A bowtie antenna is a type of antenna that reputedly provides higher gain at lower radiation angles than a center-fed dipole antenna at heights considerably less than 1/2 wavelength above ground. Bowtie antennas have been around since the “spark gap” days of radio. I was curious about HF bowties so I decided to learn more about this type of antenna.
Phase I – Antenna Modeling
I selected a simple horizontal form of the bowtie antenna for modeling purposes. Figure-1 shows the target bowtie configuration.
Figure-1. 40 meter bowtie antenna configuration
How long should the radiators (L) be and how far apart (W) should they be at the ends? I started with a commonly used formula for computing the radiator lengths of a 1/2 wavelength dipole (468 / freq. MHz).
L = 468 / 7.15 MHz = 65.45’; 65.45 / 2 = ~32.72’ = 33’ (rounded up)
For W, I decided to start with 20% of the length of L. Why 20%? I had to start somewhere and I knew that I could easily adjust the value of W in the model.
W = 33’ * .20 = ~6.6′ = 7’ (rounded up)
The antenna was specified to be in a horizontal plane 25’ feet above real ground. The radiators were defined as #14 AWG stranded THHN insulated wire. The antenna was configured to be center-fed using Wireman 554 ladder line with a 4:1 balun at the source for impedance transformation purposes.
I plugged in the starting values of L & W and adjusted the model to get the best predicted SWR curve. After some experimentation, I found that a value of 6’ for W and radiator lengths (L) of ~31.25’ produced the best predicted SWR curve.
The EZNEC models used in this project can be downloaded from the links below.
center-fed dipole model bowtie model
Note: To see the predicted effects of insulated wire radiators take a look at the EZNEC models below.
center-fed dipole – insulated radiators bowtie – insulated radiators
Predicted radiation Patterns
Figure-2 shows the EZNEC predicted elevation radiation pattern for a 40 meter center-fed dipole at a height of 25’.
Figure-2. 40 meter center-fed dipole predicted elevation radiation pattern
Figure-3 shows the EZNEC predicted elevation radiation pattern for the target 40 meter bowtie antenna at a height of 25’.
Figure-3. 40 meter bowtie predicted elevation radiation pattern
Predicted SWR Curves
Figure-4 shows the EZNEC 40 meter center-fed dipole predicted SWR curve.
Figure-4. 40 meter center-fed dipole predicted SWR curve
Figure-5 shows the EZNEC predicted SWR curve for the target 40 meter bowtie antenna.
Figure-5. 40 meter bowtie antenna predicted SWR curve
Before you get concerned about the predicted bowtie antenna SWR curve, remember that balanced line will be used for the transmission line. High quality balanced line provides low transmission loss even in the presence of relatively high SWR values. What we really care about is antenna efficiency which is a measure of the effective power delivered to the antenna feed point. We will calculate the efficiency of the completed antenna later in the article when the antenna analyzer measurements are available.
Let’s take a quick look at the proposed 40 meter bowtie antenna predicted efficiency. We can use a free online transmission line loss calculator and the EZNEC SWR predictions to estimate efficiency. The online loss calculator is generally accurate to within ±5%. The Wireman 554 ladder line (50′) was selected for the transmission line. Figure 5a shows the predicted antenna efficiency at three frequencies across the 40 meter band.
Figure-5a. 40 meter bowtie antenna predicted efficiency
Comparison of Center-fed Dipole and Bowtie EZNEC Models
As far as predicted radiation patterns go, the center-fed dipole is a “cloud warmer” with negative gain at radiation angles less than 45 degrees. On the other hand, the bowtie exhibits modest gain at radiation angles of 25 degrees and higher. The 40 meter bowtie predicted SWR curve and antenna efficiency prediction are acceptable.
According to the model predictions, a 40 meter bowtie antenna should provide good local/regional coverage. Encouraged by the predictions, I decided to build a bowtie antenna.
Phase II – Construction
The characteristics of the wire used and environmental conditions present will affect the physical lengths of the radiators needed to achieve the best SWR curve. If you build the antenna, be sure to allow extra radiator length to compensate for wire characteristics and factors such as ground quality and “capacitance” caused by proximity to trees or buildings. The extra wire can be left dangling at the radiator ends and trimmed as needed to achieve the best SWR curve.
TIP: If you don’t have 65’ feet of horizontal space for the antenna, let the radiators dangle at the far ends so you can fit the antenna in the space you have. Keep the lengths of the dangling portions of the radiators the same length and limited to 20% or less of the radiator length (L) and the antenna will still perform well.
End spreaders: The spreaders used to maintain the end spacing were made from 2 X 2 X 8 treated wood furring strips. To facilitate attaching the radiators and support ropes, I inserted 1/4 X 20 eyebolts at the ends of the spreaders. Figure-6 shows a diagram of the spreader bar.
Figure-6. diagram of spreader bar
TIP: If you use 2X2 furring strip spreaders, select the strips carefully to find two that aren’t full of knots, split, bowed, or dripping with preservative. After you insert the eyebolts, spray paint the spreaders and bright metal hardware with dull finish camouflage paint to reduce visibility.
Figure-7 shows one of the completed end spreaders.
