A VHF 1/2-Wave Center-Fed Dipole Antenna
for High Altitude Balloon Use
January - 2015

My ½-λ Center-Fed Folded Dipole and End-Fed J-Pole antennas are excellent performing antennas, but I wanted something that weighed less, was easier and faster to build and much shorter in length.

¼-λ whips are sold for HT (hand-held transceiver) use and some simply using a ¼-λ length of wire. But a ¼-λ whip on a HT isn't a complete antenna and only half of a ½-λ center-fed antenna and relies on capacitive coupling to the operator’s body for the second half of the dipole. Like wise, a ¼-λ antenna element on a car roof or with ground radials are only half of a ½-λ antenna that use the car roof or ground radials for the other half. But there is no one holding a tracker with a ¼-λ whip in a foam box high above the earth to complete the antenna so a ¼-λ wire 'counterpoise' must be provided which means an antenna shorter than a ½-λ length isn't possible, at least not without sacrificing much of the performance of a proper ½-λ center-fed antenna (and similar to the performance loss one will experience using helical rubber duck and any other type of antenna shorter than ½-λ length).

For those wanting more information about ¼-λ and ½-λ antennas, check out these Antenna Notes For A Dummy and Antenna Theory pages that I found, are easy to read and have a fair bit of relative information.

A center-fed dipoles impedance varies with the angle between the elements and for in-line elements it's ≈ 72Ω, for a vertical ¼-λ ground plane with radials at 90° it's ≈ 23Ω and with the radials sloped down 45° it's ≈ 50Ω.

I began by building 3 antennas that were quick to build using 1/32" dia. music wire and one sided PCB. The PCB pattern was created using an exacto knife to cut the copper surface and then pulling off the unwanted copper while heating the copper with a soldering iron to soften the glue.

The 1st was similar to the familiar ¼-λ ground plane antenna with 4 sloped ground radials, but with only 2 radials; the 2nd was similar to the 1st, but with only 1 radial so simply a ½-λ center-fed dipole and the 3rd was similar to the 2nd except for having in-line elements.

The PCB used for the last 2 was 0.5" x 1.4" and they only weighed 5g or 15g with a 15" RG-316 feed-line with a SMA connector.

There was no real reason to build the 1st antenna, but it only took a few minutes and was used with the 2nd one and one with 4 sloped radials to confirm that the number of radials doesn't affect the impedance or SWR.

 

 

 

 

 

But impedance & SWR is affected by the angle between the elements and the 2nd antenna with elements at a 135° angle measured 50Ω with a SWR of 1.3 while the 3rd with in-line elements measured 70Ω with a SWR of 1.7, but it's hard to know what the actual impedance & SWR was with using a 36" feed-line rather than taking the measurements right at the antenna. A balun should also have been used to provide a balanced feed, but most consider the effect of using an unbalanced coax feed line to feed a 70Ω center-fed dipole directly too little to make any real difference. I'm sort of a perfectionist and would prefer an antenna with a much better match, but I'm also realistic and realize that even a perfect match would make little difference and having a simple easy to make antenna to use is much more important in this case.

The ground plan, or grounded element(s), of a ground plan or ½-λ center-fed vertical dipole antenna are usually positioned below the active element, but for HAB use the elements should be inverted, in other words, with the ground element(s) positioned above the active element so the signal is directed down to the earth where it's needed and will be received.

The 2nd antenna was used for BEAR-5 by simply letting it hang horizontally by the feed-line a few inch's below the payload box, but this didn't work out as well when we did the same for BEAR-6.

BEAR-6 tracker transmissions suddenly stopped being received at 25K ft. during descent and after arriving in the area where we figured the payload most likely landed, searching for 30 min. listening for a tracker transmission before receiving one and locating the payload nearby, we discovered why.

Wire, if allowed to flex where it's been soldered, will quickly break at that point and, as this photo shows, is exactly what happened to our antenna coax outer braded shield connection.

After the BEAR-6 problem I decided to make the PCB larger so the coax could be fastened in a way that prevented further coax problems. I also decided to test using a ferrite toroid core as a balun.

A ferrite toroid core on a coax feed line near the feed point of a balanced antenna acts as a RF choke and prevents the coax shield from coupling with the grounded dipole element and becoming part of the antenna and radiating power. Without such a balun, antenna impedance measurements become sensitive to the configuration of the feed cable and the antenna radiation pattern will be distorted by radiation from the cable, but with our tracker antenna feed line lengths being so short (< ¼-λ length in most cases) I didn't expect to see a toroid core make much difference, however I was wrong and it made using the antenna analyzer much easier by reducing the SWR and Impedance changes seen each time the analyzer was handled and/or moved. However I doubt any noticeable difference in the received signal would ever be seen when tracking a payload, especially with how much the antenna pattern is already affected by the constant antenna movement during a flight, but decided to use a toroid core anyway as this would also prevent coax movement and further coax problems.

My Latest HAB Tracker Antenna Design

I needed to make a number of antennas so I began by using a CAD program to create a cutting pattern that was copied and pasted as many times as could be printed on a sheet of paper. Each sheet was then cut up into individual patterns that were each glued to a PC board using Elmer's rubber cement to make later removal easy.

The PC boards were cut to the required size from larger sheets of single sided 1/16" (.062") copper clad fibreglass PC board material.

The ferrite toroid cores are nothing special and simply what I had salvaged in the past and had a hole dia. the same, or only very slightly larger, than the outer dia. of the RG-316 coax used. (I never toss anything without first saving everything from the item that I may possibly need someday and these cores with a small inner hole dia. are the type sometimes found used for RFI suppression on small dia. audio lines and other such wiring.)

