Saturday, June 18, 2011

The Optimized Wideband Antenna (OWA) Yagi

The Optimized Wideband Antenna (OWA) Yagi


Introduction

A remarkable method for increasing the VSWR bandwidth of Yagi antennas as well as allowing a direct dipole feed was recently discovered by Jim Breakall, WA3FET. Over the past year, Nathan Miller, NW3Z, has designed a series of contest grade monoband antennas using this method. These antennas are currently being built by NW3Z at the K3CR contest station and Tim Duffy, K3LR, is planning to replace many of his stack antennas with these designs.

The taper schedule for each of the antennas discussed is shown at the end of this article.

The Optimized Wideband Antenna

The Optimized Wideband Antenna (OWA) is antenna feed-point matching method developed by WA3FET that provides an increased feed point impedance while widening the VSWR bandwidth. This method can be used to improve several common amateur antennas: the dipole, monopole and the yagi. We will briefly discuss applications of the OWA for dipoles and monopoles before exploring the applications to Yagi antennas.

Dipole and Monopole Application

The OWA is implemented in a dipole or monopole antenna by placing a parasitic element very close (£ .01l ) to the fed element. By placing the element very close, the antenna approximates a wire with a diameter equal to the spacing between the driven element and the OWA element. This large apparent diameter results in a correspondingly larger bandwidth as well as an increased radiation resistance. In a monopole antenna, the increased radiation resistance is extremely useful in matching the antenna to a feedline. For the dipole (which has a free space impedance of about 72 ohms) the increased impedance is a problem that must be overcome by using some type of transformer at the feed point although the increased bandwidth dramatic in both cases. In the case of an 80m OWA dipole, coverage can be extended to the entire band. Because the OWA element is very close to the fed element the pattern has only a slight distortion from the single wire case.

Application to Yagi Antennas

The OWA is implemented in a yagi antenna by placing the first director very close (£ .05l ) in front of the driven element. Because the feed point is raised to 50 ohms, a simple 1:1 Current Balun is all that is needed to connect a coaxial feedline; replacing the more complex gamma or beta matching systems used to match the low impedance of most yagi antennas.

The only negative aspect of OWA yagis is that they give up a slight amount of forward gain (typically less than .5dB) from a conventional yagi. This small degradation is impossible to notice in operation and is well worth the simplicity of direct feed and the excellent bandwidth.

NW3Z has recently designed a series of 10, 15 and 20m monoband yagis for use at both the Penn State station K3CR and the K3LR Super-Station. These designs were created using the most recent Numerical Electromagnetics Code (NEC4) and NEC-OPT optimization software.

Construction of the Elements

The elements are made of standard .058" wall 6061-T6 aluminum tubing which is available from a large number of sources throughout the country. The taper schedules, which have around 90mph wind survivability, are variants of those shown in the ARRL Antenna Handbook. Where the tubing overlaps, the joint is coated with anti-oxidant and secured with 4 aluminum pop rivets set 90 degrees apart.

Mounting the Elements on the Boom

The parasitic elements are mounted to the boom using four U-Bolts and a plate is made of 4" x .375" thick aluminum bar stock. The plate is 8" long for the 20m elements and 6" long for the 15m and 10m elements. Because the OWA yagi uses a direct dipole feed, the driven element must be split in the center as well as insulated from the boom. To insulate the element, the mounting plate is made of 3/8" thick Garolite, a phenolic material that is available in sheets from many suppliers. A gap of 2" is left in the center of the driven element and a 24" section of fiberglass rod of appropriate diameter is inserted inside the tubing to provide strength and to prevent crushing. Two U-Bolts are used on each half of the driven element for additional support. It is important to note that when measuring the driven elements, all measurements are made from the center of the gap, not the beginning of the tubing. For example, if the taper schedule calls for 48" of 1.000" tubing for the inside part of the element half, the tubing is actually cut to 47" to leave the gap in the center.

Boom Supports

All of the antennas at K3LR and K3CR use 48� booms, which must be vertically guyed to eliminate most of the sag. The 20m antennas use a boom of 3"(.120" wall) aluminum while the 10 and 15m antennas use a 2.5" (.120" wall) boom. Stainless Steel Eye-Bolts are put through the boom and PhillystranÔ , a PVC protected kevlar rope, is used for the support cables. Because Phillystran is non-conductive, interaction is avoided.

20m - 6 Element Antenna on a 48� Boom

Free Space Optimized

The first antenna design was optimized in free space and has the following remarkable characteristics across the band: VSWR < 1.2, F/B > 23dB and Gain > 10dBi.

20m - 6 Element Antenna on a 48� Boom

6/6/6 Stack Optimized

While the previous antenna is excellent when used by itself, problems develop when it is stacked vertically. Unfortunately, when this antenna is stacked on a 175� tower such as at K3LR or K3CR, the pattern is marred by a tremendous back lobe. The antenna was re-optimized in the stack in order regain the excellent F/B but the VSWR had to be sacrificed in order to meet this goal although it is still under 1.6 across the band. At K3CR, the antennas are stacked at 60�, 120� and 175� giving a takeoff angle of 7 degrees.

