One of these is an antenna for 160 meters. So I sat down to review my options and design something that will get me through the winter season. I decided to shelve my original 160 meter antenna plan as too problematic and time consuming to be undertaken this year, if ever.
A horizontally polarized antenna for 160 meters is a poor choice even with a 43 meter tower; that's still only ¼λ, far too low for effective DX communication. So it has to be a vertical. I could shunt feed the big tower, except that for SO2R and multi-op contesting use of the tower in this fashion pretty well precludes use of any of the antennas for other bands mounted on the tower. Besides which, with all the yagis and mast on it the electrical length would be longer and therefore less predictable. Radials have to be rolled up in the spring before the hay grows and the ticks return.
The trees near enough to be useful as vertical supports are under 25 meters height and near to other planned antennas or the power line. In any case I don't have a good way to send a rope and wire that high and through dense foliage. It's a solvable problem but not expedient.
Since my long term plan for 160 meters remains unsettled I am happy to pursue a temporary solution to get me through this coming winter and contest season. Doing nothing isn't an option since I am missing too many QSOs and multipliers with makeshift approaches such as loading my 80 meter inverted vee last contest season. I worked some DX (over 30 countries) with great difficulty, and almost all were contest super-stations.
I settled on a what is a fairly popular antenna: a wire vertical top loaded with a 2-wire capacity hat. It is effective, can hang off the big tower with long rope and is efficient with a modest number of radials. The design is also popular in vertical yagi arrays since multiple elements for multiple directions can be hung around a big tower. The tower serves as the driven element.
Basic design and potential problems
After some geometry and experimentation with EZNEC I settled on a design where the vertical is 21.3 meters tall and each half of the top hat 13 meters long. It resonates at 1.830 MHz, near the centre of the band segment for CW DXing and contesting. The 'T' is at an angle of 45° so that it can be placed 20 meters from the tower and the catenary rope tied off near the top of the tower. Impedance at resonance is 27 Ω.
The choice of distance from the tower allows enough space in the hay field for 30 meter long radials without the need for an extremely long transmission line. The area is to the southwest of the tower, approaching the stone wall surrounding the yard area, and overlapping the south guy anchor. The catenary rope will be tied off to a tree or a stake in the ground, which will be determined during final layout of the antenna.
|Transparent 60 m diameter circle roughly marks the extent of the antenna's radial system|
The antenna is easy to model, with a few cautions. The tower is grounded for lightning safety with a 10' long ground rod (9' in the ground). This cannot be directly modelled in NEC2 so I used the MININEC ground option in EZNEC. The impedance of the ground connection must be estimated and simulated with a resistance load at the bottom end of the wire representing the tower. The tower itself is simply modelled as a 30 cm diameter wire, which is consistent for a lattice triangular structure with 20" (50 cm) faces.
I tested the model with both grounds before adding the tower to identify any modelling anomalies. MININEC ground is required in the full model since the tower is grounded. The antenna is raised slightly above ground (20 cm) so that it and the radials can be modelled in NEC2. The 27 Ω resonance impedance mentioned earlier is from the real, high-accuracy ground option in EZNEC. Switching to MININEC ground the reported impedance drops 10% and the resonant frequency moves downward an insignificant 2 kHz (-0.1%). The reported gain also shows a small change due to ground loss differences in the two ground models. MININEC reports 1.0 db gain versus 0.4 db for real, high-accuracy medium ground.
Based on this comparison the antenna is remarkably similar with both EZNEC high-accuracy ground and MININEC ground. This provides some assurance that the full model that uses MININEC ground is useful and reasonably trustworthy.
The antenna and capacity hat are 14 AWG TTHN house wire. There are 16 30-meter long radials made of 24 AWG insulated copper wire that are tentatively planned to be the individual strands taken from stripped bulk data cable. With this many radials the system begins to show the characteristics of a non-resonant ground plane, so the chosen length can work well. More radials are of course better though not enough to justify the trouble for a temporary antenna.
The major difficulty using 24 AWG wire for radials is their fragility. I²R losses are negligible at 1,000 watts since the current divides among the radials. It is quite low in each, low enough to substantially minimize loss. It works out to a small fractions of a decibel.
