Ground Station

Eggbeater antennas are quite simple in concept, but I have discovered that they require a bit of skill to design and put together. The loops are one full wavelength of whatever you are trying to receive, with a little extra thrown in to make bends and connections with. Specifically, the formula is 1005 / F (MHz). This is equal to 1005 / 145 = 6.93 feet or 2 meters 11.26 cm for the VHF band (also called the 2 meter band), and UHF is the same formula which comes to 1005 / 435 = 2.31 feet or 70.4 cm. You will need a little extra on the ends, so give yourself a little wiggle room here. An extra inch or two will do.

Your loop, should it have continued, completing the circle, should be precisely one wavelength. That said, you do not wish for those ends to meet, but the closer you get to a circle, the better off you are. Bend the aerial material sharply from this point to form attachment points for the feed line and phasing loop. The loops should be built from a substance that weathers well (I used 12 ga aluminum fence wire), and cannot tough at the top. To make sure of this, I used two small zip ties around each loop and through each other. When tightened they maintain the geometry while insulating the loops. Use black outdoor zip ties to make the whole thing (and your radio) last longer.

The phasing loop made from RG-68 coaxial cable is another item which requires precise cuts according to the output of a formula. It is a quarter wavelength, but radio waves travel at a different rate in a piece of coaxial line, so we have to take all these things into account. The formula is 246 x coax. velocity factor / F (MHz). This is equal to (246 x 0.86) / 145 = 1.46 ft or 44.5 cm for VHF or (246 x 0.86) / 435 = 0.486 ft or 14.8 cm for UHF bands. Again, from a practical standpoint give yourself some extra to attach things to and to strip back the coax.

The coaxial cable used for the phasing loop must be RG-68 due to two properties of this cable. At 90 ohms impedance, it is close to the 50 ohm impedance of each of the loops in series (2*50=100), and it contains the right properties such as the velocity factor to phase delay the circular polarized signals picked up by the loops and make it into a good signal for the receiver. Unfortunately, RG-68 is the cable the old Thin Net (10base2) networks used to run on when we still used coax instead of ethernet. Very little of it is still around, and this was the delay in getting the ground station up and running with a comfortable margin. The good news is I have bought a thousand foot spool of it for the kits, so if anyone needs any, just drop me a line and I'll make you a sweet deal.

Connecting the feed line (a piece of RG58 or RG6 (50 ohm) transmission line, the phasing line, and the aerials is simple, but turned out to be the most difficult to do. I soldered mine, but soldering copper and silver to the aluminum I used for the aerial material was a real pain. For the next version I will be using small ground lugs to make the electrical connections on the aerial loops and crimp terminals for connection to the phasing and feed lines.

The sequence to connect the various lines and aerials for right hand circular polarization looking down from the top of the antenna is as follows:

1. Starting with the aerial loop end at the 12 o'clock position attach both the center conductor from the feed line as well as the center conductor of the phasing loop.
2. Moving to the end at the 6 o'clock position, connect both the shield braids of the feed line and end of the phasing loop of which you attached the center conductor in the first step.
3. At the 3 o'clock position, connect only the shield braid from the other end of the phasing line.
4. At the 9 o'clock position, connect only the center conductor from the other end of the phasing line.

If you need left hand circular polarization, simply reverse the connections made in steps 3 and 4 above.

I was looking for an ideal antenna body. On a trip to Home Depot, I picked up a 24" long 3" diameter black PVC tube for the body of the antenna. The diameter was overkill, but it attaches firmly to a wall, eave, or other surface with a fence clamp. The black PVC was chosen because it is more resistant to UV weathering and looks better. For the prototype used in the demo, I simply cut 4 slots for the aerial loops to exit the antenna body at 90 degree intervals (cross cut) just wide enough for the wire used for the loops to friction fit, and deep enough for the wire loop to pass through with the cap fully seated. This will be unnecessary in the final version since it makes more sense for the loops to be screwed down instead of clamped in the body and soldered.