When the antennas are stacked, the VSWR is a maximum of 1.6 at the bottom of the band and nulls near the top of the phone band. The 6/6/6 stack pattern is shown in the following graph.

15m - 6 Element Antenna on a 48� Boom

The original plan was to use 7 elements on a 48� boom but after several models were run, it was found that there was no advantage to the 7th element. This antenna definitely looks like its missing an element but�. For 15m we were fortunate in that the pattern of the optimized free space design is not degraded at all when placed on the 185� tower at K3CR. This design has the following free space performance across the band: VSWR < 1.3, F/B > 23dB and Gain > 11.2dBi. When this antenna is modeled on the 120� tower at K3LR there is some loss of F/B at the top of the band but the overall performance is excellent. The free-space performance of this antenna is shown below.

10m - 7 Element Antenna on a 48� Boom

Once again the optimum free space design works on both the 175� tower at K3CR and the 100� tower at K3LR. The antenna has the following free-space characteristics.

Application to 40m

The OWA has also been applied successfully to the (2) 40m antennas at K3LR. This remarkable antenna was designed by WA3FET and K3LR and has been proven in several contests. These 4 element antennas are on 48� booms and are stacked on a 190� tower. The OWA allows continuous coverage of both the phone and CW portions of the band. More details of these antennas will be published in the future.

TAPER SCHEDULES

The spacing of each element is shown from the reflector and all lengths represent the exposed length of tubing. Double wall sections are indicated below the table. The taper schedule of the elements can be adjusted using YO (using the W6QHS algorithm). All element lengths assume the use of the plate clamps discussed previously.

The 20m Antenna Optimized for Free-Space or Stand Alone Operation

1

0.875

0.75

0.625

0.5

0.00

48

24

44

36

65.73

90.00

48

24

44

36

58.70

139.52

48

24

44

36

48.80

226.70

48

24

44

36

42.62

388.44

48

24

44

36

42.63

570.00

48

24

44

36

35.39

* The 1.000", 0.875" and 0.750" sections are double wall sections

The 20m Antenna Optimized for Stacking

1

0.875

0.75

0.625

0.5

0

48

24

44

36

66.7765

90.05

48

24

44

36

59.3972

137.74

48

24

44

36

51.573

238.04

48

24

44

36

43.9092

348.84

48

24

44

36

43.5015

570

48

24

44

36

38.9487

* The 1.000", 0.875" and 0.750" sections are double wall sections

The 15m Design

0.75

0.625

0.5

0

36

24

83.5746

82.4547

36

24

79.2662

118.239

36

24

74.0373

221.131

36

24

70.3034

408.314

36

24

66.9878

570

36

24

63.5933

*the .75" section is double wall

The 10m Design

0.75

0.625

0.5

0

24

18

64.81

63.98

24

18

62.13

86.97

24

18

57.49

157.73

24

18

54.84

291.46

24

18

51.47

439.21

24

18

51.07

570

24

18

47.28

*The .75" Section is a double wall

Conclusions

The 6/6/6 20m antennas have been installed at K3CR and the switching system should be completed by the end of the year; we can only select individual antennas at this point. We have done extensive on-the-air testing and the design seems to be performing as expected. Due to cost constraints, the middle antenna is fixed on Europe while the top antenna is rotated with a Tailtwister and the bottom is turned with a TIC Ring. A single 15m antenna has been finished and installed on a 55� crankup tower for testing. This antenna is absolutely phenomenal in both pattern and performance. The other (2) 15m antennas as well as (4) 10m antennas will be constructed throughout the winter to be placed on the tower next spring.

I will update this page as construction and testing progresses. If you want to build any of these antennas feel free to email me with any questions and please let me know how they work for you. I also have several other designs that you might be interested in but did not have the time to incorporate into this article. Keep your eyes out for the Feb or March �98 issue of QST; we will have an article in there on an interesting 3 element 40m antenna that could be quite useful for smaller stations.

GO FRC!

Nathan A. Miller NW3Z

nam109@psu.edu

PS. Why is it important to check the antenna performance in the stack?

As a contrast to the K3CR 15m system (6/6/6 on a 185� tower), a stack of (6) four element monobanders (Gain = 8.3dBi and F/B = 30dB for a single antenna in free space) from the ARRL antenna book was modeled on a 200� tower at 21.200 MHz. The K3CR stack has slightly more gain but its big advantage is in the pattern. The monster stack has very poor F/B; the main rear lobe is only 15dB down from the front lobe. The many advantages of the three antenna system are obvious; the reduced complexity of erection, maintenance, the ability to rotate all antennas, etc.

No comments: