Area 51 for Antenna Modelers & Builders (See Post #1)

[300ohm]

Interesting, putting curves on the outside and inside of the mesh produces dips in the gain curve. Ive tried various patterns on the mesh, including one where I followed (in 2 X 4 block style) the outline of Autofils GH10 rod model. I havent found a pattern yet that does better than the ordinary rectangle mesh pattern.

[firimari]

Revisiting the reflector-less Single Diamond Koch Fractal antenna, I thought to myself

Hey, why can't I set the impedance on this thing to 75 Ohms and just attach the coax directly?

So I ran the optimizer first to minimize the SWR, then again to maximize the gain while keeping the SWR low, and came out with this gain curve:



Not at all shabby, keeping it above 7 between Channel 22 and 44

NEC File:

CM Experimental Antenna Design X-077
CM Parameterized Reflectorless Single Diamond Grey Hoverman 
CM Second Iteration Koch Curve
CM UHF Antenna 75 Ohm impedance
CM AWG Gauge 12 Copper Wire
CE
GW 2 3 0 0.2502373 0 0 0.2300478 0.0201895 1.0262e-3
GW 3 3 0 0.2300478 0.0201895 0 0.2374377 0.0477688 1.0262e-3
GW 4 3 0 0.2374377 0.0477688 0 0.2098584 0.0403789 1.0262e-3
GW 5 3 0 0.2098584 0.0403789 0 0.1896689 0.0605684 1.0262e-3
GW 6 3 0 0.1896689 0.0605684 0 0.1970588 0.0881477 1.0262e-3
GW 7 3 0 0.1970588 0.0881477 0 0.2246381 0.0955375 1.0262e-3
GW 8 3 0 0.2246381 0.0955375 0 0.2044486 0.115727 1.0262e-3
GW 9 3 0 0.2044486 0.115727 0 0.2118385 0.1433063 1.0262e-3
GW 10 3 0 0.2118385 0.1433063 0 0.1842592 0.1359164 1.0262e-3
GW 11 3 0 0.1842592 0.1359164 0 0.1640697 0.1561059 1.0262e-3
GW 12 3 0 0.1640697 0.1561059 0 0.1566799 0.1285266 1.0262e-3
GW 13 3 0 0.1566799 0.1285266 0 0.1291004 0.1211369 1.0262e-3
GW 14 3 0 0.1291004 0.1211369 0 0.1089109 0.1413264 1.0262e-3
GW 15 3 0 0.1089109 0.1413264 0 0.1163008 0.1689057 1.0262e-3
GW 16 3 0 0.1163008 0.1689057 0 0.0887215 0.1615158 1.0262e-3
GW 17 3 0 0.0887215 0.1615158 0 0.068532 0.1817053 1.0262e-3
GW 18 3 0 0.068532 0.1817053 0 0.0869267 0.2000999 1.0262e-3
GW 19 3 0 0.0869267 0.2000999 0 0.1120542 0.193367 1.0262e-3
GW 20 3 0 0.1120542 0.193367 0 0.1053214 0.2184947 1.0262e-3
GW 21 3 0 0.1053214 0.2184947 0 0.123716 0.2368893 1.0262e-3
GW 22 13 0 0.123716 0.2368893 0 0.251586 0.2368893 1.0262e-3
GX 22 011
GW 1 49 0 -0.250237 0 0 0.2502373 0 1.0262e-3
GE 0
GN -1
EK
EX 0 1 25 0 1 0
FR 0 39 0 0 470 6
RP 0 37 73 1001 -180 0 5 5
EN

SVG for one quadrant of the element (scaled 1mm=1mm so that I can use it as a bending template!)

