Fractal TV Antenna Research & Development

[300ohm]

Just for the heck of it, after watching the NOVA PBS program on fractals, I fractalized Autofils GH4 VHF antenna. Heres a picture of what it looks like :



Each of the element pieces are 6 inches, the stubs are 12 inches. The VHF gain is pretty darn good at 12.5 to 13 db with a SWR of close to 1 for the range of 174 to 216 mhz (channels 7 to 13). The UHF gain is pretty miserable at 5 db or below.
Basically by fractalizing the antenna, its size is reduced 20%, going from 24 inches per leg to 20 inches. Also by making more bends, the structure of it is stiffened somewhat. Ill work on fractalizing the GH10 UHF antenna, to make that monster a bit smaller. I know I can get the antenna even smaller by reiterating the fractalization, but as a practical matter it would drive anyone nuts trying to bend and build it, heh. (with the GH4 VHF, you may be able to do one more reiteration without it being impossible)

Im not planning on building this GH4 VHF antenna because I also need good channel 6 performance in an antenna, but if anyone else is interested, I could draw up the simple plans on graph paper.

Heres the NEC file :

CM GH4 (VHF-Hi) Collinear Rods: Fractalized by 300ohm
CM Array and reflector elements = 5/16" diam Alu
CM All element pieces 6 inches, Stubs are 12 inches.
CM AGT = 1.0 (0 db)
CE
GW 1 3 0 1.875 -28.970589 0 6.11764055 -24.727948 0.1377953
GW 2 3 0 12.1176539 -18.727935 0 16.3602945 -14.485295 0.1377953
GW 3 3 0 6.11764055 -24.727948 0 12.1176406 -24.727935 0.1377953
GW 4 3 0 12.1176539 -18.727935 0 12.1176406 -24.727935 0.1377953
GW 5 3 0 16.3602945 -14.485295 0 12.1176539 -10.242654 0.1377953
GW 6 3 0 6.11764055 -4.2426406 0 1.875 0 0.1377953
GW 7 3 0 12.1176539 -10.242654 0 12.1176402 -4.2426536 0.1377953
GW 8 3 0 6.11764055 -4.2426406 0 12.1176402 -4.2426536 0.1377953
GW 9 3 0 1.875 0 0 6.11764055 4.2426405 0.1377953
GW 10 3 0 12.1176539 10.2426539 0 16.3602945 14.4852945 0.1377953
GW 11 3 0 6.11764055 4.2426405 0 12.1176406 4.2426543 0.1377953
GW 12 3 0 12.1176539 10.2426539 0 12.1176406 4.2426543 0.1377953
GW 13 3 0 16.3602945 14.4852945 0 12.1176539 18.727935 0.1377953
GW 14 3 0 6.11764055 24.7279484 0 1.875 28.970589 0.1377953
GW 15 3 0 12.1176539 18.727935 0 12.1176402 24.7279354 0.1377953
GW 16 3 0 6.11764055 24.7279484 0 12.1176402 24.7279354 0.1377953
GW 17 3 0 -1.875 0 0 -6.117641 -4.2426406 0.1377953
GW 18 3 0 -12.117654 -10.242654 0 -16.360295 -14.485295 0.1377953
GW 19 3 0 -6.117641 -4.2426406 0 -12.117641 -4.2426544 0.1377953
GW 20 3 0 -12.117654 -10.242654 0 -12.117641 -4.2426544 0.1377953
GW 21 3 0 -16.360295 -14.485295 0 -12.117654 -18.727935 0.1377953
GW 22 3 0 -6.117641 -24.727948 0 -1.875 -28.970589 0.1377953
GW 23 3 0 -12.117654 -18.727935 0 -12.117641 -24.727935 0.1377953
GW 24 3 0 -6.117641 -24.727948 0 -12.117641 -24.727935 0.1377953
GW 25 3 0 -1.875 28.9705891 0 -6.117641 24.7279485 0.1377953
GW 26 3 0 -12.117654 18.7279351 0 -16.360295 14.4852946 0.1377953
GW 27 3 0 -6.117641 24.7279485 0 -12.117641 24.7279347 0.1377953
GW 28 3 0 -12.117654 18.7279351 0 -12.117641 24.7279347 0.1377953
GW 29 3 0 -16.360295 14.4852946 0 -12.117654 10.242654 0.1377953
GW 30 3 0 -6.117641 4.2426406 0 -1.875 0 0.1377953
GW 31 3 0 -12.117654 10.242654 0 -12.117641 4.2426536 0.1377953
GW 32 3 0 -6.117641 4.2426406 0 -12.117641 4.2426536 0.1377953
GW 33 3 0 1.875 28.970589 0 6.11764055 33.2132295 0.1377953
GW 34 3 0 12.1175415 39.2131305 0 16.3600698 43.4558835 0.1377953
GW 35 3 0 6.11764055 33.2132295 0 12.1176406 33.2131305 0.1377953
GW 36 3 0 12.1175415 39.2131305 0 12.1176406 33.2131305 0.1377953
GW 37 3 0 -16.360295 -43.470589 0 -12.125 -39.220615 0.1377953
GW 38 3 0 -6.1176406 -33.21323 0 -1.875 -28.970589 0.1377953
GW 39 3 0 -12.125 -39.220615 0 -6.1176539 -39.213229 0.1377953
GW 40 3 0 -6.1176406 -33.21323 0 -6.1176539 -39.213229 0.1377953
GW 41 3 0 -16.360104 43.4560165 0 -12.1176 39.2132391 0.1377953
GW 42 3 0 -6.117641 33.2132296 0 -1.875 28.9705891 0.1377953
GW 43 3 0 -12.1176 39.2132391 0 -12.117641 33.2132391 0.1377953
GW 44 3 0 -6.117641 33.2132296 0 -12.117641 33.2132391 0.1377953
GW 45 3 0 1.875 -28.970589 0 6.117641 -33.21323 0.1377953
GW 46 3 0 12.117654 -39.213243 0 16.3529486 -43.463217 0.1377953
GW 47 3 0 6.117641 -33.21323 0 6.117654 -39.21323 0.1377953
GW 48 3 0 12.117654 -39.213243 0 6.117654 -39.21323 0.1377953
GW 49 3 0 16.3600698 43.4558835 0 28.3600698 43.455411 0.1377953
GW 50 3 0 -16.360104 43.4560165 0 -28.360079 43.455411 0.1377953
GW 51 3 0 16.3529486 -43.463217 0 28.3529483 -43.460591 0.1377953
GW 52 3 0 -16.360295 -43.470589 0 -28.360286 -43.470272 0.1377953
GW 53 3 0 -1.875 0 0 1.875 0 0.132
GW 54 11 -10.94 -1 43.455 -10.94 -35 43.455 0.132
GW 55 11 -10.94 1 43.455 -10.94 35 43.455 0.132
GW 56 11 -10.94 1 -43.455 -10.94 35 -43.455 0.132
GW 57 11 -10.94 -1 -43.455 -10.94 -35 -43.455 0.132
GW 58 11 -10.94 -1 14.485 -10.94 -32.5 14.485 0.132
GW 59 11 -10.94 1 14.485 -10.94 32.5 14.485 0.132
GW 60 11 -10.94 1 -14.485 -10.94 32.5 -14.485 0.132
GW 61 11 -10.94 -1 -14.485 -10.94 -32.5 -14.485 0.132
GS 0 0 0.0254 ' All in in.
GE 0
EK
LD 5 0 0 0 2.49e7 0
EX 0 53 2 0 1 0
GN -1
FR 0 8 0 0 174 6
RP 0 1 10 1510 90 0 0 20 0 0
EN

