All posts by Martin Blustine

I studied physics and went on to work in infrared optics, millimeter wave and microwave engineering until retirement. My interests lie in teaching, music, radio astronomy, infrared systems and microwave and antenna engineering. I enjoy writing technical papers about ham radio topics. When I am not operating CW, I enjoy homebrewing ham gear and restoring vintage HP and Agilent test and measurement equipment.

Antennas, Featured, General, Homebrewing, Newsletter, Station Equipment

Highly Efficient L-Matching Networks for End-Fed Half-Wave Antennas

June 11, 2022 Martin Blustine

Introduction

End-fed half-wave (EFHW) antennas provide a convenient way to move the coaxial feedpoint for a half-wave antenna from the center to one end. There are three types of antenna feeds in popular use: 1) transformer feed, 2) L-matching network feed, and 3) Zepp.

The transformer feed variety may be broadband and, depending upon its construction may be useful over the entire HF spectrum. On the other hand, an L-matching network will only work on a single band. There are distinct advantages and disadvantages for each type of feed. This article will discuss the L-matching network feed in detail. The subject of ferrite transformer feeds will be discussed in detail in another article. The subject of Zepp antenna feeds is discussed in great detail in many antenna books [1]. One may think of a Zepp antenna as a folded full-wave antenna usually fed 0.25 wavelength from one end. When one unfolds the antenna, we have an off-center-fed (OCF), full-wave antenna. Depending upon the feed section, a single or double Zepp antenna may be designed to have the advantage of fundamental and harmonic operation.

If fed with an L-matching network, an EFHW antenna may possess an efficiency that exceeds 95%. Here are some things to consider when designing and using this antenna.

First, the load impedance at the end of the half-wave wire must be determined by iteration or through electromagnetic analysis. The load impedance will dictate the values of the L-matching network components. Parameters that have the greatest effect on load impedance are antenna height above terrain, physical properties of the antenna conductor, ground properties, and immediate surroundings.

If we may digress for a moment, the radiation resistance of a center-fed half-wave dipole is well understood, and it varies with wavelengths above the terrain, as shown in Figure 1. For example, if we were to elevate a 40m dipole to 33′ (10m) above the terrain, we would expect a radiation resistance of about 80 ohms at the bottom of the band. This has been verified by modeling the antenna in EZNEC [2] (81.3 ohms) and by field measurement (78.8 ohms) [3]. One would expect the end resistance of a half-wave wire to vary in the same way, although at a vastly different impedance level.

Figure 1. Radiation Resistance of a Dipole Above a Perfectly Conducting Plane. Radiation resistance values for horizontal and vertical dipoles are shown as a function of wavelength above a perfectly conducting plane. The dotted line shows how the radiation resistance of a horizontal dipole departs from the graph when the dipole is close to real ground. It is expected that the end resistance of our EFHW antenna will follow similar variations as a function of wavelength above ground, although at very different impedance levels.

This article will present test results measured at 5 frequencies in 5 bands to demonstrate how end resistances vary at a fixed height above terrain. We allow the number of wavelengths above ground to vary as a function of frequency while holding the antenna height constant. There is not enough test data to draw any conclusions other than to say that the end resistance appears to vary as a function of wavelength above the terrain. It is hoped that the measurement program will be completed as future work.

Second, if the antenna is fed close to where the radio operator will sit, as may be the case for portable operation, the electromagnetic fields at the end of the antenna may rapidly exceed those recommended by the FCC for controlled and uncontrolled environments. For this reason, it is recommended that both ends of the antenna be elevated, with the possible exception of QRP operation.

Third, any end-fed half-wave antenna will require a counterpoise. Sometimes, the counterpoise is provided by the coax that feeds the L-matching network, in which case the feedline will definitely carry common mode currents that will tend to spoil the antenna pattern of the half-wave wire. Since the coax that feeds the L-matching network steals current from the half-wave wire, it reduces the main-beam efficiency. This problem can be remedied by placing a common mode choke in the feedline at a distance of 0.05 wavelength [4] from the L-matching network. Since the common mode currents are on the outside of the shield, the value of 0.05 wavelength is measured in air, not in coax. Any effect due to the outer jacketing will be small. Be advised, however, that the 0.05 wavelength value is not cast in concrete. It is a starting point. To be absolutely sure, the currents on the transmission line shield, counterpoise, and antenna wire should be measured with an RF current probe. One such probe is the MFJ-854 [5]. As will be shown by electromagnetic analysis with EZNEC in another article, there will always be some RF currents where we don’t want them to be.

While a portion of the coaxial feedline may serve as a counterpoise, the alternative is to co-locate the common mode choke with the L-matching network so that a separate counterpoise wire can be connected to the L-matching network ground. In any event, the counterpoise conductor should be perpendicular to the half-wave radiator, regardless of whether it is horizontal or vertical, to minimize the interaction of counterpoise fields with the main antenna beam.

Methodology

A lowpass L-matching topology is employed as shown in Figure 2 – a series inductor followed by a shunt, open-ended coaxial capacitor consisting of RG-316/U. (RG-316/U coax has a capacitance of 29 pf/ft.) It is widely known that if the shunt-matching element is in parallel with the load, the transformation will be from a higher load impedance to a lower source impedance [6]. This configuration is very easy to tune. However, this configuration will not bleed static charge from the antenna wire. While a series capacitor followed by a shunt inductor in parallel with the load will provide a similar transformation and static protection, this topology is more expensive to realize and more difficult to tune. We bring this point to your attention because static protection is often omitted from antenna installations.

The matching schematic below, and tutorial from Professor Stephen Long, Emeritus, ECE, UCSB, may be found online [7].Figure 2. Lowpass Topology. The inductor may be air-wound or toroidally-wound, while the capacitor consists of an open-ended piece of RG-316/U. While the capacitor appears to be grounded at the bottom of the capacitor, for this implementation, the ground connection is soldered to the shield-end closest to the inductor. R_s is the 50-ohm source impedance, real, while R_p is the antenna end impedance, real.

An online calculator that performs all of the calculations for a variety of L-Matching Network topologies may be found on John Wetherell’s website [8]. While online calculators are great, the reader is encouraged to perform at least one calculation if only to verify that the calculator yields the same result as theirs.

A quick way to arrive at the values for the matching elements is to estimate the impedance transformation ratio required. For example, if we needed to transform from a 50-ohm source to a 4800-ohm load impedance, the impedance transformation ratio would be 1:96. From that, we obtain the unloaded Q and the reactance values required.The unloaded Q = 9.75, and the reactance required for the inductor at the design frequency would be 488 ohms. The reactance required for the capacitor would be 492 ohms. If we were designing for 80m, we would calculate the inductance and capacitance from the reactance formulas at, for example, 3.6 MHz.

