Page