Solid Bow-Ties?

[k6sti]

Today I was reading the chapter on TV receiving antennas by Edward B. Joy in Johnson & Jasik's Antenna Engineering Handbook. In the section on triangular-dipole antennas, I found this:

A dipole formed from two triangular sheets of metal is sufficiently broadband with respect to gain and VSWR (with respect to 300 ohms) for all-UHF-channel reception. . . . The triangular antenna has many of the same pattern and input-impedance characteristics of the biconical dipole discussed in Chap. 4, but it is lighter in weight and simpler to construct. Further simplification of the triangular dipole to a wire outline of the two triangles results in significant degradation of the broadband performance. However, the metal triangles can be approximated with wire mesh, provided the mesh spacing is less than one-tenth wavelength. [Italics added]

As an example, the book shows a sheet-metal bow-tie with cutout holes 15" wide overall with a flare angle of 70 degrees at the vertex of each triangle. A graph shows measured gain for two of them stacked 23" apart, each in front of its own 22" x 19" screen, as 10.5 dBi at 470 MHz quickly rising to between 13 and 14 dBi from 530 to 700 MHz (and higher beyond).

It seems like the bow-tie designs discussed in this forum use just two wires. When I modeled one, I was surprised to find no impedance broadbanding over an ordinary dipole. I wonder if anyone has looked into using sheet-metal or mesh bow-ties.

Brian

[stampeder]

A few years ago there was this exchange between two OTA Forum members that I copied and pasted into the OTA FAQ regarding bowties. Your post reminded me of it:

Originally Posted by Zozo

Why are the wires on the CM 4228 and CM 4221 at angles to the horizontal? I have been reading up on the fundamentals of antenna theory and one of the things that I learned is that the polarization of the transmitter and antenna must be the same in order to get the best signal reception. I have read that TV broadcasts use horizontal polarization.

My question now is why are the receiving wires (bow ties?) on these antennas in the V shape with angles away from the horizontal. Would it not be better if all the wires were horizontal - parallel to the ground?

Originally Posted by Rop

Hi Zozo,

You can't just look at a single wire, but have to take the resulting effect of the entire bowtie into account (in fact, all those bowties and the reflector affect each other greatly as far as electromagnetic field goes). The two-wire bowtie is actually a crude approximation of a cone-shaped solid conductor, which turns out to work over a wide frequency range compared to just a single wire. Engineers found out that by making them with just two wires instead was almost as effective, and much cheaper, less wind sensitive etc. In any event, the nett effect is that it's horizontally polarized, and has a wide bandwidth. Just what you want for a TV antenna.

-Rob-

Rop mentioned conic bowties, but in your post the triangular bowtie is a different animal, right?

[k6sti]

Originally Posted by stampeder

Rop mentioned conic bowties, but in your post the triangular bowtie is a different animal, right?

Yes, that's right. A biconical antenna is a broadband antenna analyzed in textbooks. It expands as a cone in both directions (infinitely far, for some analysis). Truncated biconicals with larger cone angles are pretty broadband and a flat triangular approximation to them is what we call a bowtie or bow-tie (I think I like the one with no hyphen better!). Last week I modeled the two-wire approximation (four wires, both sides), which I think is what people call whiskers. I was surprised that SWR was no lower over the UHF-TV band than for an ordinary dipole. I thought broadband SWR was the whole point of a bowtie antenna. I did notice that gain increased somewhat higher in the band. I attributed that to the bowtie effectively stacking two dipoles at a short distance, with the high current portions being separated more at high frequencies where the bowtie acts more like a fullwave dipole than a halfwave dipole. I just figured this slight gain advantage (I think it was half a dB or so) was why people were using the whiskers. Now I'm not so sure. The Jasik book is quite definite about the loss of bandwidth unless a good approximation to a solid bowtie is used. Someone here must have been through this before.

Brian

[mclapp]

The impedance on a bowtie is generally higher than a dipole which can be more attractive when stacking mulitple bowties since you end up with a more reasonabe impedance.

The higher impedance usually lends itself to a wider band width since it takes quite a large swing in feed point impedance to turn into a poor SWR.

The whisker antennas I have computer modeled have shown a wide SWR curve generally swinging from a lower impedance on frequencies below resonance and higher above resonance.

Over that range they go from (2) 1/2 wave dipoles in phase to a 5/8 wave extended double Zepp style antenna. With a single bowtie you should see closer to 3db over a single 1/2 wave dipole when it gets into the 5/8 wave extended zepp range.

I'm not sure, were you working with half wave dipoles and half wave whiskers or full wave?

[k6sti]

Originally Posted by mclapp

I'm not sure, were you working with half wave dipoles and half wave whiskers or full wave?

Various lengths, but their impedance bandwidth never seemed any better than that of a dipole with a single conductor.

You can set the dipole length to maximize gain without regard to SWR and then match it later. That's not a bad approach. You can do that and then try to optimize the 2-D geometry to lower SWR as well, although I'm not sure solid surfaces can be adequately modeled without using a lot of wires.