The series inductance would be 21.6 μH, and the shunt capacitance would be 89.8 pf. Then, we would build the network with a coaxial capacitor that had been cut too long and load the network with a 4800-ohm non-inductive resistor. An antenna analyzer would be connected to the network as the source, and the coaxial capacitor would be pruned a tiny bit at a time (it is easy to overshoot the mark) until the reactance at the design frequency had been canceled. If the antenna analyzer could be connected to a PC, tuning might be performed in real-time. A convenient display to use would be the Smith Chart view.

Two tables of component values at frequencies in commonly used ham bands are provided in Table 1 for CW and Table 2 for SSB. The tables are parametric and provide component values for a variety of EFHW antenna end resistances from 2000 to 4800 ohms.

A good starting value for any design is 3200 ohms, and in most cases, it will deliver satisfactory results. One need not rely upon the design frequencies in the tables. You now have all tools that you need to design networks for any frequency and any load impedance!Table 1. CW Band Segment Component Values as a Function of End-Fed Half-Wave Antenna End Resistance.Table 2. SSB Band Segment Component Values as a Function of End-Fed Half-Wave Antenna End Resistance.

Determining Component Values

Access to an Inductance-Capacitance-Resistance (LCR) meter is highly desirable to ensure that the L-matching networks are being built to the design values. Fortunately, these instruments have become inexpensive and accurate, and they are quite useful for identifying unknown component values.

Implementation

A 30m L-matching network test article that employs an air-wound inductor is depicted in Figure 3. The sketch in Figure 4 shows the network interconnects in greater detail. The final iterated load impedance was 3024 ohms, the inductance value was 6.16 μH, and the final capacitance value was 41.7 pf. The resulting VSWR was 1.04:1 at 10.133 MHz. The air inductor was been wound on a scrap length of ½” PVC conduit. The OD of this conduit is 0.840” (21.34mm). The inductance wound for this test article consists of 22 tightly-wound turns of #20 AWG enameled magnet wire, but #18 AWG is highly recommended for the final L-matching network. Hot glue was used to hold the turns together during tests, but non-corrosive silicone is recommended once the inductance has been finalized. You can use an online calculator to estimate the number of turns required for the air-wound inductor of a specific diameter and length [9], but please use an LCR meter for precision. The capacitor consists of an open-ended piece of RG-316/U coax that was chosen for its maximum voltage rating of 1,200 volts and 29 pf/foot capacitance. The approximate length of our coaxial capacitor is 17 inches. In a separate article, we will discuss voltage requirements for transmission lines when subjected to elevated VSWR.

Figure 3. Lowpass L-Matching Network Topology for 30m Test Article. The air-wound inductor consisting of #20 AWG enameled magnet wire is connected from the center conductor of the SO-239 UHF connector straight through to the antenna terminal. Hot glue holds the turns in place. The coaxial capacitor center conductor is connected directly to the antenna terminal. The other end of the coaxial capacitor remains open-circuited. The photo shows a short piece of clear shrink tubing that protects the open end of the coax (see text). A ground connection to the coaxial capacitor shield is soldered near the antenna terminal end of the coax with a piece of buss wire (see Figure 3). The other end of the bus wire is grounded to the flange of the SO-239 UHF connector with a solder lug. The housing is a Bud Industries PN-1320.

Figure 4. Lowpass L-Matching Network with Air-Wound Inductor Interconnects. The sketch shows where an optional, separate counterpoise may be connected if the coax is not used as the counterpoise. It also shows where the ground connection is soldered onto the coaxial capacitor shield. The dotted lines show where to connect the test load resistor to the network.

Figure 5 shows the configuration for a 15m L-matching network test article. The magnet wire gauge used is #18 AWG because the inductor will fit the housing. Hot glue is used to keep the turns in place. The final inductance value is 2.91 μH, and the coaxial capacitor value is 20.7 pf. The test frequency is 21.09 MHz, while the end resistance is 3042 ohms. The uninsulated open circuited coax is visible. The braid and jacket must be pruned back and insulated to prevent RF arcing. A 1 pf capacitor is soldered in parallel with a coaxial capacitor to correct the capacitance. This capacitor combination was replaced with a longer piece of coax for the final configuration.

Figure 5. A 15m L-Matching Network Test Article. A 1 pf capacitor (visible in the white circle) is soldered in parallel with the coaxial capacitor to tune the network without changing the coax. The coaxial capacitor will have to be changed before transmitting because the capacitor is not rated for high RF voltage.

In order to prevent the coaxial capacitor from arcing from its center conductor to the shield near its open end, it is a good idea to snip back the coax shield a tiny bit near the open end. This has to be done while the coaxial capacitor is being tuned to its final value, however. Once one arrives at the final value, the end of the coax may be covered with a short length of shrink tubing to protect it. When tuning these networks for the higher bands, namely 20, 17, 12, and 10 meters, it may be beneficial to complete the tuning with the coax coiled in its final configuration. The pruning procedure can be somewhat fussy, and altering the capacitor configuration at the very end can lead to disappointing results.

From the network configuration, one can see that there is always access to the inductor and capacitor for measurement. It is a good idea to record the as-built values of the network once it has been tuned into the test load. Many times it will be found that the as-built values will not be exactly the same as the design values. The reason is usually the inter-turn capacitance of the air-wound coil. It’s alright as long as there is a good match into the test load resistor. We will explore the inter-turn capacitance in a separate article.

One final note on construction: The inductors for the 160, 80, 60, and 40m bands will require a large number of turns. If air-wound, they may be too long to fit within your enclosure. One solution has been to build the network in a bigger enclosure. That will preserve the low-loss properties of the L-matching network. There will be a separate article that describes the construction of a 600W L-matching network with an air inductor for 40m. For portable use, however, compactness may be a virtue, and you may wish to wind the inductors in some other way at the expense of some RF efficiency. For such cases, consider the use of small ferrite or iron cores. You will want to choose a core type for the frequency range of interest, most likely #2 (red) iron material [10] or #61 ferrite material [11]. The permeability of iron is much lower than it is for ferrite. More turns will be required on an iron core. You may also want to choose a core size that is large enough for several turns. If the core doesn’t have enough turns, it will be difficult to fine-tune the inductance. If there are too few turns, even a small adjustment will dramatically affect the inductance, and the required value will not be obtained. A 40m L-matching network that employs a toroidal inductor is depicted in Figure 6. The sketch in Figure 7 shows the network interconnects in greater detail. An FT114-61 ferrite core has been used for 40m. Eleven turns of #18 AWG magnet wire were required to obtain 10.4 μH. The coaxial capacitor consists of approximately 19” of RG-316/U. The inductance of the toroid may be altered by spreading and compressing the turns and/or by adding and removing turns.