[testproggie]

I would be interested in the hand drawn version. I will only have upper VHF and UHF channels after the switchover. And I have a channel 13 problem child. There are two from nearly around 70 degrees (i think) apart and it's a challenge to get the weaker although closer one. Oddly enough they are both the same network . However the antenna spends most of it time pointed toward the weaker one because that's where most of the other channels we watch come out of and my wife doesn't like fooling with the rotor. I have a mid priced UHF/VHF that I use currently for the VHF and a 4228 for the UHF. I wanted to get a new VHF up before winter really sets in and I was debating buying versus building. I haven't found a VHF that I was interested in building yet. This one, however, shows some promise. I have some material already acquired to build a SBGH2 but, I wasn't going to put it up till spring. Now I can use that to build this one instead, if the gain numbers are as promised. The antenna I was thinking about purchasing was the Winegard YA 1713, which I think would have been adequate for my situation. Thanks for any help you can provide to get me started.

[300ohm]

And I have a channel 13 problem child. There are two from nearly around 70 degrees (i think) apart and it's a challenge to get the weaker although closer one. Oddly enough they are both the same network

That does sound like a challange. Can you post your TVFool report or your zip code / coordinates so that we can see whats going on ?

Gain isnt the only problem in a case like that, but rather a high F/B ratio and narrow beam pattern would do it. Unfortunately, that GH4_VHF antenna above has a nice wide pattern, which is one of its advantages. One way to narrow the beam pattern would be to put an 8 ft by 8 ft screen mesh behind it. It can be done, but isnt not a real practical thing to do IMO.
I would go with the Winegard YA 1713 as you had planned, and see if it does the trick.