I just now made a model using wires that outlined each triangle and had three wires within the interior from the vertex to the far edge. Its SWR variation is less than 2 or 2.5 from 470-700. I don't know how well the wire model mimics a solid surface, but this model does seem to do something. No dipole of any length (or thickness--I tried that first) varies that little.

I'm just wondering if anyone has measured solid bowties and compared their SWR from 470-700 MHz with the two-whisker dipoles. In other words, is there anything to gain? The Jasik book definitely said there was, but I'd like to see some data.

Brian

[300ohm]

I just figured this slight gain advantage (I think it was half a dB or so) was why people were using the whiskers. Now I'm not so sure. The Jasik book is quite definite about the loss of bandwidth unless a good approximation to a solid bowtie is used.

I dont. I use the closed end bowtie in my bowtie modeling and builds. The main reason being it makes the bowtie more structually sound against the occasional bird landing on it.

Bandwidth is not really an issue with the uhf bowtie. It has oodles, unlike the Hoverman. And uhf bandwidth has been reduced to channels 14 to 51 from 14 to 83, so thats not an issue anymore with the GHs either.

[k6sti]

Originally Posted by 300ohm

I use the closed end bowtie in my bowtie modeling and builds.

What dimensions do you use?

Brian

[300ohm]

What dimensions do you use?

The length of each individual bowtie is 8.67 inches and spread is 4 inches at the wide end. Its indented on the wide end 1 inch inward at the middle. And the bowties are spread forward about 20 degrees.


The NEC file of the driven elements. I used a 40 X 40 inch swept 2 X 4 mesh reflector on it with 1/2 inch mesh in the middle. So with all the wires, the NEC file is very large. If you want, I can post the whole thing.

CM 300 ohms Baltimore Antenna
CE
GW	1	1	25.85	-0.15	0	25.85	0.15	0	0.01794527
GW	2	11	25.85	-0.15	0	15.85	-0.15	0	0.012
GW	3	11	25.85	0.15	0	15.85	0.15	0	0.012
GW	4	1	15.85	-0.15	0	15.25	-0.3	0	0.025
GW	5	1	15.85	0.15	0	15.25	0.3	0	0.025
GW	6	1	14.35	-0.625	0	15.25	-0.3	0	0.05
GW	7	1	14.35	0.625	0	15.25	0.3	0	0.05
GW	8	5	14.35	-0.625	0	14.35	-0.625	4.5	0.04040404
GW	9	5	14.35	0.625	0	14.35	0.625	4.5	0.04040404
GW	10	5	14.35	-0.625	0	14.35	-0.625	-4.5	0.04040404
GW	11	5	14.35	0.625	0	14.35	0.625	-4.5	0.04040404
GW	12	7	14.35	-0.625	-4.5	14.35	-0.625	-9.5	0.04040404
GW	13	7	14.35	0.625	-4.5	14.35	0.625	-9.5	0.04040404
GW	14	7	14.35	-0.625	4.5	14.35	-0.625	9.5	0.04040404
GW	15	7	14.35	0.625	4.5	14.35	0.625	9.5	0.04040404
GW	16	3	14.35	-0.625	-11.25	14.35	-0.625	-13.5	0.04040404
GW	17	3	14.35	0.625	-11.25	14.35	0.625	-13.5	0.04040404
GW	18	3	14.35	0.625	11.25	14.35	0.625	13.5	0.04040404
GW	19	3	14.35	-0.625	11.25	14.35	-0.625	13.5	0.04040404
GW	20	3	14.35	0.625	11.25	13.95	-3.398e-15	10.5	0.04040404
GW	21	3	14.35	-0.625	9.5	13.95	-3.398e-15	10.5	0.04040404
GW	22	3	14.35	-0.625	11.25	14.75	-3.398e-15	10.5	0.04040404
GW	23	3	14.35	0.625	9.5	14.75	-3.398e-15	10.5	0.04040404
GW	24	3	14.35	-0.625	-9.5	14.75	-3.398e-15	-10.5	0.04040404
GW	25	3	14.35	0.625	-11.25	14.75	-3.398e-15	-10.5	0.04040404
GW	26	3	14.35	0.625	-9.5	13.95	-3.398e-15	-10.5	0.04040404
GW	27	3	14.35	-0.625	-11.25	13.95	-3.398e-15	-10.5	0.04040404
GW	28	14	14.35	0.625	13.5	15.35	9	11.5	0.05094856
GW	29	3	15.35	9	11.5	15.35	8	13.5	0.05094856
GW	30	14	14.35	0.625	4.5	15.35	9	6.5	0.05094856
GW	31	14	14.35	0.625	4.5	15.35	9	2.5	0.05094856
GW	32	3	15.35	9	6.5	15.35	8	4.5	0.05094856
GW	33	3	15.35	9	2.5	15.35	8	4.5	0.05094856
GW	34	14	14.35	0.625	-4.5	15.35	9	-2.5	0.05094856
GW	35	14	14.35	0.625	-4.5	15.35	9	-6.5	0.05094856
GW	36	3	15.35	9	-2.5	15.35	8	-4.5	0.05094856
GW	37	3	15.35	9	-6.5	15.35	8	-4.5	0.05094856
GW	38	14	14.35	0.625	-13.5	15.35	9	-11.5	0.05094856
GW	39	14	14.35	0.625	-13.5	15.35	9	-15.5	0.05094856
GW	40	3	15.35	9	-11.5	15.35	8	-13.5	0.05094856
GW	41	3	15.35	9	-15.5	15.35	8	-13.5	0.05094856
GW	42	14	14.35	-0.625	13.5	15.35	-9	15.5	0.05094856
GW	43	14	14.35	-0.625	13.5	15.35	-9	11.5	0.05094856
GW	44	3	15.35	-9	11.5	15.35	-8	13.5	0.05094856
GW	45	3	15.35	-9	15.5	15.35	-8	13.5	0.05094856
GW	46	14	14.35	-0.625	4.5	15.35	-9	6.5	0.05094856
GW	47	14	14.35	-0.625	4.5	15.35	-9	2.5	0.05094856
GW	48	3	15.35	-9	2.5	15.35	-8	4.5	0.05094856
GW	49	3	15.35	-9	6.5	15.35	-8	4.5	0.05094856
GW	50	14	14.35	-0.625	-13.5	15.35	-9	-11.5	0.05094856
GW	51	14	14.35	-0.625	-13.5	15.35	-9	-15.5	0.05094856
GW	52	3	15.35	-9	-15.5	15.35	-8	-13.5	0.05094856
GW	53	3	15.35	-9	-11.5	15.35	-8	-13.5	0.05094856
GW	54	14	14.35	-0.625	-4.5	15.35	-9	-2.5	0.05094856
GW	55	14	14.35	-0.625	-4.5	15.35	-9	-6.5	0.05094856
GW	56	3	15.35	-9	-2.5	15.35	-8	-4.5	0.05094856
GW	57	3	15.35	-8	-4.5	15.35	-9	-6.5	0.05094856
GW	5576	11	14.35	0.625	13.5	15.35	9	15.5	0.01598073
GW	5577	3	15.35	9	15.5	15.35	8	13.5	0.01598073
GS	0	0	0.0254		' All in in.
GE	0
EK
LD	5	0	0	0	1.66667e7	0
EX	0	1	1	0	1	0
GN	-1
FR	0	1	0	0	614	0
RP 0 1 10 1510 90. 0. 0. 20. 0. 0.