Figure 6. Lowpass L-Matching Network Topology for 40m. The toroidal inductor is connected from the center of the SO-239 UHF connector straight through to the antenna terminal. The coaxial capacitor center conductor is connected directly to the antenna terminal. The other end of the coaxial capacitor remains open-circuited. A ground connection to the coaxial capacitor shield is soldered at the antenna terminal end of the coax with a piece of buss wire (see Figure 5). The other end of the bus wire is grounded to the flange of the SO-239 UHF connector with a solder lug. The housing is a Bud Industries PN-1320.

Figure 7. Lowpass L-Matching Network with Toroidal Inductor Interconnects. The sketch shows where an optional, separate counterpoise may be connected if the coax is not used as the counterpoise. It also shows where the ground connection is soldered onto the coaxial capacitor shield. The dotted lines show where to connect the test load resistor to the network.

The 80m L-matching network shown in Figure 8 was constructed with a T130-2 iron toroid core because the material was available. There are approximately 40 turns on the core for an Inductance of 21.3 μH. This core material permeability is 10, which is about one-tenth the permeability of #61 ferrite material. With so many turns, the inter-turn capacitance is apt to be large. For this design, an FT114 -61 core would be a better choice for fewer turns. This test article was built for a test frequency of 3.55 MHz with an end resistance is 4639 ohms. The coaxial capacitance value is 93.8 pf.

Figure 8. An 80m L-Matching Network. This network was constructed with a T130-2 iron core. The initial permeability of this core is too low, requiring too many turns with attendant inter-turn capacitance. The next iteration will employ FT114-61 material.

Lab Test Procedure

Once the design frequency and antenna load impedance have been chosen for the design, an L-matching network is fabricated. The network is loaded with a non-inductive load resistor (a cermet trimpot is a good choice), having the same value as our antenna load impedance design value. The network is connected to a vector antenna analyzer. If possible, the Smith Chart view should be utilized. Most vector antenna analyzers can be connected to a computer via USB, and a Smith Chart may be displayed in real-time. What we should see, if we sweep through the design frequency, is a display near the center of the Smith Chart. If the display is below the axis of reals for the Smith Chart, the coaxial capacitor will have to be pruned a bit at a time until the display at the design frequency is right at the center of the chart. Thus, the reactance has been canceled, and we are left with a perfect transformation from the load impedance to the characteristic impedance of 50 ohms. If we call for the reactance view on the vector antenna analyzer, we should be able to see the reactance pass through zero at the design frequency of the L-matching network. Finally, we disconnect the load resistor from the network and prepare for field testing.

A Smith Chart for the 30m L-matching network is shown in Figure 9. Data was collected on an MFJ-226 Graphical Antenna Analyzer with the network connected to a 3024-ohm cermet trimpot test load. Once the inductance had been adjusted with an LCR meter, the coaxial capacitor was pruned until the reactance was canceled to place the cursor at the center of the Smith Chart. The network was designed to match 3000 ohms, and a slight adjustment of the trimpot placed the cursor at the center of the chart. Tuning is a bit tricky because the 30m Band is so narrow.

Figure 9. Lab Tuning of the 30m L-Matching Network. The inductor, capacitor, and test load are adjusted to place the cursor and the center of the Smith Chart.

The VSWR performance of the 30m L-matching network is shown in Figure 10. Excellent performance has been achieved over the entire band after lab tuning.Figure 10. VSWR Performance of the 30m L-Matching Network. Excellent performance has been achieved over the entire 30m Band.

The 40m L-matching network of Figure 6 was tested in the lab. The network was loaded by a 4573-ohm load. The resulting Smith Chart that was swept over the entire 40m band is shown in Figure 11.

Figure 11. 40m L-Matching Network Smith Chart. An excellent match is observed over the entire band.

The VSWR performance of the 40m L-matching network is shown in Figure 12. Excellent performance has been achieved over the entire band after lab tuning.

Figure 12. VSWR Performance of the 30m L-Matching Network. Excellent performance has been achieved over the entire 40m Band.

Field-Test Prep

In preparation for field-testing, we will want to prepare the counterpoise that will consist of a length of coax that has been cut to 0.05 wavelength at the design frequency. If the common mode choke will be connected to the L-matching network at the SO-239, a wire should be cut to 0.05 wavelength at the design frequency. A grounding stud should be provided to the network ground for counterpoise connection. This may be a separate ground stud mounted on the case and connected to the SO-239 connector flange, for example. Please refer to one of the interconnection diagrams for the counterpoise connection point.

Field Test Procedure

In order to test the EFHW antenna with an L-matching network feed, two end supports will be necessary. The supports may be a pair of nearby trees, a pair of telescoping masts, or one of each. Each support should be fitted with a halyard so that each end may be raised and lowered. Ideally, testing should be performed at least 0.25 wavelength above ground, except, of course, for the case of an inverted-L. The L-matching network and common mode choke will be heavy, so the end support has to be sturdy enough to support the weight of the transmission line, matching network, counterpoise, and choke. Once that is arranged, testing may begin.

Without altering the L-matching network or counterpoise in any way, the length of the antenna wire should be pruned to cancel any remaining reactance at the design frequency as viewed on the antenna analyzer. Once that has been accomplished, the real part of the impedance may be read. It should be close to 50 ohms. If it is not, then we know that we have chosen the incorrect impedance transformation ratio and, consequently, the incorrect value for unloaded Q.

Figure 13 shows the field-test Smith Chart for the 30m L-matching network. The swept measurement places the cursor very close to the center of the Smith Chart. The actual performance is very good for the entire 30m Band.Figure 13. Field-Test Smith Chart for the 30m L-Matching Network. Measured performance is very good over the entire 30m Band.

The field-test VSWR performance of the 30m L-matching network is shown in Figure 14. Excellent performance has been achieved over the entire 30m Band.

Figure 14. Field Test VSWR Chart for the 30m L-Matching Network. VSWR performance is very good over the entire 30m Band.

If the VSWR is satisfactory, testing may be completed. If it is not, iteration is the next step. Usually, the real part of the transformed antenna resistance will hint at whether we have chosen a transformation ratio that is too high or too low.