There is no real need for the radials to be 40 meters long since the system is non-resonant, and in any case they are electrically longer due to the adjacency to ground which lowers the velocity factor (VF). A good place to learn more about this at N6LF's site where he has extensively documented his many careful experiments on low band verticals and radial systems.
Now, about that tower...
The antenna itself is unremarkable. Many hams use this or a similar design on 160 meters. There are enough of them out there to assure that it can work quite well. My model of the antenna confirms that and it appears reliable based my explorations with EZNEC.
Now we come to what is perhaps my greatest concern: the impact of the 150' (really ~143' or 43 meter) tower on antenna performance. Notice in the EZNEC model view above that there is significant current on the tower since, at 43 meters and grounded, it is resonant at or near the 160 meter band. Worse, the distance between the tower and vertical is 20 meter (⅛λ), an ideal separation for a 2-element yagi!
With only a single antenna for 160 meters I am concerned that if the tower acts too much like a parasitic element that the pattern will discriminate against directions that are useful for DX and contest contacts. When loaded with mast and antennas it will be electrically longer and may act as a reflector which will inhibit my ability to work Europe (northeast). That would be bad. Some hams use this effect to create a directional set of verticals surrounding a tall tower. Since I will have only one an omni-directional pattern is preferred.
Of course there would also be gain towards the southwest, covering the bulk of the US. This is not a great trade off since the gain is small. F/B is the greater concern. That's what yagis do.
Exploring the landscape
To investigate the range of possible outcomes I ran the full model while varying the height of the tower. This is not exactly equivalent to a 43 meter tower with a capacity hat (top mounted yagis) however as long as the effective (electrical) length is not too much more than the physical height the modelled results will be more than close enough to gain an understanding of the tower's impact. I need insight more than I need high accuracy.
I varied the height from 35 meters to 55 meters in steps of 2 meters. At each step I used the model to measure the gain, F/B, feed point impedance and resonant frequency (X = 0). The plot is quite interesting. Tower grounding (ground rod) is simulated by a 20 Ω load in series with the ground connection.
The first thing to notice is that at a shorter height the tower acts as a director element. Since the pattern is with respect to the southwest I show that reversal as a negative F/B, while keeping the gain relative to the main lobe. At 41 meters height the tower's action as a reflector is accentuated.
The true maximum F/B is very sensitive to tower height and my chosen steps of 2 meters is too coarse to show how high the F/B can go, which is in fact more than 12 db. On 160 meter a yagi has a very narrow effective bandwidth.
|Azimuth plot for an electrical tower height ~43 meters;|
almost, but not worst case for high F/B
Of most interest to me is that the gain and F/B only gradually return to normal as the electrical height of the tower increases. With a mast projecting 3.5 meters above the tower and holding two large yagis this is what I can expect to experience. That is, a loss of 2 db towards Europe and a similar gain towards the US midwest. Anecdotal data from hams with a similar antenna is consistent with the model. There is little effect in other directions.
Impedance at the greater (electrical) heights is a respectable 21 Ω, a drop of only 10%. The antenna will need an L-network to lower the SWR from slightly over 2 at resonance. This is easy to build. Most likely I'll reuse the box for the 80 meter feed of my previous tower back in Ottawa. I'll do the design once the antenna is built and I measure the actual R and X values. The antenna's SWR bandwidth is good and will exhibit a good SWR across 1.8 to 1.9 MHz with a simple L-network at the feed point, provided the tower resonance is outside its yagi zone. L-network tuning is expected to change as yagis are added to the tower since that will change the mutual impedance with the vertical.
On to construction
This antenna project will go ahead later this month. It'll be done after the prop pitch rotator shelf and mast are raised to avoid the catenary interfering with the lift. I am also waiting to see whether my neighbour will bother to take a second hay crop off the field. There may not be enough there for him to take the trouble, so I am likely free to proceed when ready.
I am looking forward to operating 160 meters this winter. Never before have I had a good antenna for that band so it'll be a new experience. Most of my 160 meter operation in the past has been contests from other stations. I expect to have some fun bumping up my DX totals and contest scores.
Once the tower project is complete you can expect a return to more articles like this one, about antennas and operating. Building support structures is important but is not my primary interest.