I have a mid priced UHF/VHF that I use currently for the VHF

Do you know the make and model of that UHF/VHF combo ? Maybe it can be retuned for better channel 13 performance. Models of the popular Radio Shack UHF/VHF combo antennas are available.

[testproggie]

TV Fool now

and after the switchover
http://i85.photobucket.com/albums/k6...-Digital23.png

They are actually 143 degrees apart (mind is going ) I see now the situation reverses after the switchover. 13 from 190 degrees will be stronger than 13 from 333 degrees. This will probably help my situation enough to make a difference (I hope). I will still put up a new VHF antenna of some type because the current one just doesn't give me the signal level I expect from these two. Obviously a huge screen behind to help with the antenna described above in this case is prohibitive if I want it on the top of that tower. Thanks for your input BTW.

[300ohm]

I see now the situation reverses after the switchover. 13 from 190 degrees will be stronger than 13 from 333 degrees. This will probably help my situation enough to make a difference (I hope).

Its only a small difference, so I wouldnt get my hopes up. Yeah, the GH4 VHF antenna is very large height and width wise, as you can see below. Thats why yagis and lpdas are more practical, they take up less height.

Looking at Ken Nists site, hdtvprimer.com/ANTENNAS/comparing.html, youll notice that a lot, if not most commercial antennas like to drop off after channel 12. (I guess they figure channel 13 unlucky, heh)
Thats why if you know the make and model of your antenna, and if its one of the ones he has listed, I may be able to model a retune of it. Most likely all that would be involved is a little snipping.

[300ohm]

The GH4 VHF Fractal build plan:
=============================



Forgive the poor quality compressed scan. Heres a recap of the dimensions. The overall height is 86.91 inches. The overall width is 70.0 inches. Modeling was assuming the use of 5/16 inch diameter aluminum rod for the elements and reflectors.
Reflector dimensions:
--------------------
The middle co-linear reflector pair rods are 31.5 inches long each. The top and bottom co-linear reflector pair rods are 34.0 inches long each. The co-linear reflector rod gap is 2.0 inches between the rods. (34 plus 34 plus 2 gives the 70 inch overall width) The reflector to active element distance is 10.94 inches. Tolerances for the VHF-Hi band would be around 3/8ths of an inch.
Active element dimensions:
--------------------------
The stubs are 12.0 inches. All other element distances are 6 inches between bends. All the angles are either 90 degrees or 135 degrees. The feedgap distance between the elements is 3.75 inches.

Raw Gain Chart:


SWR Chart:


Not too shabby at all. Only trouble is the height of the thing, heh.

Making another fractal reiteration would mean the distances between bends would become 1 1/2 inches versus the 6 inches now (and many more of them), which would mean the overall width and height would decrease another approx 7 inches.

[Xauto]

300ohm: I like the shape. I have change my double diamond to your frac. shape,to test. My question is can the stubs be shorten by division of 4 and then making it look like this -v- ?

[300ohm]

My question is can the stubs be shorten by division of 4 and then making it look like this -v- ?

Actually in all practicality yes. I just ran a model of that, replacing all 4 straight 12 inch stubs with -v- 10.2426405 inch ones. (3 inches plus (3 inches times square root of 2) plus 3 inches equals 10.2426405 inches total length of the new stub)

Top gain went from 13.0 dbi to 12.93 dbi. No big whopt, a 12.93 dbi VHF-Hi antenna is still a very powerful antenna, heh.

On another note, by removing the reflector entirely, you get a very nice 180 degree pattern antenna that increases smoothly linearly in raw gain from 8.18 dbi at 174 mhz to 10.2 dbi at 216 mhz, with a slight increase in SWR to around 2. That still compares very nicely against those antennas listed at hdprimer.

[DogT]

300ohm,

Wouldn't it be easier to use the SY cards to make a scalable antenna for any frequency? I'm not an expert at the 4nec2, but have played around with the SY cards for some other antennas, and I think all the elements and spacings could be put in a "symbol" and then change the frequency and use the F12 key to run different situations or frequencies? I have taken some models and replaced specific dimentions with variables (symbols) and it seems to work with the limited use I have done. I would love to see a fractal GH10 or even a fractal DBGH10. If I can make the time, I could possibly work on translating your GH4 into a scalable SY card.

By the way my old DBGH had some problems with the solder coming loose at some of the joints, need to make better mechanical connections, but right now I have it up on a pole only about 20' and it is not working to my desired expectations, so I have a new rope over a live branch at about 50' so I will probably be hauling it up there pretty soon since the leaves are off the trees now. The Old Lady is wanting to watch something on CH4 on Thursday, so maybe tomorrow if it is not too cold or windy.