There is another approach that is worth considering. An L-matching network can be constructed from a tapped series inductor and a shunt air variable capacitor. Calculations may be performed in the field to alter the impedance transformation ratio by adjusting the inductor and capacitor. Once satisfied with the result, the inductor and capacitor values may be made permanent. If you happen to have a home with a second-story window, you might consider stringing the antenna from the window out to a tree or mast. Then, the L-matching network adjustments can be made from the open window. Please take care to keep children and pets away from any open window.

In any event, the antenna wire length will have to be restored to its full starting length so that it may be pruned enough to cancel any residual reactance that might be observed for the iterated L-matching network. Tiny wire rope cable clamps [12] shown in Figure 15 provide a way to splice pieces of wire back onto the antenna.

Figure 15. Wire Rope Clips. The 2mm and 3 mm variety are handy for splicing pieces of antenna wire together.

By now, you have probably figured out how to perform some reverse engineering from the field-test results. If the values of the inductance and capacitance have been determined experimentally, it is possible to work backward to determine the unloaded Q and then the impedance transformation required to obtain a match. From this, you will be able to estimate the end resistance of the wire.

Field Test Results

Some field-test results are summarized in Table 3. Many of the load impedances for the L-matching networks have been iterated, and these values appear in the table. Because of weather considerations on the test range in Florida in spring 2020, the test program was suspended. Nonetheless, all of the networks were tested and iterated, with the exception of the 80m L-matching network. This may have been fortuitous, as there was no means to elevate the 80m antenna to a height exceeding 0.25λ (66 ft) with the masts available.

All of the measurements in Table 3 were made with 50′ of RG-8X coax. The reference plane of each measurement was moved to the plane of the SO-239 connector on each L-matching network by means of an open/short/load calibration load kit.

Table 3. Spring 2020 Field-Test Data Summary. The as-built component values and VSWR results are as reported. The L-matching networks and coax were supported by a 33’ (10m) mast. The mast was judged to be of insufficient strength to support an additional common mode choke. Consequently, all measurements were made with a 50’ length of RG-8X where the entire coax acted as the counterpoise. No attempt was made to optimize the length of the counterpoise. The 80m L-matching network could not be tested because of the weather. See text.

References

[1] ARRL Antenna Book, 21st ed., Ch. 7, p. 2.
[2] https://www.eznec.com/
[3] Blustine, M., L-Matching Networks – Field Measurements, March 17, 2020.
[4] https://www.aa5tb.com/efha.html
[5] https://mfjenterprises.com/products/mfj-854
[6] https://web.ece.ucsb.edu/~long/ece145a/Notes5_Matching_networks.pdf
[7] Ibid.
[8] https://home.sandiego.edu/~ekim/e194rfs01/jwmatcher/matcher2.html
[9] https://www.66pacific.com/calculators/coil-inductance-calculator.aspx
[10] http://www.qrz.lt/ly1gp/amidon.html
[11] https://www.66pacific.com/calculators/toroid-coil-winding-calculator.aspx
[12] https://www.amazon.com/Rannb-Stainless-Simplex-Single-Diameter/dp/B07DLVNR54/ref=sr_1_7?crid=26694JFBAHN04&keywords=cable+clamp+3mm&qid=1654922072&sprefix=cable+clamp+3mm%2Caps%2C85&sr=8-7

Martin, K1FQL

Antennas, Featured, General, Newsletter

A Flagpole Antenna Project for Residential Settings

May 23, 2022 Martin Blustine

Introduction

A flagpole antenna has been fabricated that can be installed in any residential setting. The fiberglass tubing has been designed with a pair of 2′-long coupling sleeves so that the entire tubing assembly will separate into 3 sections. The flagpole has been designed to contain a 6-BTV vertical [1], including an 80m resonator and stinger. When equipped with the 80m resonator and stinger, the entire flagpole stands 25′ high. The flagpole is painted with Marine Epoxy finish [2] to preserve the fiberglass. The tilt-over mechanism [3] for the 6-BTV is contained within a white deck post sleeve [4]. A remote tuner [5] and common mode choke is concealed within a white circular pallet that has been fabricated from PVC trim lumber. The pallet also conceals the radial plate [6]. A standard flag pole truck [7] with a finial ball [8] caps the tubing for a finished appearance.

Flagpole Antenna Construction Details

Concrete Pour, Radial Plate, 6-BTV Concealment, and Common Mode Current Suppression

The concrete has been poured, and a radial plate [9] installed for a 6-BTV [10] Vertical Antenna. The foundation for the 6-BTV is an 8” DIA x 4’-0” long concrete form tube [11]. A 1-1/2″ OD seamless stainless tube [12] was embedded in the concrete for attachment of the 6-BTV. A pipe or tube no larger than 1-1/2″ OD must be used, if the entire tilt-over assembly is to fit within an 8″x 8″ deck post sleeve. A total of 60 radials [13] were bolted to the radial plate [14], Figure 1. These vary in length from 35 ft (10.67m) to 8.5 ft (2.59m). The radials are anchored with 6″ x 11 gauge landscape staples [15]. In one week, the radials were nearly concealed by the lawn. The 6-BTV is concealed within 2.5″ OD, 2.26″ ID, fiberglass tubing [16]. The tubing was joined with short sleeves of 2.75″ OD, 2.51″ ID, and fiberglass tubing [17]. A 1-1/2″ (38.1mm) PVC electrical conduit [18] is also visible on the far left. It is used to carry 75 ft (21.9m) of DX Engineering DXE-400MAXDX075 [19] from the service entrance to the antenna. A common mode choke [20] was placed right at the antenna to suppress common mode currents on the coax shield.

Figure 1. DX Engineering Radial Plate. Sixty ground radials are visible. Each bolt has been double-nutted so that the radials may be disconnected without lifting the ground plate. In order to save time, each of the 60 radials was prefabricated to the plot plan to fit the available space.

Figure 2 shows the brick border [21] that was added to the radial plate installation. It serves to protect the antenna, radials, and conduit from the lawn mower. It also serves to hide the lawn that is close to the radial plate. As the radials converge on the radial plate, there is considerable lawn die-off. Once this patch of lawn is covered with a tree ring landscape apron [22] as pictured in Figure 3 and a decorative pallet, it is no longer visible.

Figure 2. Brick Border. The border provides a convenient way to protect the radial connections and radial plate while mowing. The underground conduit is also visible.