<?xml version="1.0" standalone="no"?> 
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"  
  "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd"> 
<svg width="215.9mm" height="279.4mm" version="1.1" 
  xmlns="http://www.w3.org/2000/svg" 
  xmlns:xlink="http://www.w3.org/1999/xlink"> 
<defs> 
  <polyline id="Element" fill="none" stroke="black" stroke-width="2.0524e-03"
            points="0.2502373,0 0.2300478,0.0201895
0.2300478,0.0201895 0.2374377,0.0477688
0.2374377,0.0477688 0.2098584,0.0403789
0.2098584,0.0403789 0.1896689,0.0605684
0.1896689,0.0605684 0.1970588,0.0881477
0.1970588,0.0881477 0.2246381,0.0955375
0.2246381,0.0955375 0.2044486,0.115727
0.2044486,0.115727 0.2118385,0.1433063
0.2118385,0.1433063 0.1842592,0.1359164
0.1842592,0.1359164 0.1640697,0.1561059
0.1640697,0.1561059 0.1566799,0.1285266
0.1566799,0.1285266 0.1291004,0.1211369
0.1291004,0.1211369 0.1089109,0.1413264
0.1089109,0.1413264 0.1163008,0.1689057
0.1163008,0.1689057 0.0887215,0.1615158
0.0887215,0.1615158 0.068532,0.1817053
0.068532,0.1817053 0.0869267,0.2000999
0.0869267,0.2000999 0.1120542,0.193367
0.1120542,0.193367 0.1053214,0.2184947
0.1053214,0.2184947 0.123716,0.2368893
0.123716,0.2368893 0.251586,0.2368893" />
</defs>
<g transform="scale(3543.307)">
<g transform="translate(-0.05,0.015)">
<use xlink:href="#Element" />
</g></g>
</svg>

[300ohm]

I see you already got the stubs on the X077 shortened. I was going to try to put balls (“capacitive hats”) on the stubs, like the old rabbit ears trick that Ken Nist reminded me of here : http://www.hdtvprimer.com/ANTENNAS/R...s.html#Getting

Four 1 inch balls, 26 gauge, 30 inches off to the side, starting with wire 1000,
Geometry and Segment checked :
http://www.wuala.com/300ohm/Documents/4_1inchBalls.nec

Now, to figure out where to get hollow 1 inch aluminum and/or copper balls, heh. (I know I can crumple up aluminum and copper foil, but that looks so amateurish)

Revisiting the reflector-less Single Diamond Koch Fractal antenna, I thought to myself
Quote:
Hey, why can't I set the impedance on this thing to 75 Ohms and just attach the coax directly?

You still need a balun since youre going from a balanced antenna to an unbalanced coax line.

[firimari]

You mean I can't just solder my 75 ohm line in the middle like the hentenna?

I was actually considering building a 'symmetric log periodic balun' as soon as I can figure out what all the measurements need to be for a UHF one.

[300ohm]

A novel ultra-wideband balun with 3 to 6 GHz passband

That sounds good.

[RamKat]

Now, to figure out where to get hollow 1 inch aluminum and/or copper balls, heh. (I know I can crumple up aluminum and copper foil, but that looks so amateurish)

Just to see if it will work and to have a good sphere I suggest you get polistyrene (sp) balls that you can get a bag full off (different sizes) at the arts section of any $ store and wrap it in aluminum foil. Pre-drilling the hole and lining it with a foil sleeve (that you flare on the outside) before you do the wrapping will ensure a good electrical contact between the sphere and the rod. Cover the hole on the other end with a piece of hard plasting to prevent the rod from poking through the foil.

Keep us posted on your results

[300ohm]

arts section of any $ store and wrap it in aluminum foil.

Yeah, I thought about that too. On firimani's earlier versions of the small fractal GH, the stub lengths were keeping the antenna from being as compact as it could be. But on the above version, he already has that optimized for the shorter stubs.

[Walter Dnes]

After a bit of experimenting, I've managed to come up with an double-sized bowtie plus reflectors (24" wide x 9" high x 5" deep) that gets raw gain of 8 db across channels 14..69, with a SWR just below 3. If optimized for channels 14..51, it's 8.5 db raw. Is it worth posting it here, or is that rather weak for an antenna that size?

[RamKat]

Is it worth posting it here, or is that rather weak for an antenna that size?

Go for it Walter_Dnes. There is always someone looking for a wide band low profile antenna for closeby transmitters and that could be installed in small attics (or I guess on valuable real estate on an antenna mast) A smaller antenna also has in general a better WAF when mounted on the outside of the house (WAF = Wife Approval Factor)

I too have worked on a wide band double bay version of a bowtie. Will post my results soon.

[300ohm]

double-sized bowtie plus reflectors (24" wide x 9" high x 5" deep) that gets raw gain of 8 db across channels 14..69, with a SWR just below 3. If optimized for channels 14..51, it's 8.5 db raw.

Yeah post it here. I built a single bay version of that a long long time ago, (far, far away, NOT) for vhf, heh. Have you run the vhf-high charts on it ?

[stampeder]

Originally Posted by Walter Dnes

Is it worth posting it here, or is that rather weak for an antenna that size?