When I brought the pipe into the lab, I realized that not only was the diameter overkill, but the length was as well. I cut the pipe in half and it is much more portable. Theoretically, I could have cut the pipe in half again for a 6" length, but that would make it difficult to attach the ground plane, mounting hardware, get the aerial loops to exit the side within the proscribed distance from the ground plane and install the end caps. A minimum of an 8" length is needed for the UHF model, while I calculate that a 12" length is needed for the VHF model or a model which receives in both bands.

The ground plane is non-critical, except for four factors:

1. It must be connected to the signal ground (6 o'clock in our sequence above).
2. It must be level with the ground while perpendicular to the body of the antenna.
3. It must extend at least 1/4 of a wavelength from the body of the antenna ((1005/F)/4).
4. It must be mounted 1/8 of a wavelength below the bottom of the aerial loops ((1005/F)/8).

The lower the frequency, the more you can get away with murder in the construction of the ground plane. In VHF, you only need 8 aerial protruding from the body. For UHF, I used aluminum flat struts cut to length, bent, and fastened with an automotive hose clamp (which also held the ground wire). I covered the aluminum struts with aluminum window screening which is secured by folding over the excess and stapling the two sides together. for L-band or above, I would recommend the use of a solid ground plane.

The feed line on the prototype was terminated with a BNC connector, but any RF connector will do. Once all the connections were made inside the tube, the top and bottom were covered with the end caps. I did not glue them as I would for a production model since I expected to tear it down and rebuild the device multiple times.

Using a Realtek RTL2832u chipset USB HDTV receiver dongle, I attempted to receive the UHF signals the antenna was designed to receive. There were several issues, most of them dealing with specific hardware issues on the target laptop. It worked once under windows, and a windows update screwed up the USB driver or the way the software communicated with the dongle. Under Linux I had version compatibility issues in PERL. This is something I will be fixing in the next month, but simply ran out of time for the Yuri's night video. I am just glad I got a recording of it working before windows thought it was smarter than I was.

Lessons learned: Using an early netbook with some proprietary oddities and has seen some "love" over it's long live was probably not a good choice of host PC. I will be using a more conventional laptop for testing later in the month. While I was originally thinking about putting the radio dongle in the PVC pipe enclosure of the antenna, this would have been a bad idea when it comes to the temperatures I can reasonably expect the enclosure to see if mounted outside. Use coax to run between the dongle and the antenna, keeping the dongle in a temperature controlled room. When selecting the dongle, don't get one with a barrel of F-connector. chose instead an MMCX or other connector which will make good contact for a long time and where the connector can freely swivel instead of putting stress on the RF and USB connectors.

A gentleman from Bucketworks Makerspace asked about the noise floor issue. I discovered that if the laptop is running on batteries and the dongle is connected to the PC with a long cord (mine was 5 ft), the noise drops by a huge margin. This is due to hash noise from the digital circuits in the host PC which were filtered by the longer cable, as well as the switching power supply used on the laptop making noise. When the laptop ran on batteries, the noise floor dropped by a ton. I will have more precise measurements of the difference when I have GNU Radio running.

One of the biggest issues with the construction of the groundstation antenna as noted above is the attachment of the transmission line, phasing line, and aerials. The way it was done in the 1.0 prototype design was difficult in a well equipped lab, and unreasonable to expect a hobbyist to do it that way. We had to figure out something else.

While a ground lug would be ideal, allowing a bolt to be used to physically attach the lug and connectors to the body, most ground lugs have square ends which cause oddball issues particularly in higher frequencies such as UHF. One design, although pricey, doesn't have this issue. Panduit's ML8-CY ground lug is a rounded case design which accepts 3/16" hardware and wire diameters from #14 to #8 AWG.

Product page
Newark| $3.73
Mouser NA
Amazon $2.07
CableOrganizer.com $2.07