Figure 3. Landscape Apron. A landscape apron was installed over the radial plate to retard the growth of grass and weeds. The mounting pipe length was shortened. The underground electrical conduit that carries coax and the antenna tuner control cable is visible.

Fiberglass Tubing

A 25′ tall fiberglass flagpole has been constructed that will conceal a 6-BTV vertical antenna. A picture of the sections, as constructed, is provided in Figure 4. In order to clear the aviation hose clamps that clamp the 6-BTV aluminum tubing and resonators together, as well as the 80m coil, a pole has been fabricated from 2-1/2″ OD fiberglass tubing [23] obtained from DX Engineering [24]. The sections may be separated for storage during high winds, and these sections have been coupled together with sleeves constructed from 2-3/4″ OD x 24″-long fiberglass tubing [25] also obtained from DXE. The three sections measure, from foreground to background: Bottom: 8′ (96″), Middle: 12′ (144″), Top: 8′ (96″).

Figure 4. Fiberglass Tubing Sections. Sections of tubing are bonded together with non-corrosive RTV. If bonding is performed as shown, the assembly will spill water, once assembled. See text.

Once assembled with overlap, the overall length is 25′ long not including the flagpole truck [26] and finial ball [27]. It is anticipated that an additional 1′ of 2-1/2″ OD tubing may be removed just below the truck/ball, and this will reduce the bending moment. The fiberglass dielectric was expected to change the antenna tuning but it turned out to be negligible. The proximity of the truck and ball to the 80m “stinger” resonator was evaluated in its present configuration, and its effect was negligible.

Note that it is essential that the DX Engineering DXE-RLT Reinforced Lower Tube [28] manufactured for the BTV antenna series be utilized in order to support the additional bending moment of the flagpole.

Use Without 80m Resonator

If 80m operation is not required, the top 2-1/2” OD fiberglass tubing may be omitted. Only the bottom two sections need to be fabricated. The flagpole truck and finial ball are moved down by one section without modification to the design.

Fiberglass Tub

ing Fabrication

The following steps were performed outdoors using personal safety equipment (PSE) consisting of a 3M fine particle respirator [29], safety goggles and gloves. If the flagpole is to be painted, a suitable respirator designed specifically for the constituent paint volatiles must be employed in a well-ventilated area, preferably outdoors.

The tubing sections are cut using a straightforward technique. First, the pieces of tubing being cut are wrapped circumferentially, under the cut-line, with ordinary masking tape and the desired lengths are marked circumferentially with dots from an indelible marker. The tubing is placed on a set of V-blocks, Figure 5, that come as an accessory to a set of Stanley collapsible sawhorses [30]. The V-blocks slip onto the ends of the sawhorse.

The tubing is rotated in the V-blocks and gently scored on the dots with a small reciprocating saw [31] that holds a Lenox 8″ Diamond Blade, P/N 800RDG [32]. Once the circumference is scored, the tubing is rotated while applying constant pressure to the reciprocating saw to complete the cut all the way through. If care is taken, the fiberglass cut will be clean and straight, and there will be minimal fraying of the glass fiber. The masking tape is removed, and the cut is gently finished with plumber’s grit roll [33] or emery cloth.

The adhesive chosen for assembly is Loctite SI 5011 CL Non-Corrosive RTV [34], also known as “ULTRAFLEX”. This is a neutral cure adhesive/sealant that cures by emitting non-corrosive methanol. It takes some additional time to cure, but it is worth it.

Corrosive silicone adhesives continue to cure over a long period by absorbing water and emitting acetic acid. It is a good idea to avoid using them.

Loctite SI 5011 CL adhesive comes in a caulking tube, and this project will consume over half a tube. The pieces of tubing can be wrapped with masking tape so that when the tubing is assembled and the excess adhesive is removed, a clean adhesive line can be obtained.

Figure 5. Sawhorse V-Blocks. (Top) The V-Blocks slide onto each end of the expanding top rail of the sawhorse. (Bottom) The V-Blocks are used to rotate the fiberglass tubing while it is being cut with a diamond reciprocating saw blade. Rotating the work results in a square cut. See text.

Please spend some time studying the photograph, Figure 4, and the assembly sketch, Figure 6, so that the adhesive is applied in the correct places. This will ensure that the assembly will shed water and slide apart.

Figure 6. Fiberglass Tubing Assembly Drawing. The adhesive is applied to the 3 larger diameter sleeves so that the sections will disassemble into 3 pieces. Please study the drawing at the 204″ mark. You will see where two of the outer sleeves touch. When assembled, the sleeves will spill water. In the final version, the 80m resonator and top tubing were removed. See text.

Once the tubing is correctly joined, the flagpole will come apart, as shown in the photo. Please note that the assembly sketch is just that – a sketch – it’s not to scale. The white gap that is visible in the upper section is an air gap for the 80m coil. It provides some additional airflow around the coil. The adhesive is applied in a number of long stripes along the 2-1/2″ OD tubing. The 2-1/2″ OD tubing is then slowly rotated within the 2-3/4″ OD sleeving until the adhesive is spread uniformly between the tubes being joined, and the correct overlap is obtained, as is shown in the photograph, Figure 4, and assembly sketch, Figure 6.

Excess adhesive is removed with a paper towel or cloth. Care is taken to provide a continuous bead of adhesive around each joint so that each joint is waterproof. Once the adhesive is dry, the masking tape is removed. The flagpole is assembled with adhesive so that all of the sleeves “spill water”, i.e., water does not enter the sleeves. This makes tubing disassembly easy once the flagpole antenna is tilted on its OMNI-TILT mount [35].

Flagpole Truck, Finial, Halyard, and Cleat

A standard 2-1/2″ flagpole truck [36] and decorative finial ball [37], visible in the title photo, finishes the project. These may be obtained together or separately from Amazon or eBay. The flag halyard is made from polyester double braid rope that is 1/4″ in diameter. This material provides very good sun resistance. A good source of supply is Miami Cordage [38]. A Nylon halyard cleat [39] is Ty-wrapped to the lower 2-1/2″ OD tubing section. Do not drill the fiberglass tubing. Drilling will weaken it. Nylon cleats may be obtained at Home Depot [40] or Lowes [41]. The flag should be correctly sized for the flagpole height [42]. The length of the flag should measure at least one-quarter the flagpole height. For a 20′ flagpole, the flag should measure 3′ x 5′.

Flagpole Finish

Although the DXE tubing is advertised as having sun resistance, a coat of paint is advised. A one-part or two-part epoxy paint will work well provided that the work has been prepared. Shades that utilize little or no carbon pigment are best. Interlux topside finishes are very good and may be obtained from West Marine [43]. Please make a note of all cautionary notices that accompany these products.