Turned 65 today, WOO HOO. Medicare here we come.

DogT

[300ohm]

Wouldn't it be easier to use the SY cards to make a scalable antenna for any frequency?

Yes and no. 4nec2 already has a built in feature to Rescale any or all wires, just by entering in the Old frequency, New frequency and the Scaling factor. So that makes that particular task pretty much a no-brainer, heh. Geometry editing in 4nec2 is a very powerful and easy to use feature, once you get the hang of it. It also has features to copy and paste any wire or any group of wires, and the ability to rotate any or all wires around any of the 3 axis, by any amount of degrees. SY card formulas are very easy to lose in 4nec2 without careful saving, so I try to use them mostly only when optimizing a particular variable.

I would love to see a fractal GH10 or even a fractal DBGH10. If I can make the time, I could possibly work on translating your GH4 into a scalable SY card.

Oh yeah, there is tons of work yet to be investigated. For example, whats better, an outie or innie bend ? Or maybe alternating the bends or some other pattern ? Or maybe a /5 or /6 fractal works significantly better than the /4 fractal. Or maybe some wild equation fractal. The possibilities are endless, heh. I just used the /4 fractal because it was one of the first ones by Dr Nathan Cohen that looked like it was in style with the GH shape.
On the other hand, practical aspects have to be considered. Who would want to be making 90 degree bends every darn 1/4 of an inch, heh.
From what I have modeled and have read so far, fractalizing doesnt really produce any more gain, but decreases size and increases bandwidth some. Of course, by decreasing element size and leaving the reflector size the same, that should generally result in some more gain.
Oh and from what Ive also read, all current modeling engines dont handle fractals that well, which is another reason I wouldnt feel confident modeling more than about 2 reiterations.

[Xauto]

I have made two gen1 SBGH's. One to print and one at 104%. I seened that 300ohm was working a fractal. I now have rebent the 104% to a fractal. I have tested it today and it looks like there is no difference in reception, it may be working better with my undersized slit screen 27X36 inches for Ch14.

[300ohm]

Very nice Xauto.

Yeah, even though I havent modeled the GH UHF fractal yet, IMO performance shouldnt go down hardly at all from a non-fractalized version. But its big advantage is the approx 20% size reduction. And maybe some increased bandwidth ??

Also I noticed that your top stubs go -v- and the bottom stubs go -^-. The model I did in the previous post above had them all going -v-. Probably no significant difference.

And I see you corrected my model drawing error, on the last leg before the stubs, heh. Again, it probably makes no significant difference. (the NEC file and the Raw Gain and SWR figures are from the model as drawn, even with my cross-eyed drawing)

Tape some aluminum foil to another big piece of cardboard (25mm slit in the middle) and put it behind the antenna with some 100mm spacers for some real gain.


The innie and outie bends (-v- and -^-) are one of the first things Ill try to investigate as to which is better.

[Xauto]

Thanks 300ohm. I'm very pleased with the looks. If I build another it will be a fractal (out of wire for now). I have a sheet of cardboard with Al.-foil and 20mm gap. Work's good. All inside. I have a 4" boad that gives me the 100mm space. The element on one side the reflector on the other. The boad has two screws on each side (paper punched hole in the carboard). Just hang, make change antennas easy. From 4 bay bowtie(8.8X8) to GH(gen2&gen1,frc.) to folded dipole in an X-array.

[300ohm]

I'm very pleased with the looks

Yes, it does have that modern sculpture work of art look to it.

Just hang, make change antennas easy.

Yeah, that would be my favorite way of making an attic antenna, nice, easy and cheap. I would use styrofoam panels (which I have) to mount the elements on instead of cardboard because of the lower conductivity of styrofoam. But again, its only a teeny minor difference, unless the cardboard got wet.

[300ohm]

Heres a picture of the antenna models to be compared. For simplicity, the reflectors will be left off for now.



On the far right is the original SBGH gen1 model with a 89mm feedgap, 7 inch straight legs, 5.6 inch stubs.
On the far left is the /4 fractalized version of the original SBGH gen1 model with a 89mm feedgap. The size of each of the original legs has been subdivided into 4 sections and the 2 middle sections bent with the 90 degree shape as shown.
In the middle is a SBGH model with straight legs that is the size of the fractalized version. It has a 89mm feedgap, 6 inch legs, and 4.75 inch stubs. Including this antenna model helps to give insight as to what is going on with UHF fractalization.