[k6sti]

Originally Posted by 300ohm

The length of each individual bowtie is 8.67 inches and spread is 4 inches at the wide end. Its indented on the wide end 1 inch inward at the middle.


The NEC file of the driven elements. I used a 40 X 40 inch swept 2 X 4 mesh reflector on it with 1/2 inch mesh in the middle. So with all the wires, the NEC file is very large. If you want, I can post the whole thing.

Thanks, but I don't use NEC files. If you could just give a verbal description.

Do you use two wires, each side, that fan out? Are these two wires joined by a single wire with the feedpoint at its center? Is 8.67" the total length of the element or its half-length? Is the 4" spread the distance from top to bottom or top/bottom to centerline? The indentation I think I understand (!). What is its purpose?

Brian

[300ohm]

Yes, the description above is for just one of the eight wire triangles you see above. 8.67" is the length of each side of the triangle. So the perimeter of one triangle is 8.67 plus 8.67 plus 4 inches equals about 21.34 inches.

Is the 4" spread the distance from top to bottom or top/bottom to centerline? The indentation I think I understand (!). What is its purpose?

The 4" spread is top to bottom. 2" to the center. By indenting, it helps make it more closely resemble a cone, and helps with the higher channels. An important thing to note is that I specifically tweaked it for the channel range of 38 to 46, my Baltimore stations. (plus 11 and 13 vhf, which the reflector is responsible for)

The phasing lines are the ones in mclapps pdf.

The triangles are soldered at the tips, after wrapping around a nail to make a hole for the bolt to go thru. Then trimmed and filed.

[k6sti]

Thanks for all the info. As usual, impatience got the better of me and I ran a bunch of bowtie models before seeing your response. You'll find the results in another thread.

Sorry, I missed seeing the image of the antenna the first time. It answered most of my questions.

I really like the idea of soldering the connections. Recently I had reason to believe that one of my FM Yagis wasn't performing quite as it should. I got out my ohmmeter and measured an open circuit at the end of the twin-lead feedline. The antenna uses folded dipole driven elements so I should have seen something. When I pulled the antenna off the roof, the twin-lead was securely held by the wing nuts at the feedpoint. Everything was very tight. But when I unscrewed them, I found all kinds of corrosion. I haven't had this trouble before and it was hard for me to believe that all the pressure at the joint didn't result in a decent connection. I cleaned it all up, put it back on the roof, and it has worked fine ever since.

Brian

[300ohm]

When I pulled the antenna off the roof, the twin-lead was securely held by the wing nuts at the feedpoint. Everything was very tight. But when I unscrewed them, I found all kinds of corrosion.

Galvanic corrosion from mismatched metals will do it every time, heh. Consult a galvanic corrosion chart.