Deck-Post Sleeve for Tilt-Over Mechanism Concealment

The vinyl 8″ x 8″ deck-post sleeve [44], obtained from Penn Fencing that conceals the DX Engineering DXE-OMNITILT-1P [45] antenna tilt-over mechanism, is shown with its removable cap [46] and skirt [47] in Figure 7. The 2-3/4″ OD hole that is visible in the cap clears the 2-1/2″ OD fiberglass flagpole tubing. The gap is waterproofed with an Oatey neoprene rain collar [48] to keep water from running down the pole and into the tilt-over mechanism.

Figure 7. Deck-Post Sleeve. The antenna base and tilt-over mechanism is concealed in an 8” x 8” vinyl deck-post sleeve.

Tilt-Over Mount Hardware Modifications

The DX Engineering 6-BTV OMNI-Tilt Vertical Antenna Tilt Base, DXE-OMNITILT-1P [49], Figure 8, has received minor modifications so that it may be concealed within an 8″ x 8″ vinyl deck-post sleeve manufactured by Superior Plastics and sold by Penn Fencing [50]. The lengths of some of the stainless steel hex head bolts were shortened and some stainless steel jam nuts were used in place of stainless hex nuts, because they are not as thick. The actual OD of the ground-mounting pipe cannot exceed 1-1/2″, or the mechanism will not fit. Some bowing of the sleeve is evident, but this is always present. The mechanism has been fit-checked. A right angle adapter [51] provides access to the SO-239 connector.

Changes to DX Engineering hardware kits and ground mounting pipe are easy, and the changes have been listed below. Replacements were obtained from Fastenal [52] in 18-8 (304) stainless steel for all hardware kit changes. A stainless 1.5″ OD x 0.156″ wall thickness ground mounted pipe was obtained from OnlineMetals.com [53].

Use 1.5″ OD x 0.156″ wall thickness ground mounted pipe (the actual 1.5″ OD dimension is critical)
Change the DXE-OTMC-250C [54] kit from 5/16″-18 x 4-1/2″ long Hex Bolt, Full Thread, 18-8 Stainless Steel to 5/16″-18 x 2-3/4″ long Hex Bolt, Full Thread, 18-8 Stainless Steel.
Change the DXE-OTMC-250C kit from 5/16″-18 Hex Nut, 18-8 Stainless Steel to 5/16″-18 Hex Jam Nut, 18-8 Stainless Steel.
Change the DXE-OMNITILT-1 kit from 5/16″-18 x 1-1/2″ long Hex Bolt, Full Thread, 18-8 Stainless Steel to 5/16″-18 x 1-1/4″ long Hex Bolt, Full Thread, 18-8 Stainless Steel.

The changes to all listed items are critical and must all be completed, or the mechanism will not fit within the 8″ x 8″ deck-post sleeve. Once carefully tightened, the thin jam nuts will be fully threaded. No changes to the flat washers and lock washers provided by DX Engineering have been made.

Figure 8. DXE-OMNITILT-1 Tilt-Over Base. The DXE-OMNITILT-1 tilt-over mechanism is shown clamped to the 1-1/2″ OD stainless steel pipe that has been embedded in a concrete foundation. One of the 6-BTV base insulators is visible, as is the RF right-angle connector.

Figure 9 provides a top view of the deck-post sleeve with the DXE-OMNITILT-1 and 6-BTV base (red cap) installed. The entire assembly fits within the 8” x 8” sleeve once thin jam nuts have been installed.

Figure 9. Top View of Deck-Post Sleeve. The 6-BTV base (capped) and DXE-OMNITILT-1 tilt-over mechanism clamped to the 1-1/2″ OD mounting pipe are visible.

Figure 10 provides a better view from the bottom of the deck-post sleeve. The DXE-OMNITILT-1 is clearly visible as is the base insulator of the 6-BTV and right angle RF adapter.

Figure 10. Bottom View of Deck-Post Sleeve. The RF connections to the 6-BTV base are visible, as are the tilt-over mount and stainless steel mounting pipe.

Static Charge Bleeder Resistor and Ground Straps

Figure 11 shows a 3 MEGOHM carbon film bleeder resistor [55] connected between the antenna feed-point and ground. This has been waterproofed within a piece of shrink tubing. The ends have been sealed with Loctite SI 5011 CL Ultraflex Non-Corrosive RTV [56]. This will bleed static charge from the antenna when the DC grounded step-down transformer or shunt matching coil is not present, as will be the case on 10 and 15m. The tinned ground braids in the photo are temporary, and copper ribbon will become the permanent fix to provide a low resistance path to the radial plate.

Figure 11. The 6-BTV Base. The RF connections, bleeder resistor (see text), and temporary ground straps are visible. The straps were replaced with solid copper ribbon.

Deck-Post Sleeve Modifications and Rain Collar

The deck-post sleeve cap has to be modified so that the flagpole 2-1/2″ OD fiberglass tubing (not the mounting pipe) can pass through it. Please refer to Figure 9 to see the 6-BTV tubing base with its red protective cover. The 2-1/2″ OD fiberglass tube will slide over this. The dimensions that were used to locate the hole in the deck-post sleeve cap are shown below in Figure 12. The hole has been bored out to 2-3/4″ to ensure 1/8″ clearance all around when the 2-1/2″ tubing is slipped into place. The positional dimensions should be verified for each and every installation, and a cardboard template should be constructed to ensure that the location of the hole is correct, as these caps are expensive. Please note that the hole is positioned in reference to the center of the deck-sleeve cap. Clamp the work, drill a small pilot hole, then use a new hole saw at low speed to bore through slowly.

Figure 12. Boring Dimensions for Deck-Post Sleeve Cap. Please note the location of the center of the deck-post sleeve cap.

The finished deck cap is shown in Figure 13. If the cap has been bored correctly, the 6-BTV base tubing will be centered in the hole. If not, the fiberglass tubing will not be centered and will not clear the hole.

Figure 13. Top of Deck-Post Sleeve Cap. The 6-BTV and fiberglass tubing will clear the cap with room to spare.

In order to prevent water infiltration down the pole and through the cap, an Oatey All-Flash Rain Collar [57], Figure 14, 1-1/2 in. to 3 in., has been used. Rain collars are available at home improvement and plumbing stores. The reason that the hole in the cap was bored to 2-3/4″ and not 2-1/2″ was one convenience for assembly. There would be is plenty of lateral play in the deck-post sleeve and the flagpole if the hole were bored to 2-1/2″. In either case, a rain sleeve would be required.