2 music wire antenna elements, each about 21" long, will also be required along with some RG-316 coax and an appropriate connector.

Left - an exacto knife was used to cut through the patterns and transfer the layout pattern to the boards.

Center - the paper pattern is then removed and the cuts inspected to ensure they cut though the copper layer completely.

Right - The unwanted copper is removed by heating it with a soldering iron to soften the glue and then simply pulling the copper off. If there's a problem with getting a good transfer of heat from the soldering iron to the copper, simply use some solder and the heat will transfer nicely and make the copper easy to remove as the solder is flowed across the surface.

Left - Holes for the elements are drilled, the board is cut to shape and the opening the ferrite core fits into is cut out. Note that 2 sides of the opening are bevelled for a better fit and to help keep the core positioned while it's glued in place.

Center - One end of each element is bent into the shaped shown, the elements are installed through the holes and then the longer section of each one is bent down to lie flat against the backside of the board. Once soldered, the excess length of the shorter bent section of each element is cut off flush with the board using a dremel tool (as seen in the right hand image below of the board back side).

Right - The elements are soldered in place, the coax is fed through the toroid and then the core is glued in place with epoxy. It's important to have the feed line, or other small length of coax, fed through the toroid while it's glued to ensure it's positioned exactly right as a fraction too low and you wont be able to fit the coax through later or a fraction too high and the coax won't lie flat against the PCB.

The right hand image shows the back side of the board and how the ferrite core is installed partially through the board so the coax may lay flat against the board where it's soldered. This image also shows how the excess length of each element that protruded through the board has been cut off flush with the board.

You may have noticed that the photos show 3 different antenna styles. Above left is one with the active element sloped down, above center with it straight down and left is a combination of the two and the one I've built antennas for, and myself, have used the most.

The last 2 steps are to cut the elements to the right length so as to end up with each element 19-5/8" long from the center point between the two elements to each tip after providing some form of eye protection at the tips which otherwise could easily poke an eye out. These 2 steps take some planning of course to not end up with the elements a bit too short or too long. Forming a tiny loop and filling it with solder to ensure nothing can get snagged through it is the hardest, but provides the best protection and actually pretty easy once you get the hang of it. (Filling the loop with solder may be overkill, but I don't like to leave anything to chance - plus it looks nice.) Something much easier is to simply bend the end of each element back over itself and flow solder over the fold to smooth things out and give the end a more finished look. A nice big round ball of solder at the tips could also be used, but forming a ball rather than simply tinning the end of an element is much harder than it looks and takes some skill and practice.

For additional element length details see this detail.pdf and feel free to download and use, modify or whatever my PCB Layout Sheet.pdf.

A few additional notes:

I started out using .031" dia. music wire to save weight, but SABLE-4 identified a problem with that and now use .039" dia. (17 gauge) music wire.

Music (some call it piano) wire is easiest to find in most any hobby store dealing with R/C aircraft. It's sold in 36" lengths and ≈ 21" is needed for each element so there's a fair bit of waste if building a number of antennas and you may want to find a different source like I did. Piano tuners buy piano wire in large spools and I found one close by that was willing to sell me several hundred feet.

Always use RG-316, or a similar type teflon coax, that isn't affected by heat and melt when soldered.

I've been making our payload boxes from 1-1/2" styrofoam and installing the antennas by simply cutting a slot in the foam, making a hole midway along the slot large enough for the SMA antenna connector to pass through and after the SMA connector and feed line has been fed through the hole into the box simply pressing the antenna PC board into the slot until the pear edge of the board is flush with the inside of the box.

Knowing antenna theory and that an antennas impedance and SWR is optimum is all well and good, but in the end, the only thing that really matters is how well the antenna performs in actual use and after a number of flights I've seen no noticeable difference between using these center-fed antennas with either in-line elements, elements at 90° or elements at 135° to each other.

Most don't consider short rubber duck type antennas < ¼-λ length to be a real antenna, at least not one suitable for HAB use and a problem with any full size antenna is that they are too large to fit inside most payload boxes and the elements almost always end up becoming tangled in the lift line during post burst chaos and descent. I've considered setting up a UHF i-gate station for our balloon flights so smaller antennas that would fit in payload box could be used, but after seeing no difference between using my antennas with in-line, 90° and 135° elements I decided to try something else first on our BEAR-13 flight.

The payload box was a 13" cube made from 1.5" thick foam and I decided to try simply bending the elements as needed to fit the box. The yellow arrow points to where the antenna PCB was located. The blue line shows the path of the grounded element along the top inside corner of a side wall and then, after a 90° bend, pressed into a slot cut down the center of the front wall top edge. And the red line shows the path of the active element down the back inside corner, through the foam bottom and then, after a 90° bend, pressed into a slot cut in the bottom foam next to the rear wall.

A back-up tracker with an in-line antenna was used in addition to the payload tracker with the bent-to-fit-the-payload-box antenna during the flight and I was surprised to see no difference between the two in the number of position reports logged by aprs.fi or we received directly while tracking the payload. As a result, I'm planning to make future payload boxes no smaller than a 11" cube and install a bent-to-fit antenna in each as the boxes are made.

Some of the antennas made before running out of salvaged ferrite cores.

There may be a better core material to use than whatever my salvaged cores are made of and with no more cores to use I've been searching for what's best to purchase.

Selection is very limited with needing needing a specific inner dia. hole size, but I managed to find one the same size as one of my salvaged cores that's specified for VHF use and is an Amidon FB-31-0202.

Be sure to also see Part 2  - A VHF 1/2-Wave Center-Fed Dipole Antenna for a HAB Back-Up Tracker
To BEAR Home Page