Absolutely its worth posting here. To paraphrase someone from somewhere sometime... more discoveries happen by asking "Why Not?" than "Why?"

[RamKat]

This is addicting - time goes way much too fast

I played with this Bowtie design and basically gave the optimization algorithm a free hand in adjusting every parameter. The only parameter on the rods that are the same are the gaps between the rods. Even the bow tie upper and lower parts were allowed to be a different lengths as well as different angles.

The bowtie lengths ended up 9.83 and 9.32 inches and the angles at 24.5 and 23.8 degrees - Close but different - It might have been just the optimization tollerance? I have also allowed the bowtie seperation and the feed separation to be different.

It was set to optimize over the whole UHF band Channel 14 to 88 in 24MHz steps.

It was set-up for 10 AWG for all elements - Here is what it looks like.


(Total size is about 33 inches Wide by 30 inches High)



Here is the response curve - still have to do the spread sheet thing to determine raw gain

[300ohm]

It was set to optimize over the whole UHF band Channel 14 to 88 in 24MHz steps.

Heh, the UHF band never went up to 88 even in the old days. But you are showing the incredible bandwidth the bowtie has.

By splitting the reflectors like that, youre giving up vhf-hi (174 - 216 mhz) gain on the bowtie. (with the high SWR and low raw gain, the net gain is well into the negative numbers) Try it with long (28 - 44 inch) non-split reflectors. The bowtie and GH are quite different in that respect.
Also, the top, bottom and middle reflectors are probably contributing very little to your overall gain.

[firimari]

Hey, you forgot the NEC file!

Yeah, I'm currently trying to add the balun into the simulation. It is, of course, messing everything else. It's certainly a case of 'Why Not?'

[Walter Dnes]

X-90 is a bowtie with reflectors. As noted in the comments, set variable SKALE to 1.11 to optimize for channels 14..69 and 1.20 to optimize for channels 14..51. With SKALE = 1.11, beware of the sharp dropoff just below channel 14.

Here is the code...

CM Bowtie with reflectors
CM Channel 14..69 version use SKALE=1.11
CM Channel 14..51 version use SKALE=1.20
CE
SY E_RAD=0.1
SY FEED=1
SY FB2=FEED/2
SY ANGL=22.5
SY COS_ANGL=COS(ANGL)
SY SIN_ANGL=SIN(ANGL)
SY BASE_E_LEN=8.5
SY BASE_R1_LEN=10.5
SY BASE_RH_SEP=5.0
SY BASE_RV_SEP=2.0
SY SKALE=1.11
SY E_LEN=BASE_E_LEN*SKALE
SY R1_LEN=BASE_R1_LEN*SKALE
SY RH_SEP=BASE_RH_SEP*SKALE
SY RV_SEP=BASE_RV_SEP*SKALE
SY ZZ1=E_LEN*SIN_ANGL
SY YY1=E_LEN*COS_ANGL+FB2
GW      8       15      -RH_SEP -R1_LEN RV_SEP  -RH_SEP -FB2    RV_SEP E_RAD
GW      9       15      -RH_SEP FB2     RV_SEP  -RH_SEP R1_LEN  RV_SEP E_RAD
GW      10      15      -RH_SEP -R1_LEN RV_SEP*2        -RH_SEP -FB2  RV_SEP*2 E_RAD
GW      11      15      -RH_SEP FB2     RV_SEP*2        -RH_SEP R1_LEN RV_SEP*2E_RAD
GX      4      001
GW      1       3       0       -FB2    0       0       FB2     0       E_RAD
GW      2       9       0       -FB2    0       0       -YY1    ZZ1     E_RAD
GW      3       9       0       -FB2    0       0       -YY1    -ZZ1    E_RAD
GW      4       9       0       FB2     0       0       YY1     ZZ1     E_RAD
GW      5       9       0       FB2     0       0       YY1     -ZZ1    E_RAD
GW      6       15      -RH_SEP -R1_LEN 0       -RH_SEP -FB2    0       E_RAD
GW      7       15      -RH_SEP FB2     0       -RH_SEP R1_LEN  0       E_RAD
GS      0       0       0.0254          ' All in in.
GE      0
EK
EX      0       1       2       0       1       0
GN      -1
FR      0       1       0       0       470     6
RP      0       1       37      1510    90.     0.      0.      10.     0.     0.

And the gain graph



And the SWR graph