Figure 14. Oatey Rain Collar. It prevents rain from running down the fiberglass tube and into the antenna base.

6-BTV Feed-Point Resistance and UNUN

The 6-BTV was installed long enough to tune the traps. An MFJ-226 Vector Analyzer [58] was used at the feed-point of the 6-BTV. Ideally, the feed-point resistance for a quarter-wave radiator above a perfectly conducting ground plane should be half that of a center-fed dipole, or 36.5 ohms. Sadly, a feed-point resistance nowhere near 36.5 ohms would be observed on 3 of the 6 bands. The 10m trap could be adjusted so that the VSWR over the entire CW band was under 1.2:1. On 15m, the VSWR could be adjusted to under 1.3:1. Not so for the 20m trap. There, in spite of hours of tinkering with the trap sleeve and tubing below the trap, an acceptable match of under 2:1 could not be obtained. A great deal of time was spent inspecting the trap and searching for antenna and trap resonances. Convinced that the antenna resonant frequency had been found, the resistance was measured at the feed-point. The process was repeated for the 30m trap. There is no 40m trap. For the 20, 30, and 40m bands, the feed-point resistance was measured between 25 and 18 ohms. A shunt-matching coil like the DX Engineering DXE-VMN-1 [59] was considered in order to raise the feed-point resistance to a manageable one. Since K-type ferrite was available, a Jerry Sevick 2.25:1 UNUN was constructed for the feed-point. It provides a transformation to 22 ohms and is easier to construct than the one that transforms to 25 ohms. If you encounter similar matching problems, Balun Designs can supply a ready-made step-down transformer of this type, Model 1232 [60]. The mismatch is acceptable for the 20, 30, and 40m bands after the UNUN is installed. An alternative is to locate a remote antenna tuner at the base of the antenna. No difficulty was encountered when tuning the 80m resonator and stinger. The match over a 30 kHz segment at the bottom of the CW band was better than 2:1.

Line Choke

A line choke was added right after the UNUN to choke-off common mode currents that would otherwise travel to the operating position. The UNUN places the antenna at DC ground potential, so no other method is required to bleed static charge from the antenna. If the UNUN is not used, then the bleeder resistor shown in Figure 11 will bleed static from the antenna.

Matching to the 6-BTV Feed-point and Protecting the Remote Tuner

The 6-BTV is very well matched on 10 and 15m, but antenna impedance matching is required on the 20, 30 and 40m bands. An SGC-239 tuner [61] was tested at the base of the antenna, and it worked very well. Unfortunately, CW operation is restricted to 200W. This is inadequate for operation with the Ameritron ALS-600 linear amplifier [62]. Consequently, a 1 kW PEP Stockcorner JC-4 [63] was installed at the 6-BTV base. In order to conceal the tuner, a decorative circular pallet 24″ (61cm) in diameter was constructed from a white PVC trim board, Figure 15, and the tuner was fastened to the bottom of the pallet along with an input line choke, also visible.

Figure 15. PVC Lumber Pallet. A synthetic, PVC lumber pallet (Top) conceals the Stockcorner JC-4 remote antenna tuner and a homebrew common mode choke (Bottom). Height is sufficient to prevent the tuner from sitting in a puddle of water.

The pallet height, 6-1/4″ (15.9 cm), is sufficient to keep the tuner above rainwater. The deck-post sleeve, Figure 16, now sits on top of the pallet instead of the radial plate.

Figure 16. Flagpole Antenna Base Pallet. A deck-post sleeve and synthetic PVC lumber pallet conceal the tilt-over mount and remote antenna tuner. An Oatey rain collar prevents water from penetrating the deck-post sleeve cap.

Common Mode Noise Suppression

Seventy-five feet (23m) of LMR-400 [64] runs in buried conduit from the vertical to the shack. Control cable consisting of 75 feet of Carol C2543A, #18 AWG/4 with a shield [65] (try eBay for short lengths of this cable), has been pulled through the conduit, as the Stockcorner JC-4 [66] remote tuner requires a manual tune command as well as DC power. Line chokes, Figure 17, have been connected at each end of the LMR-400. These line chokes were tested after construction, and the procedure and test results have been provided in a separate article. As might be expected, the noise floor of the system is greatly improved. The photo below shows 11 turns of RG-400/U wound on a stacked 2 x FT240-43 core [67]. The 6th turn is a Joe Reisert crossover winding [68]. Whatever common mode choke is used, the real part of the impedance must be greater than the reactive part.

Figure 17. Homebrew Common Mode Choke. The choke consists of 11 turns of RG-400/U wound over 2 x FT240-43 ferrite cores.

Stockcorner JC-4 Remote Tuner Control Box Repackaging

The control box that was provided with the Stockcorner JC-4 remote tuner has been repackaged, Figure 18, with a 4-pin microphone connecter [68] and 2.1mm power connector[69] for convenience. These connectors are readily available on Amazon or eBay. The housing is a Bud Industries PN-1320-DG [70]. In the future, the radio will control the remote tuner.

Figure 18. Remote Antenna Tuner Control Box. (Top) The tune command is asserted by depressing the red push button. The A or B antenna is selected with the toggle switch. (Bottom) The tuner is powered through the 2.1mm DC connector on the side of the box. Control signals are routed to the tuner through a 4-pin microphone connector.

References

Antennas, Featured, General, Homebrewing, Newsletter

Magnetic Loop Antenna

April 11, 2022 Martin Blustine

Recently, a magnetic loop antenna was fabricated for use by my brother, Allen, W2GYD, in his NYC apartment. The smallest dimension of his apartment window is 24″, and that sets an upper limit on the loop diameter. The direction of maximum directivity lies in the plane of the loop. There will be deep nulls perpendicular to the plane of the loop.

A diameter of 23″ is just right for standard coax lengths of 6′, 12′, and 18′. The coax is supplied by DX Engineering, type 400MAX[1]. The 6′ length works well for 1-turn, Figure 1, while the 12′ and 18′ lengths were coiled into 2-turn, Figure 2, and 3-turn, Figure 3, loops, respectively.

Figure 1. Single Loop, 6′ (183 cm) Configuration

Figure 2. Double Loop, 12′ (366 cm) Configuration

Figure 3. Triple Loop, 18′ (548.6 cm) Configuration

Pipe clips[2] were acquired from a plumbing supplier in the U.K., as 10mm plumbing supplies are readily available, and 10mm is a good match for the 0.32″ O.D. of the 400MAX coaxial cable. Pipe clips are employed to maintain the shape of the loops and to fasten the loops to the oak frame. The frame was constructed from red oak hardwood flooring. This serves as a stable measurement fixture for the loops.