Net (not Raw) Gain Summary (in dbi):

--------------- SBGHg1 ---------- SBGH------------- SBGHg1
Mid------------ 7 inLeg ----------6 inLeg -----------Fractalized
Channel--------Net Gain----------Net Gain-----------Net Gain
14 ********** 5.51 ********** -1.64 ********** 2.12
15 ********** 5.92 ********** -1.00 ********** 2.80
16 ********** 6.32 ********** -0.37 ********** 3.46
17 ********** 6.69 ********** 0.24 ********** 4.08
18 ********** 7.04 ********** 0.84 ********** 4.67
19 ********** 7.37 ********** 1.42 ********** 5.23
20 ********** 7.69 ********** 1.98 ********** 5.75
21 ********** 7.98 ********** 2.53 ********** 6.22
22 ********** 8.26 ********** 3.06 ********** 6.67
23 ********** 8.52 ********** 3.57 ********** 7.08
24 ********** 8.76 ********** 4.07 ********** 7.43
25 ********** 8.98 ********** 4.54 ********** 7.75
26 ********** 9.19 ********** 5.00 ********** 8.03
27 ********** 9.39 ********** 5.43 ********** 8.28
28 ********** 9.57 ********** 5.84 ********** 8.49
29 ********** 9.73 ********** 6.24 ********** 8.66
30 ********** 9.88 ********** 6.61 ********** 8.81
31 ********** 10.02 ********** 6.96 ********** 8.93
32 ********** 10.14 ********** 7.29 ********** 9.04
33 ********** 10.26 ********** 7.61 ********** 9.11
34 ********** 10.36 ********** 7.89 ********** 9.18
35 ********** 10.45 ********** 8.17 ********** 9.24
36 ********** 10.53 ********** 8.41 ********** 9.28
37 ********** 10.60 ********** 8.65 ********** 9.31
38 ********** 10.66 ********** 8.86 ********** 9.34
39 ********** 10.71 ********** 9.05 ********** 9.37
40 ********** 10.75 ********** 9.22 ********** 9.40
41 ********** 10.78 ********** 9.39 ********** 9.42
42 ********** 10.79 ********** 9.53 ********** 9.45
43 ********** 10.79 ********** 9.66 ********** 9.48
44 ********** 10.77 ********** 9.78 ********** 9.51
45 ********** 10.74 ********** 9.89 ********** 9.55
46 ********** 10.67 ********** 9.98 ********** 9.59
47 ********** 10.58 ********** 10.06 ********** 9.64
48 ********** 10.46 ********** 10.15 ********** 9.69
49 ********** 10.30 ********** 10.21 ********** 9.76
50 ********** 10.09 ********** 10.26 ********** 9.83
51 ********** 9.83 ********** 10.32 ********** 9.91
52 ********** 9.49 ********** 10.37 ********** 9.99
53 ********** 9.09 ********** 10.40 ********** 10.08
54 ********** 8.60 ********** 10.44 ********** 10.16
55 ********** 8.02 ********** 10.48 ********** 10.24
56 ********** 7.36 ********** 10.51 ********** 10.30
57 ********** 6.60 ********** 10.53 ********** 10.33
58 ********** 5.78 ********** 10.55 ********** 10.31
59 ********** 4.90 ********** 10.58 ********** 10.19
60 ********** 3.98 ********** 10.59 ********** 9.96
61 ********** 3.06 ********** 10.61 ********** 9.56
62 ********** 2.19 ********** 10.62 ********** 8.97
63 ********** 1.41 ********** 10.63 ********** 8.13
64 ********** 0.78 ********** 10.64 ********** 7.05
65 ********** 0.33 ********** 10.63 ********** 5.75
66 ********** 0.09 ********** 10.60 ********** 4.26
67 ********** 0.04 ********** 10.58 ********** 2.65
68 ********** 0.14 ********** 10.52 ********** 0.99
69 ********** 0.35 ********** 10.45 ********** -0.59
70 ********** 0.62 ********** 10.33 ********** -1.91
71 ********** 0.93 ********** 10.18 ********** -2.79
72 ********** 1.25 ********** 9.97 ********** -3.13
73 ********** 1.57 ********** 9.71 ********** -3.02

As you can see, fractalization throws the gain curve to the higher channels, but not as much as the equivalent size 6 inch leg SBGH.

The benefits of fractalization are not as great for the UHF SBGH as they are for the VHF SBGH version. The overall element area is reduced from 24.32 by 30 inches to 21.70 by 25.35 inches. However, the element structure is stiffened with the additional bends.

Next Ill work on fractalizing the Gen2 UHF SBGH.