Figure-7. completed spreader bar
Feed Point Connector and Balanced Line Support
The feed point connector was made by inserting an eyebolt in the center of a plastic dog bone insulator. The balanced line support was made from rigid composite landscape edging. The support “sandwiches” the balanced line between two pieces of edging. Nylon bolts (1/4 X 20) and wing nuts were used to fasten the two halves of the support. The support is attached to the feed point connector with a large plastic zip tie. Ring terminals were crimped on the ends of the radiators and balanced line. The radiators were connected to the balanced line with 3/4” long #8 pan head stainless steel machine screws, flat washers, lock washers, and wing nuts. Figure-8 shows the completed feed point connector and balanced line support.
Figure-8. feed point connector and balanced line support
Phase III – Analysis
The completed bowtie antenna was hoisted and analyzed with an AIM-4170C. Figure-9 shows the measured SWR curve after the antenna was tuned. Notice the measured SWR curve is lower than the model predicted SWR curve and is relatively flat across the 40 meter band.
Figure-9. 40 meter bowtie AIM-4170C measured SWR curve
40 Meter Bowtie Calculated Antenna Efficiency
The AIM-4170C SWR readings and the online transmission line loss calculator were used to calculate the efficiency of the completed bowtie antenna. Figure-10 shows calculated antenna efficiency at three frequencies across the 40 meter band.
Figure-10. 40 meter Bowtie antenna efficiency (online calculator)
A balanced line matching network is the preferred approach to matching the bowtie antenna to a transmitter. Figure 10a is a diagram of a balanced line antenna matching network.
Figure-10a. Balanced line matching network
Unfortunately, most of us don’t have a balanced line matching unit readily available. Integrated autotuners found in most amateur radio transceivers are designed to accept unbalanced coax transmission lines.
The good news is that we can connect the balanced line to an unbalanced antenna tuner by inserting a balun between the balanced line and the tuner input. Since the experimental antenna has a 4:1 balun connected to the balanced transmission line, a matching circuit such as a high pass T-network can be used.
We can get an idea of how a high pass T-network matching circuit will affect antenna performance by using a free T-network tuner simulator (courtesy of W9CF). Figure 10b is a diagram of a typical adjustable T-network matching circuit.
Figure-10b. Typical adjustable T-network circuit
To use the simulator we need to know the load R ±j values at each frequency of interest. Figure-10c shows the source R ±j values measured by the AIM-4170C.
Figure-10c. AIM-4170C source impedance measurements
Fortunately, we can use the free TLDetails transmission line loss calculator (courtesy of AC6LA) and the AIM-4170C measured source R ±j values (Figure-10c) to get the load R ±j values needed to use the tuner simulator. Figure 10d shows the TLDetails load impedance values for 7.15 MHz. Look for the cursor arrow (at load) on the left side of the results panel.
Figure-10d. AIM-4170C source impedance measurements
Enter the frequency and load impedance values from TLDetails into the tuner simulator. Click Autotune. The simulator will display the SWR match and estimated tuner loss. The values Autotune selected for C1, L, and C2 will also be displayed. We can vary the tuner component values using Setup to reflect the characteristics of a real T-network tuner. Figure 10e shows the tuner simulator display for 7.15 MHz. The simulator indicates that a SWR of 1:1 can be achieved. Calculated tuner loss is 4.8%.
Figure-10e. T-network tuner simulator display for 7.15 MHz
We can refine calculated antenna efficiency by subtracting estimated tuner loss from net power available. There will be additional loss incurred with the 4:1 balun depending upon the impedance.
The completed bowtie is reasonably efficient for a 40 meter wire antenna at a height of 25’ (~.28 wavelength). Most autotuners should have no problem matching the antenna to the transmitter.
Phase IV – Operational Test
The operational test configuration for the bowtie antenna consisted of a IC-706 MKIIG transceiver and a Palstar AT-500 manual antenna tuner. The Wireman 554 balanced line TL was connected to an LDG RBA-4 4:1 balun. The RBA-4 was connected directly to the AT-500 using a double-ended PL-259 connector. The AT-500 was connected to the IC-706MKIIG with an 18″ RG-8X coax jumper cable. Figure-11 shows the operational test configuration.
Figure-11. Operational test configuration
Bowtie Operational Test Results
An operational test was conducted on September 17, 2013. QTH Cary, North Carolina running 100 watts SSB.
I’m a member of the Navy Amateur Radio Club (K4NAR.org) that operates a 40 meter net on 7.245 MHz (0800-1000) daily. I was able to check in with the duty net control located in Georgia with a 59 signal report. The net routinely has check-ins from all over the eastern half of the United States. I was able to clearly hear check-ins from New York to Florida and as far west as Ohio. Check-ins from North Carolina were also clearly heard. I checked in with the daily Communications Ontario Net (Chatham, Ontario, Canada) on 7.153 MHz with a 59 signal report. I was able to hear numerous net check-ins from all over the Eastern half of the country on ECARS (7.255), MIDCARS (7.258), and SOUTH CARS (7.251).
The bowtie antenna performs well as a local/regional antenna. In my case, it covered the Eastern half of the United States and lower Canada. The bowtie antenna would make a good Field Day antenna that could easily be suspended between trees or a couple of portable masts. As a bonus, the antenna will perform acceptably on the 20, 15, 12, and 6 meter bands if an antenna tuner is used.