A shielded coupling loop, Figure 4, was constructed from RG-8X, also supplied by DX Engineering[3]. The coupling loop is 14.4″ (36.6 cm) long, and it provides good coupling performance. The total length is a combination of interconnects within the Coupling Loop Assembly, Figure 5, the adapters, and an external 12″ length of RG-8X coax. The Bud Industries housing is PN-1331-MB[4].

Figure 4. Coupling Loop. The position of the loop is vertically adjustable using the sliding block and pegs.

Figure 5. Interior View of Coupling Loop Box. Only the outer shield carries RF.

The capacitor, Figure 6, in the Main Loop Tuning Assembly is an M73-type that was custom manufactured by Oren Elliott Products[5].

Figure 6. Custom-Made Oren Elliott Capacitor with Reduction Drive

The Bud Industries housing is PN-1333-MB[6]. As installed in the Main Loop Tuning Assembly, Figure 7, the minimum capacitance is 16 pf, while the maximum capacitance is 428 pf. This provides adequate range to tune the 1, 2, and 3-turn loops over a very large range from 80m through 12m. The minimum capacitance prevents tuning of the 1-turn loop to 10m, and a coax shorter than 6′ is required.

Figure 7. Oren Elliot Capacitor in a Bud Industries Enclosure. The reduction drive and tuning knob are visible to the right. An insulated tuning shaft protects the operator from high voltage.

Some adjustment of the position of the coupling loop is highly desirable. This has been verified by direct measurement of all loops on some or all bands from 80m to 12m. A sliding block arrangement, Figure 8, provides adjustment of the coupling loop vertical position with respect to the main loop. The holes in the sliding block and channel were drilled with a doweling jig[7]. Without this device, it is very difficult to drill a series of holes in a straight line. The sliding block is fixed in the desired vertical position with dowel pegs.

Figure 8. Sliding Block Assembly with Coupling Loop. Holes bored in the sliding block provide adjustment.

The magnetic loop is coarsely tuned to resonance on receiver noise. Fine-tuning is performed with an antenna analyzer. A Diamond Antenna CX210A SPDT switch[8] switches the antenna between a QRP radio and the antenna analyzer. The switch provides 70 dB of isolation. This is adequate to protect the antenna analyzer from QRP transmitter power. This setup works well, and there is no need to connect and disconnect coax to provide precise tuning.

Oren Elliott Products also provides a 6:1 friction reduction drive[9] so that 3-turns are required to tune from minimum to maximum capacitance. The reduction drive is essential for fine-tuning.

Interconnects within the Main Loop Tuning Assembly Figure 9, marked High Voltage, are made between the connectors and the capacitor with a 1″ wide x 0.012″ thick copper ribbon supplied by Georgia Copper[10]. This material turned out to be very easy to handle. Thicker materials, 0.016″ and 0.022″, were difficult to pattern and install. Consequently, these thicknesses were abandoned. The 0.012″ material was clamped between two thin sheets of craft plywood obtained at Michaels[11]. The plywood makes it possible to drill the 0.625″ holes for the SO-239 connectors in the ribbon. Patterning was accomplished by drilling with successively larger drill bits until a diameter of 0.625″ was achieved. Without the plywood to constrain the ribbon the drill grabs and tears the copper, even when cutting oil and a low cutting speed is employed. More than one attempt was required, as it was not possible to grind the drill cutting edges for use on copper. Metal stamping or boring would be better for creating the connector pattern in copper ribbon.

Figure 9. Tuning Assembly Mounted on the Antenna Frame

Another view of the Main Loop Tuning Assembly is provided in Figure 10. The 10mm plumbing clips are visible and serve to space the turns. A roll of copper ribbon used to connect the tuning capacitor to the connectors on top of the Main Loop Tuning Assembly is visible in the upper right of the photo.

Figure 10. Frame and Tuning Assembly Under Construction

The sharp tuning null for the 3-turn loop on 40m is shown in Figure 11. For the 3-turn loop at a frequency of 7.093 MHz, the -3 dB bandwidth of the loop is 19 kHz for a Q of 373. The efficiency of the loop is 7.9% which may be expressed as -11.0 dB. Since the directivity is 1.8 dB, the resulting gain is -9.2 dB. On the higher bands, the efficiency improves.

Figure 11. Sharp Tuning Null of 3-Turn Loop on 40m

Martin, K1FQL

Magnetic Loop Antenna Related References

[1] https://www.dxengineering.com/search/part-type/coaxial-cable-assemblies

[2] https://www.toolstation.com/talon-hinged-clip/p90494

[3] https://www.dxengineering.com/search/department/cable-and-connectors/part-type/coaxial-cable-assemblies/coaxial-cable-type/rg-8x?cm_mmc=ppc-google-_-search-_-part-type-_-keyword&gclid=CjwKCAjwrqqSBhBbEiwAlQeqGrlMyXXkAQdLqkPrCkkJ39Ug9Pm_xCPB3jkHF42

[4] https://www.budind.com/product/nema-ip-rated-boxes/pn-series-nema-box/ip65-nema-4x-box-with-mounting-brackets-pn-1331-mb/ – group=series-products&external_dimension

[5] https://www.orenelliottproducts.com/product-category/air-variable-capacitors/m73-m90-m97/

[6] https://www.budind.com/product/nema-ip-rated-boxes/pn-series-nema-box/ip65-nema-4x-box-with-mounting-brackets-pn-1333-mb/ – group=series-products&external_dimension

[7] https://www.rockler.com/dowel-pro-jig?country=US&sid=V91040&promo=shopping&utm_source=google&utm_medium=cpc&utm_term=&utm_content=&utm_campaign=PL&gclid=CjwKCAjwrqqSBhBbEiwAlQeqGtcSEMJ9SNtgzGkl_7dxBONE-E3BjhOV-tR2Y5lS9HaNICw6kOyUOxoCvLsQAvD_BwE

[8] https://www.diamondantenna.net/cx210a.html

[9] https://www.orenelliottproducts.com/product-category/planetary-reduction-drives/

[10] https://gacopper.com/strap.html

[11] https://www.michaels.com/midwest-plywood-economy-bag-12ct/D508855S.html

Nashua Area Radio Society

All posts by Martin Blustine

Radio Amateurs Developing Skills Worldwide